It does if there is no PF correction to compensate for it. Look at it this way: If you suddenly added 1,000,000 of these devices to the grid then you most certainly *would* have to generate an extra 100MW to handle that (if you want to maintain the same output voltage). Until such time as the power company figures out there is this extra poor PF load on the grid and installs more PF correction gear to compensate for it. Without the PF correction gear there is no free lunch. And the further up the grid you install the PF correction gear the more losses in the wires and transformers further down the grid.
@@EEVblog nope, it does not. The generator will not require 100 MW of additional power at the *shaft*. It will require an additional power equal to the increased current by the resistance of the wires (and windings).
@@vihai Where does this magical additional power come from if not from the generator and it's fuel? And if the generator is producing the same voltage, that increased current *must* translate to real power at some point.
@@EEVblog it comes from the rotor (+turbine) inertia for half cycle, it goes back for the other half cycle, that's why you need some rotating mass as flywheel for apparent power "production".
@@vihai Ok, but surely this translates into a larger generator? Or more generators if that one is already maxed out. At some point there must be no free lunch.
I've been the designer for several consumer products, and the pressure to reduce BOM costs is completely INSANE! I've attended dozens of meetings and wasted hundreds of hours all for the reduction of 12 cents on a product. Since I opposed this insane cost reduction, I no longer work in the design of consumer products.
That is because there is one of those insane "economists" in the company. When a device is sold for amount of X and the BOM cost is Y, those people calculate that the factor between turnover and BOM cost is X/Y and therefore the effect of reducing the BOM cost by 12 cent will be X/Y times higher on the profit margin. That is of course a big delusion, but you cannot hammer that out of an "economist".
The whole fun of working in this area is to design a good product and still be able to squeeze every possible cent out of the BOM. It's not that hard to design things without budget constraints, boring.
But then you get the other end of the scale where a product seems to do wonders with very little in it and we all sit around gwaking at it to work out how the heck does it do that.(ie:ZX81s running video without any video chip)
I keep trying to do the calculation for you, but my calculator keeps spitting an error...seems anything divided by zero gives calculators the heebeejeebies.
Smoke alarms are put in by builders and builders couldn't care less how much you will be paying for power. So if they can save a few nickels with a poor design, they'll do it every time...
All ways presumed that these things were low energy every thing else we fit is supposed to be glad Dave’s done this vid is there no regulatory body that’s supposed to sort this out and check Electricians that fit these know minimal and micro electronics Yer sure they taught about power factor but they don’t check it or have the meters to check it that I’m aware of Great video Dave opened my eyes
@ats89117 Absolute truth. I imagine over the term of a 30 year mortgage, replacing 5 smoke detectors with better, lower power ones would give you a net cash back in energy savings.
@@anthonyselby8337 Again, multimeters with power factor are more expensive, testing takes time and that is more expensive. Not going to happen in the construction industry, no matter how well taught the electricians are.
@@chrisw1462 - over a THOUSAND years, replacing a THOUSAND smoke detectors with better lower power factor ones would save exactly ZERO dollars in energy savings. Domestic customers are not billed in kVA. Commercial customers have lagging PF (from motors) so the capacitors in soke detectors etc make the PF better, not worse.
inductive droppers??? Proper PFC + LLC switching converter will be cheaper, take less physical space and cost less than 1/2 of inductive dropper with even lower standby power. Standart flyback converter or PFC flyback converter versus inductive dropper for low power products - there is no contest there
In fact in older days we had problem with too much inductive load: quiscent power of transformers, luminescent lamps with inductive ballast, motors etc. If this is still true, these capacitors actually take part of PF corrector, because industrial PF correctors are EXACTLY huge capacitors with commutators to turn on as many as needed.
@@PhilfreezeCH PFC LLC setup will be cheaper than a inductive-dropper, forcing ( 1/2 devices cap dropper + 1/2 devices inductive dropper ) out of existence
@@volodumurkalunyak4651 Why would anyone design a Flyback for that? Non-isolated buck is the way to go. You can even try to find the one that syncs the current waveform with the input voltage for active PFC... This beats the transformer cost easily.
@@EEVblog Makes sense. I'm in Los Angeles and we have the same kind of absurd building codes here. Want to add a garage to your 100 year old Crafstman home in Pasadena? Well, your garage has to be engineered to handle an 8.2 on the Richter scale centered directly underneath you.
@@EEVblog In Germany you just need smoke detectors with a lithium battery under the newest regulations (not a rechargeable but a so-called "10-year lithium" battery). I can't believe we actually got at least one new regulation here that makes more sense than regulations in other countries.
"It is a safety product" -- I'm curious how frequently you'd hear this argument to justify the poor design characteristics that are unrelated to the actual safety.
Kidde smoke alarms aren't even reliable. The last one I took apart had the wires to the speaker routed right over the big 1/2 W resistor, which melted the insulation and shorted out the wires. So if anything the poor design makes them less safe.
@@jayleno2192 Sad... On the bright side, maybe it'll _start_ a fire, then, if still working, _detect_ it and warn the occupants... :-( Thereby demonstrating it's usefulness and justifying it's existence. ;-)
Mine are powered by a battery, which last 10 years, so no 230 V supply. They work, if something is smoking on the stove they switch on. Such products can be also found in other places, I have already seen a TV antenna amplifier which was rated for 150 Watts (which means it consumes 1300 kWh per year - a unbelievable waste of energy)
@@cambridgemart2075 I just had the optical Ei Electronics 650 smoke alarm beep at me while i was cooking in the middle of the night just a couple nights ago. I made quite a jump to silence it, luckily at least the button is huge.
There are ones out there that only react to heat. They ignore cold particles. Most of them are dual sensor and you can disable the optical one. They are designed to be used in kitchens or other areas with steam or dust in the air.
Your exactly right; everyone thinks of just one running, not the millions around the world installed. Think of the millions of products doing exactly the same. No interest in conserving energy at all.
Omit the whole pcb? :D Actually best not suggest that or we'll end up with dummy fire detectors.. This is the sort of thing that need an energy rating and maybe legislation to set a max power draw..
5x 1 Watt smoke detectors vs one 60 Watt light bulb: 120Wh for 24 hrs of smoke detection = 2 hrs of light. Probably best to get those light bulbs in check.
My son just asked me, "Why don't we just turn it on when there's a fire?" As I stare at him, I'm thinking "saving up a college fund might be a bit too optimistic".
@@saddle1940 - Your son is a GENIUS. Let's use a Peltier (or similar Heat > Power device) to power up the alarm using the fire's heat for power! We could also include a small wind turbine to harness the energy in the convection currents.
@@johncoops6897 Thats pretty crappy smoke detector. When the peltier is heated up enough to produce power, there is already fire mayhem. Smoke detectors detect smoke, not fire.
@@Brian_Of_Melbourne Yeah, they need to work right-way-up, since they're mounted upside-down on a ceiling which is itself already upside-down. If you get the wrong model for your hemisphere the smoke falls out before it can be detected. I'm very fortunate: I'm an ex-pat Aussie at the moment living in Singapore, and I've found that we're so close to the equator that both types work fine here.
AFAIK contracts in France have been in kVA for quite a while (if not forever). However old meters were only able to measure kW so that’s what it was in practice. The new meters tripping on kVA instead of kW would actually be a good thing (force individuals to correct their PF) *if users were notified when they trip because of bad PF*
There is NO electricity wasted here. The capacitive dropper in these creates a LEADING power factor, which improves the inherently lagging power factor on the supply grid. Dave is wrong.
@@johncoops6897 No, he gives Watt and VA rating .. and even the Watt rating is way too high (because of the Zener diode solution). Additionally the ohmic loss (think supply line) increases if the current increases, which it also does if only the reactive power increases.
I'm typing this reply while on night shift at a production facility with no other employees in for the weekend, yet heat and light is on full for almost a square kilometer of buildings. A 2W smoke detector is the least of my worries.
Hmmm... I'll pass this onto my company, we're building a new campus and we're interested in total CO2 emissions and TCO of all operations. We put in a bunch of stuff that's supposed to be "energy efficient" it might actually be a good idea for us to audit the hardware the site planners have selected.
You should check the facts first. The use of capacitive droppers like these will IMPROVE your site's power factor. That will take the load of the expensive on-site Power Factor Correction devices that you will be installing to add capacitance and reduce your kVA loading onto the grid. Capacitance is capacitance... add it with these or pay thousands to add it anyway. Don't listen to Dave, he is clueless about this stuff, despite a squeaky voice and massive fanboi base.
@@lucaskendc - Of course loads with lots of distortion and harmonics (THD) will not be perfectly inductive or capacitive. However, in order to be significant to the power grid, those "distortion components" have to be associated with a large load. These smoke detectors are a tiny tiny load, so the THD is literally "a drop in the ocean" compared to a (say) a 150W switchmode PSU, even if it has higher PF and lower THD..
I think you're over-blowing the situations. You'd have a point if this was a grid of nothing but smoke detectors, but the truth is they share the grid with things like fans, A/C compressors, refrigerators and other devices that don't have a power factor of "1". In the case of residential power, the power company takes care of it by adding capacitors to the power line (since most of the problem is inductive loads, caps are the "anti-inductive" fix.). All that these smoke detectors do is shift the amount of correction by a small about. (Small because their current is small compared to motors.)
@@EEVblog Ok, but good engineering is the art of compromise. How much are you saving compared to the additional parts (and their impact on the environment)? I still think you picked the wrong poster child. I'd worry more about bigger electronics and modern LED lighting systems. Big Clive routinely talks about how bad the PF is on the bulbs he takes apart.
I wonder if the electricity companies are seeing a shift from the mostly-inductive load as it was in the past (requiring addition of capacitors) to a mostly-capacitive load today. Or maybe it is still mostly inductive, but far less than it was in the past. After all, many loads like mains transformers and small electric motors have been replaced by new designs that use high-frequency switching and present a capacitive load if not corrected well (in which case it will be resistive).
Great video. Addresses EXACTLY my concerns about power consumption of late. So i'm building my own switching power supplies with standby circuits for stuff more and more lately, its not rocket science, you go to a semiconductor catalogue and find a chip that allows the features and there is a plethora of stuff in the 'ù' amp range for hibernate/standby. Build that chip into the design. Its time people started lobbying the politicians: demand very strict energy efficiency. Revise it every 18 months as the stuff that keeps coming out is getting better and better. Another gripe: Don't want E-waste? Then mandate a 10 year manufacturers warranty for EVERYTHING. That'll elliminate dodgy power supplies, cheap electrolytic caps etc in no time ;-) Solutions literally sitting on the shelf!
There is no such thing as "generation" of reactive power. Apparent power just causes an increase in current in the transmission cables, transformers and generators. Of course the increased current causes losses due to resistances and ultimately some additional active power to be produced but nowhere near the total apparent power.
You could keep your bitcoin wallet up-to date with that kind of power... And they say Bitcoin consumes too much power... (yeah, i know mining will always take more.)
I've thought for some time that it would make sense to standardize around some DC voltage (12v?) to extra pins on the household mains sockets, and have a central AC-DC converter(s). You'd have to deal with volt drops around the house, so you might need more than one AC-DC converter in the house to keep the cable runs short, but you wouldn't need one per socket. If you had specified a nominal 12v it would actually be between 8 and 16v, and a max of 500mA maybe. That would be enough for clocks, smoke alarms, and the standby circuits in lots of the consumer electronics (TVs, CD players, microwaves clocks, oven clocks... etc). 2 main advantages: 1) The central converters could be much more efficient, maybe even specified in building codes. 2) It would reduce the cost and complexity of all devices that can take advantage of it, especially those that only need low current DC - a cheap small switcher and you're done. Unlikely to happen, but I think it's an idea with some merit...
the same here, most new lamps are low voltage LED connected to main power, millions are installed every year for example. If you have PV even better, avoiding losses converting to DC to AC
the voltage drop would be a big problem since people would not like to invest in the correct wire gauge to lower the voltage drop and it would be another wire running around your house making the wiring more complex and creating more confusion for people who can barely wire the line and neutral wires the correct way around.
Real eye opener here. So much waste. I wonder how many other crappy mains devices are out there. Perhaps make a section about this dave? We should all inform our local representatives to take notice so products like these come with a “waste” warning, or better still, are banned from the market unless they do it right.
Here in Europe we have energy labels per product. But I haven’t seen those on these “low” consumption devices. They only use it for big devices such as refrigerators. And I’m not even sure if that label incorporates the power factor to measure the actual consumption. Either way, I’ll try to ask for more info on this. Thanks a lot for reporting this! I always suspected that there was a lot of waste with these rectifiers, but never knew it was so massive.
Instantaneous power is P = V*I. In an AC system with a resistive load (PF=1.0), the V*I is always positive, thus the load is always consuming power. As soon as the PF is not 1, then there are parts in the AC wave where P = V*I that is negative. Because of the phase difference between I and V. Thus, I will be positive while V is negative for some instances of time, and vice versa. That means the load is supplying power for those negative instances. You still have the real I^2*R losses, as those are PF 1 loads. The generator, and all the infrastructure, must be able to supply the V and the I. So yes, if I becomes larger than what the generator is able to supply, you would need another generator (or tranformer). The reactive power on the other hand appears as a torque ripple on the generator shaft. Not ideal either. The real power, as seen by fuel usage, is only P = V*I*PF, PF the power factor as seen by the generator. The I^2*R losses will change the PF at the generator from the PF at the load. So, bad PF wastes extra I capacity that could have been used to perform actual work.
Hmm... I was thinking that for non-linear electronics load, negative V*I does not happen. Load is not sinusoidal. So V*I is either positive or zero. That will create torque ripple and generally mess things up for sure, but no extra I^2*R losses in down time. I think PF is somewhat misleading in those cases as "apparent power" is kinda BS in non-sinusoidal situations.
@@otherbasis8505 Non-linear loads does not preclude negative V*I. It just means that given a single frequency sinusoidal voltage, the current will have different frequency components than just the supply frequency. Also means that superposition does not apply. I have seen some add a PF based on the current total harmonic distortion to the normal lead/lag PF. Though that part I'm not too familiar with.
I simply cant thank you enough for this video. It is an important one. Governments need to take action. I have a modem in my house. It costs me almost $ 200 a year. The designers should care and it needs to be a topic.
Hey DAVE check the Power factor again...its leading ( capacitive ).....and it is actually better solution because its compensate our home power factor which is generally lagging in nature due to inductive load and nonlinear load like bridge rectifier ...so our generating station will get little help from each of this tiny capacitor....
@@absalomdraconis do you think that those zener can dip the power factor to .069 with 1 watt active power consumption . ??? It's connected in series with the mains capacitor and its main function to protect the output capacitor ... Its can't draw reactive power ...
100%! I just wrote a comment explaining the same thing, but it was "magically deleted". If @EEVblog wants to post a follow up, there is a very simple experiment he can do. Hook up the scope to show both mains voltage and current going into those smoke alarms (not easy - high common mode but small signal). Then compare that to a small linear power supply, or a traditional linear plug pack, with no load on it. (ie just the input transformer). In the tranny of the linear plug pack, the voltage will peak before the current (that's what inductors do), and in the smoke alarm the current will lead the voltage (you guessed it...capacitor...). When both are supplied from the same mains plug, these 2 effects will (partially) cancel out, ie the overall power factor of the 2 devices in parallel will be closer to one (current closer to being in phase with voltage) than either one of them individually. That's pF correction in a nutshell. And that is what the smoke alarm is providing. pF correction for the rest of your house. Note pf = cosine(phase angle between voltage and current) .... but that angle can be positive (inductive loads) or negative (capacitive loads).
I lived in a house that had Live and Neutral swapped on a few outlets. It caused a significant amount of voltage on the ground contacts of the coaxial cable. Never did figure out which one was the problem.
I discovered that the live and neutral were swapped on a light switch in my house when the socket went POP and showered sparks in the kitchen when I was trying to remove a broken light bulb base with a pair of needle nose pliers. I was glad that I'd decided to put on gloves at the last minute before starting work.
Here in western Europe we have symmetrical mains outlets - no way to tell which pin is live and which is neutral. So they cannot be swapped either :-) But what you describe there can occur nonetheless. It is usually not caused by "swapped live and neutral" but by a symmetrical mains filter in the equipment (small caps from either mains lead to chassis) causing the chassis to float to 1/2 the mains voltage when it is not grounded. Indeed, this can cause nasty surprises e.g. when you grab a grounded antenna cable and try to connect it to a device touching both the cable shield and the device...
Dave (as others have done) has just driven off of the road into the ditch. Homes and industry present a largely inductive (positive reactance) load on the power suppliers. Items like this and LED lamps with capacitive droppers (negative reactance) partly counter the existing positive reactance, they do NOT create a bad reactance. This is review is a dud. Ron W4BIN
Power factor correction really only applies to the device in question, and PF can even vary how the device is used. You can't simply slap a negative reactance into your system and expect it to help. Besides, with more and more devices that have negative reactance on your house mains (LED bulbs being more common), you can no longer guarantee that most reactance presented to power suppliers are positive.
@@wyrmofvt True. Our house is mostly capacitive. The few motor loads we have are sporadic (fridge, boiler pump, clothes dryer). Most of the time it is electronic stuff with switching PSUs or capacitive droppers.
Should we mention all the unnecessary ITO devices, cheap TV's and video boxes, computers that stay idle instead of powering off (my laptop refuses to turn off completely when powered from main, it draws a continuous 12W)... Edit: oh! And I forgot, Wi-Fi connected lightbulbs... I'm glad I'm running mostly on off-grid solar. As long as the sun shines, I can waste energy.
I often wonder if LED bulbs suffer from similar issues with low power factor and high peak currents on the input, not to mention the environmental impacts of the built materials (plastic and components etc). It's like we've traded lighting power consumption in return for having more e-waste to deal with, I'm not sure how LED bulb failure rate is now but a few years ago it was a pretty high.
The problem with the smoke detector is that it has to use a capacitor dropper that is sized for the max current draw (when the alarm is sounding) When the alarm isn't sounding, it still draws that power and just dissipates it as heat. A similar power supply on an LED light bulb will be sized for the light bulb itself- there is no appreciable excess. Also, it's only drawing power when the light is in use. The smoke detector draws power constantly. A capacitor dropper is fine for something like an LED bulb that is either on or off. But in a device that only needs a tiny amount of current most of the time, and occasionally needs a lot more, there should be a switching power supply. But that's more expensive than a dropper, which always draws the same amount of power whether it's needed or not.
Could it be that they chose such a "simple" power supply design because it's simplicity makes it more robust and reliable which is an important consideration in what is basically equipment used in a life safety application?
Whilst the argument might have some validity, in the real corporate world of "products built down to minimum BOM cost for maximum profit margin" the decision was probably made because capacitive dropper supplies are as cheap as it can get. A passive transformer-based linear supply would use minimal extra components, would be just as reliable as a cap dropper (if not more so), and could easily be designed to have better power factor. But such a design choice would add 20-30 cents to the BOM cost per unit.
29 years ago i went to a training for reliability , the closing speaker ( from rocky mountain institute) was addressing the issue of designing for energy economy , telling that it our responsibility as engineer to design energy efficient product especially when they are mass produced . he was given the exemple of instant start crt based TV that use 6 watts standby to keep the tv ready to turn on rapidly , he was saying that represent in USA the power of 3 nuclear power station just to keep TV ready to turn on rapidly !
Good one! I think its just so that you cant ignore the warning chirps as the battery goes flat. The mains guarantee you will be nagged to replace the battery.
It's always pissed me off how inefficient the mains versions of these things are, for exactly the reason you give. Never realised it was 65000 times - just knew it was 'really bad'. You are not the first person in Oz to notice: there is even a govt report about it: reductionrevolution.com.au/blogs/news-reviews/5842566-99-waste-the-unexpected-energy-consumption-of-smoke-alarms I just tried and failed to find a mains one that was actually efficient. I also just complained to Which that their review of 'best smoke alarms' didn't even mention the power consumption. Without some consumer pressure they are not going to spend the extra dollar to make a decent PSU.
Interesting rant on the VA of the device, it may be overlooking that this is most likely Capacitive, and 180 from the Inductive power factor of the transformer supplies in the microwave, home router, and the inductance of the fridge, and other appliances with a motor, such as heat pump, AC, and furnace? Don't overlook the inductive load that is the utility transformer on the street feeding the house. This may actually reduce the total VA connected load of the house. Individually this looks bad, but it may fix some of the large amount of connected inductive load on the utility. Have you put current transformers in your main panel to find the total power factor of your house and if it is leading or lagging?
You could also have a look at the alarm clocks sold. Probably the same amount of units per house hold and it could be battery powered or even wound up ones. But there are lots and lots of them on the market which draw more power costs in a year than you have to pay when buying them. And as they have a "clock" in them, they don't need to be sub 0.5 Watt stand-by as other devices must comply to. (ever bothered to think why lots of devices have clocks these days? :) )
@@sasodoma As far as I know, it is at least an EU regulation. See for example: ec.europa.eu/info/energy-climate-change-environment/standards-tools-and-labels/products-labelling-rules-and-requirements/energy-label-and-ecodesign/energy-efficient-products/mode-standby-and-networked-standby_en
I bet the springy metal switch contacts on the Quell unit cost a lot more in terms of storage space and automation. I bet it's a lot easier to get production line fixtures for your surface-mount button than to fit those metal spring contacts at just the right angle.
You can actually use a high voltage depletion mode MOSFET to modify any "normal" regulator (7805,LD118,...) to support serveral hundret volts on the input.
@@absalomdraconis That really depends. In standby its efficiency is near perfect, with iq of a few microamps, under load it is as unefficient as you'd expect from a linear regulator =)
@@berniwa Don't forget that all idle current of your linear regulator is REAL power, as opposed to the APPARENT power that a capacitive dropper has. In the shown circuit only the power dissipated by the zenerdiode is real power, all other power is apparent power. But drawing even only 1mA in a linear regulator that operates on line voltage will cause a big real power usage. That effectively precludes any use of conventional devices like a 7805.
@@Rob2 Sure, with a 7805 the current draw will be problematic, but technology progressed since then and TIs 7805 cousins (like the TPS783xx) have an iq of 500nA. So the real power consumed by the regulator would be a moderate 115uW at 230V ;-)
@@berniwa But once the alarm starts beeping and draws 50mA or so, your solution draws over 10W of power and dissipates that all in the regulator/fet, while the capacitive dropper solution still only draws about 1W as it always does. Not important from energy consumption perspective, but the design has to accomodate the dissipation of those 10W at least occasionally.
years ago I have seen a TV antenna rectifier which was rated for 150 watts - 1300 kWh per year, or almost 400 € in Germany (it was installed in Germany).
Capacitive power supplies help correct the low PF of inductive motors. The big waste here is not the large VA but the use of linear voltage regulation from 240V to 12V.
I had some poorly designed battery operated fire alarms before, they'd chew through a 9v alkaline battery in under 2 months, whereas the good fire alarms i replaced them with can last a few years on a 9v battery. A good lesson to learn, never cheap out on safety equipment.
I have seen the struggle to understand reactive power among engineers working at power plants ! There is this option to set pf from SCADA system in the power plant I worked, we used to set it 0.95 and that will increase the excitation current of the generator, meaning generators need to work more hard to make up the reactive power, thus needing more torque to rotate to maintain the same rpm (frequency) and more fuel. I am talking about generation sides , not load sides.
@@ryanb1874 some of the current generated is fed back into the rotor through a rotating transformer i think. When you need to start the thing up there should be enough magnetism left in the rotor to start generating enough current to get the whole thing going again. Smaller generators use slip rings and can be shunt wound/series wound and all that stuff I don't remember from uni...
@@ryanb1874 Well, it is but not directly ! The generators in the power plant I used to work had actually 2 generators in 1 genset shaft (made by ABB). Excitation controller will supply few Amps of DC to the stator of excitation generator G2, G2's rotor will make 3 phase AC which will be converted back to large DC current to feed G1's rotor with a 3 phase rotating rectifier, and finally the actual AC power comes from the stator of G1. Something like this - diary-of-electric.blogspot.com/2019/05/wiring-diagram-avr-of-3-phase-generators.html p.s. I am sorry if I explained something incorrect, it was almost a decade ago. Now, I live in embedded electronics world and care much less for electrical systems.
@@demoncloud6147 sounds complex, I was just watching a Don Smith video and the old fart could even get his Over unity, resonator to work. THANKS for the comment.
Why do you need a mains powered smoke alarm anyway? So you have a product there, with a backup battery, that you still need to check periodically and eventually have to replace. Might as well use it to power the device?
Out of curiosity, is there a benefit to overall reliability or certifiability of the circuit with the less-efficient design? I am just curious about other factors - apart from cost - that might have informed the design decision.
So in America, we have devices that just have (I think) a 24volt-ish line as well as a signal line to sound off the other alarms, as well as a 9v backup battery, but my home was built less than a decade ago, and older homes usually have just a 9v battery and the alarm line.
Power factor is difficult to understand, if the current is 90 degrees out of phase with the voltage the wattage is zero. The only power consumed is from the resistance in the wires, transformer windings and generator windings. You need to understand the difference between true power and apparent power
Hey @EEVblog, I just thought of another one. Bathroom extractor fans, mine draws a couple watts all the time since it's just a big dropper resistor and a zener diode to run the fan timer control circuit (I've seen them with capacitive droppers too). Every bathroom gets them now which is another necessary vampiric load in every home.
Been there, done that. You do it once and never forget. Though in my case the scope was fine afterwards, but the probes didn’t fare too well with a dead short on the shield conductor.
I am somewhat surprised to see how many electronic engineers somehow struggle to get their heads around the phase lag between voltage and current. The bottom line is that the overall lag in domestic installations is inductive (think motors fridges, etc. ). This is even more so for industrial installations. Adding a small capacitative lag will (slightly) improve the power factor!
uK8cvPAq Yes, I know that’s what Dave is saying, but here’s the thing... we don’t have a separate grid connection for each appliance, just one to the household. In most cases the overall household load is inductive, so biasing the load with a small capacitive load will actually *improve* the overall household power factor.
uK8cvPAq yes, I’d agree that we should fix the power factor. So, how will we do this? Again, the dominant problem is inductive load. This is prevalent in industrial applications, not so much in domestic. The standard “solution” to fixing an overly inductive load is to put a capacitor across the supply. That way, there’s a reserve of electrons to meet the out of phase inductive load. Oh dear, that’s where we started! Unfortunately, there’s a lot of BS on you tube on power factor. The only youtuber I’m aware of who calls it out is Big Clive. We need the EE community to take a stronger stand on this. Reactive power is not something else that needs to be generated. The only possible implication is *current* for line distribution companies, not power. If we have to have higher rms current to meet the power needs, we may need beefier wires. That’s a big if though! Here’s my challenge to any EE. Do the calculations. It might surprise you.
The trouble with zener-regulated supplies like this is that the series reactance has to supply the peak current demand, without dropping the voltage too much, which as you observe is in the ballpark of 15mA when the alarm is sounding.But even given that, as designed, it is drawing 70mA, so they could have used almost 4 times the reactance (reduce the capacitor value by a factor of around 4 to, say, 0.27μF) and get the same voltage drop with only 17.5mA of quiescent current, and it still would have been enough to run the alarm when sounding. I have to wonder if they are using exactly the same circuit for 120V systems. this would give a slightly less feeble justification for a whole 1μF dropper capacitor, though that would still be about twice as large as necessary and the circuit would still draw twice as much current as necessary
"If the actual product itself takes less than 50uA, why does the product have to take orders of magnitude more than that?" So the real reason for this is because this kind of power supply draws constant power (all the time, regardless of the actual useful load current consumption), and you have to design for the maximum load current that you'll ever draw. If you have a 30mA load to power a 2V LED, then that also results in a 30mA input requirement at the AC mains voltage as well--which is a lot more power than you need. If you also want to chirp the alarm while keeping the light on, and that takes another 30mA, you now have a 60mA design requirement, so you will always draw power equal to AC mains voltage * 60mA. Add another 10 to 15mA for the microcontroller, etc. Then When you also factor in that that a lot of the resources for transformerless power supply design don't even do the calculations correctly, you end up with overdesigned transformerless power supplies, or worse, ones that are reused from previous designs without optimization because "they just worked" Also, at 240V, your losses for a given load are 2x that of a 120V mains system, so that's unfortunate. 50Hz vs 60Hz doesn't matter much in this type of application because it only affects the size of the DC capacitor you put on the output. I've put together a number of helpful design resources on this very specific topic (mostly because I couldn't find anything good myself a few years ago) at www.designercircuits.com/transformerless.htm, including a document that walks through the calculations and different topology options, spreadsheets and schematics for both capacitive (like the smoke detector) and resistive version of transformerless power supplies. One of the conclusions I've come to is that this is really only a viable topology for loads in the 10s of milliamps, because beyond that, it is difficult to even reject the heat generated. But if you do need more current, a very clever approach that greatly improves the efficiency is to target an intermediate output voltage, like say 40VDC (meaning less energy is burned up in the input resistance), then use a switching power supply to efficiently step that down to lower voltages. This lets you get upwards of 100mA without much cost, because switch mode controllers in that voltage range are still pretty low-cost. Beyond that, it really is a matter of going into the design with ultra-low power consumption components. And avoid LEDs.
A quick addendum to the above; at 0.5W of real power consumption, that maps to about 2mA of current that gets consumed/dissipated. Still like 40x what is needed for a 50uA average load. A far better solution economically for the manufacturer and consumer would be to eliminate the capacitor (short across it), replace the 100ohm resistor with 1Megaohm resistor, and still have 100% margin in your load current (support for 100uA). Beef up the output capacitor to handle peak transients (beeping, LEDs, etc.)
So it looks like the most clever approach here will be to use a resistive transformerless power supply to run the logic needing 50uA constantly, and whenever it is time to beep, since 15mA is needed for that, have a separate transformerless supply that you turn on with an SCR/thyristor controlling that second transformerless supply from the low side. You still save money because you can still avoid the capacitor for 15mA, and the diodes, resistor, and SCR is still cheaper than the 1uF AC mains-rated capacitor.
He didn't even factor in hotels that have 100's of these sucking away power. Here in the US you go to any hotel and they have a smoke detector in every room I'm assuming it's the same in other countries. The amount of wasted power is insane considering the smoke detectors here and everywhere else are most likely nearly identical designs to those down under.
As an electrical engineer in the USA, there are so many reasons why 50hz is rubbish compared to 60hz. I think it would be interesting if you did a video comparing the two, and seeing your reasoning on why 60hz is rubbish.
Could you give a few reasons why 50Hz is rubbish? (To give an example of a disadvantage: 60Hz creates a bigger resistance and bigger losses in transmission lines).
Maybe a discussion about "what is beter, 115V AC or 230/240V AC for consumer mains outlets"...? Americans write that 115 is "safer" but I really question that. 115 is high enough of a voltage to get a dangerous shock, and with 115 the current of the average appliance will be double of what it is in the 230/240 system. The higher current contributes to fires due to bad connections, so 115 may be more dangerous as a whole.
The quiescent current in phone chargers is very low. They use proper buck converters to drop the voltage, so when there is no load on them, they are basically turned off. Power factor will be horrible because they are only loading the peak of the sine wave, but a low power factor on a ~zero load isn't a problem. Good ones use circuitry that follows the sine curve to give them a good power factor under load.
TVs have outstanding standby current. My dad is a service tech for Philips. 20 years ago, the training tech was HYPED about the low power consumption on standby on those sets. It was so low, the standby LED would remain powered for a good half hour after you unplugged it. They used a big capacitor so in standby the TV would just "burst" the PSU to charge the capacitor and go back to sleep at milliwatts again. My old-ish LG LED TV has a relay to completely disconnect (not just standby) the main supplies. Only the standby PSU remains powered in standby. The rest of the TV is OFF. And then the PACE cable boxes just blank the HDMI output when "off". The tuner and decoder circuitry runs 24/7 and the device is a nice space heater...
Yes...but a heat exchanger (your A/C unit) will do it more efficiently. But if you're just looking at financial cost, at even 1W, and estimate 10 cents per kWhr, then 1W * (1kW/1000W) * ($0.10/kWhr) * (24 hr/day) * (365.25 days/year) = $0.88 per year to operate a 1W smoke detector and incidentally heat your home.
@@davidhoelscher True. I have a heat pump and high efficiency gas furnace. I should probably work on shutting off electronics in the summer when I am cooling. In that case, the electronics are using electricity and I am then using more electricity to remove the heat from the house.
Gotta love how these environmental nuts whine about the craziest shit when things like THIS are being totally IGNORED as if it's all peaches and cream. Thanks Dave for highlighting this issue again!👍👍👍👍👍
What about the perils of mixing full wave and half wave rectifiers on the same incoming ac feed? If one smoke alarm manufacturer was to use a non-isolated half wave rectifier in their product and another chose to use to use a non-isolated full wave rectifier if their DC commons are tied together you will also be wanting to have on hand a battery operated smoke detector to detect the smoke the two hardwired ones will produce if they are connected to the same incoming AC feed. I think that may be why we dont see many non-isolated half wave and full wave rectified consumer products. If they have isolation transformers before the rectifiers then no problem but snce in most countries our neutrals are grounded back at the distribution panel keeping the DC commons from being connected is difficult without isolation of some kind before the rectifiers.
With things as low power as these, it's not an issue. And you'll notice that in his reverse engineered davecad circuit, the zeners simply short out one half of the mains cycle(through the capacitor and resistor), so the draw them is still reasonably balanced. Yup, toss out half the input current, then toss out most of the other half. Uugh, I hate bad design.
Dave's reverse engineered Davecad circuit is of the manufacturers current inefficient capacitive dropper circuit. I am not questioning the manufacturers use of capacitive droppers but rather Daves suggested use of half wave and full wave rectifiers instead unless they are also isolated from the mains. If the proposed rectifiers dont have isolation from the mains this could lead to mixing full and half wave rectifiers together on the same AC feed which will not end well. Adding the required isolation before the rectifiers would increase the cost of these smoke dtectors even more than Dave's proposed 250V linear regulators. In the case of a hard wired smoke detector you would also want to be connecting them together in some way to allow for one detector to set off the others in the same building. and this could very well lead to interconnecting their DC commons. Mixing a non isolated half wave rectified power supply with a non isolated full wave rectified power supply on the same incoming AC supply with their DC commons tied to ground will result in their diodes shorting each other out as they find a path through ground. Proper incoming isolation and grounding would need to be observed or the magic smoke will be released from the rectifiers diodes.
With an R-C voltage divider input to drop the mains voltage prior to rectification, won't the power factor be LEADING? Therefore, given that the vast majority of reactive loads are inductive, these smoke detectors would each serve to provide a small corrective effect on the grid. Or am I missing something here?
Hi Steve, Ever wondered how much power is used by all the installed 'smart meters'. I know this is not paid by the customer, at least not directly. I never succeeded in finding in the specs of smart meters how much power they consume. I remember that once you investigated a smart meter. If you still have that meter you might investigate how much this meters draw or how the power supply internally works. Keep up the good work.
I have seen something really "nice" in that regards: some sheds for bikes in a housing complex. To make it "more green" they included motion-sensors for the lights. The sensor takes ~2W continuous, the LED-Lamps installed are 3.5W .................. why not just a simple door contact switch with a timer. Oh yeah - nearly forgot to mention - there is not just one sensor - there is one on the outside and one on the inside. And the sheds are built in such a way that small animals like cats can easily leave them, and of course there are many cats that constantly trigger the sensors.
Reminds me of the neon nightlights with the sensor to turn them off during the day. The sensor is just a phototransistor in parallel to the neon bulb, so they consume more current when off than when on. But hey, it _looks_ like they're saving energy...
Dave is wrong about the reactive volt amps on the power grid with the low power factor, power companies use capacitors on the primary power lines to correct the power factor back to the generators, so the amps on the wires from the reactive power only affects the wires and transformers from the source to those capacitors.
Something like that. Always wanted Thunderfoot or Dave here to do the math to see how much CO2 was pumped into the air in the production and viewing of the videos to plant 20 million trees (one time, rather than the 20 million trees, twice a day, every day, it would take to make just the US carbon neutral. Oh, yeah, also have to be magic trees that grow to full size in a day.)
Spring lever clamps are way more satisfactory for retaining tension on the wires in the long term. Old installations often have loose connections where the Copper has relaxed and the screw clamps aren't holding them tightly any more.
Speaking of which, that NBN modem I received gets HOT. Do you think anyone anywhere calculated how many extra power stations we need in Oz once we install them in every single household? And I'm pretty sure it has no fundamental reason why it needs to use any of that power, especially when idle.
I wonder why to choose a linear regulator? My understanding of linear regulator is, that they just "burn" the input voltage minus the output voltage with the output current into heat. Like the old 78xx variants. I always thought you would need a switching regulator for such high dropouts. In case of 450V input, 15V output and 80mA input = output => this results in (450V-15V)*80mA ~ 35W - way more than the original "bad" design.
No, with the high voltage regular the load is only 50uA + the 10uA or so quiescent current of the regulator. The zener is a shunt device that needs higher current to regulate it's voltage.
Because he's the co-host of the world's biggest electronics engineering podcast, and for him only the active power factor matters, not the deformant power factor, which is mainly because of semiconductor rectifiers which introduce a whole lot of odd, in-phase harmonics into the grid, and is the reason why his magnificent 3-phase power analyzer shows that aberrant ACTIVE power factor.
The reason for such a power supply is obvious to me: What do you need to run the alarm: If you think microamps, you are fooling yourself. Also, who wants to design two supplies. Better design is certainly possible though. Put a switch and an additional capacitor somewhere and you could drop the quiescent power most of the time.
I always thought that a capacitive dropper could be very efficient if matched perfectly to the load. Perhaps they did the calculations based on the current draw when the buzzer is operating continuously so it's extremely miss-matched when operating in sensing mode (99.99999% of the time) and almost all of the power is being wasted away by the zener diode. If that's the case could they have the microcontroller operate a MOSFET to switch in a different value resistor or capacitor when in alarm mode? That way it would change to being extremely miss-matched and inefficient only 0.00001% of the time.
The apparent power is not real, it does not have to be produced. It does, however, cause losses in the wires because of the resistance.
It does if there is no PF correction to compensate for it.
Look at it this way: If you suddenly added 1,000,000 of these devices to the grid then you most certainly *would* have to generate an extra 100MW to handle that (if you want to maintain the same output voltage). Until such time as the power company figures out there is this extra poor PF load on the grid and installs more PF correction gear to compensate for it. Without the PF correction gear there is no free lunch.
And the further up the grid you install the PF correction gear the more losses in the wires and transformers further down the grid.
@@EEVblog nope, it does not. The generator will not require 100 MW of additional power at the *shaft*. It will require an additional power equal to the increased current by the resistance of the wires (and windings).
@@vihai Where does this magical additional power come from if not from the generator and it's fuel? And if the generator is producing the same voltage, that increased current *must* translate to real power at some point.
@@EEVblog it comes from the rotor (+turbine) inertia for half cycle, it goes back for the other half cycle, that's why you need some rotating mass as flywheel for apparent power "production".
@@vihai Ok, but surely this translates into a larger generator? Or more generators if that one is already maxed out. At some point there must be no free lunch.
I've been the designer for several consumer products, and the pressure to reduce BOM costs is completely INSANE! I've attended dozens of meetings and wasted hundreds of hours all for the reduction of 12 cents on a product. Since I opposed this insane cost reduction, I no longer work in the design of consumer products.
LOL, well done!
That is because there is one of those insane "economists" in the company.
When a device is sold for amount of X and the BOM cost is Y, those people calculate that the factor between turnover and BOM cost is X/Y and therefore the effect of reducing the BOM cost by 12 cent will be X/Y times higher on the profit margin.
That is of course a big delusion, but you cannot hammer that out of an "economist".
The whole fun of working in this area is to design a good product and still be able to squeeze every possible cent out of the BOM. It's not that hard to design things without budget constraints, boring.
But then you get the other end of the scale where a product seems to do wonders with very little in it and we all sit around gwaking at it to work out how the heck does it do that.(ie:ZX81s running video without any video chip)
I feel strongcontempt towards your ex employer
Maybe the companies asked for a low Iq but that was misinterpreted and they hired lazy engineers instead.
Haha. Good one.
Lmfao
So, how many kilometers of solar roadways would be required to power 5 of these devices?
I keep trying to do the calculation for you, but my calculator keeps spitting an error...seems anything divided by zero gives calculators the heebeejeebies.
Are trucks driving on this solar roadway?
@@Okurka. Are trucks driving on your roof?
I have a GREAT idea!!!! We could use the baterizer!
@@Okurka. Can you stop a truck on that one panel
*You know im being a fuckwit for a reason*
: P
Smoke alarms are put in by builders and builders couldn't care less how much you will be paying for power. So if they can save a few nickels with a poor design, they'll do it every time...
All ways presumed that these things were low energy every thing else we fit is supposed to be glad Dave’s done this vid is there no regulatory body that’s supposed to sort this out and check
Electricians that fit these know minimal and micro electronics
Yer sure they taught about power factor but they don’t check it or have the meters to check it that I’m aware of
Great video Dave opened my eyes
@ats89117 Absolute truth. I imagine over the term of a 30 year mortgage, replacing 5 smoke detectors with better, lower power ones would give you a net cash back in energy savings.
@@anthonyselby8337 Again, multimeters with power factor are more expensive, testing takes time and that is more expensive. Not going to happen in the construction industry, no matter how well taught the electricians are.
@@chrisw1462 - over a THOUSAND years, replacing a THOUSAND smoke detectors with better lower power factor ones would save exactly ZERO dollars in energy savings.
Domestic customers are not billed in kVA.
Commercial customers have lagging PF (from motors) so the capacitors in soke detectors etc make the PF better, not worse.
@@johncoops6897 I wasn't talking about VA vs. watts. Just plain power consumption.
Make half of the smoke alarms use capacitive droppers, and the other half use inductive droppers, problem solved!
Haha. That might correct the VA, but you still have ~1W wasted in the zener
inductive droppers??? Proper PFC + LLC switching converter will be cheaper, take less physical space and cost less than 1/2 of inductive dropper with even lower standby power.
Standart flyback converter or PFC flyback converter versus inductive dropper for low power products - there is no contest there
In fact in older days we had problem with too much inductive load: quiscent power of transformers, luminescent lamps with inductive ballast, motors etc. If this is still true, these capacitors actually take part of PF corrector, because industrial PF correctors are EXACTLY huge capacitors with commutators to turn on as many as needed.
@@PhilfreezeCH PFC LLC setup will be cheaper than a inductive-dropper, forcing ( 1/2 devices cap dropper + 1/2 devices inductive dropper ) out of existence
@@volodumurkalunyak4651 Why would anyone design a Flyback for that? Non-isolated buck is the way to go. You can even try to find the one that syncs the current waveform with the input voltage for active PFC... This beats the transformer cost easily.
So, just get the 9 volt battery operated version and change the batteries when they chirp every 5 years or so.
Yep. But here in Australia if you have a new house or renovation you have to install the mains one otherwise the job won't get signed off.
@@EEVblog North America has been doing this for years.
@@EEVblog Makes sense. I'm in Los Angeles and we have the same kind of absurd building codes here. Want to add a garage to your 100 year old Crafstman home in Pasadena? Well, your garage has to be engineered to handle an 8.2 on the Richter scale centered directly underneath you.
@@EEVblog In Germany you just need smoke detectors with a lithium battery under the newest regulations (not a rechargeable but a so-called "10-year lithium" battery). I can't believe we actually got at least one new regulation here that makes more sense than regulations in other countries.
5 years.....yeah right, dreaming. Dreaming that's interrupted by a CHEEEP in the middle of the night.
"It is a safety product" -- I'm curious how frequently you'd hear this argument to justify the poor design characteristics that are unrelated to the actual safety.
You could power several computers that would be able to take a human rated space capsule to moon with the power that these things require.
.
Kidde smoke alarms aren't even reliable. The last one I took apart had the wires to the speaker routed right over the big 1/2 W resistor, which melted the insulation and shorted out the wires. So if anything the poor design makes them less safe.
I know right
@@jayleno2192 Sad...
On the bright side, maybe it'll _start_ a fire, then, if still working, _detect_ it and warn the occupants... :-(
Thereby demonstrating it's usefulness and justifying it's existence. ;-)
Mine are powered by a battery, which last 10 years, so no 230 V supply. They work, if something is smoking on the stove they switch on.
Such products can be also found in other places, I have already seen a TV antenna amplifier which was rated for 150 Watts (which means it consumes 1300 kWh per year - a unbelievable waste of energy)
Actually, "Chip of the week" segment is a great idea for EVblog2! ))
When are the EEVblog brand smoke detectors gonna hit the market!? :D Nice work Dave.
@@Okurka. How do you figure that adding a regulator to the circuit would require firmware?
EEVSmoke(TM) (that could potentially also be a brand of cigarettes lol)
Electric ones.
With Bluetooth, so they need firmware.
Which will need upgrades.
The EEVBlog smoke detector only detects magic smoke
all this talk about smoke alarms...when are they going to make a smoke alarm that doesn't embarrass me when I'm cooking?
They do, swap out to optical ones as they don't give nuisance alarms, unlike the ionisation type.
@@cambridgemart2075 I just had the optical Ei Electronics 650 smoke alarm beep at me while i was cooking in the middle of the night just a couple nights ago. I made quite a jump to silence it, luckily at least the button is huge.
There are ones out there that only react to heat. They ignore cold particles. Most of them are dual sensor and you can disable the optical one. They are designed to be used in kitchens or other areas with steam or dust in the air.
LOL at the "woosh factor" in the replies above.
So, you're looking for a smoke alarm that gives cooking lessons?
That "chip of the week" might make a nice new video series, Dave.
Your exactly right; everyone thinks of just one running, not the millions around the world installed.
Think of the millions of products doing exactly the same.
No interest in conserving energy at all.
Its the consumer way!
Omit the whole pcb? :D
Actually best not suggest that or we'll end up with dummy fire detectors..
This is the sort of thing that need an energy rating and maybe legislation to set a max power draw..
I wonder if we could make a 250VAC rated smoke sensor
Yeah, so everybody is looking for energy effecient lights, but their smoke detectors might piss away watts 24/7
5x 1 Watt smoke detectors vs one 60 Watt light bulb: 120Wh for 24 hrs of smoke detection = 2 hrs of light. Probably best to get those light bulbs in check.
My son just asked me, "Why don't we just turn it on when there's a fire?" As I stare at him, I'm thinking "saving up a college fund might be a bit too optimistic".
@@saddle1940 - Your son is a GENIUS. Let's use a Peltier (or similar Heat > Power device) to power up the alarm using the fire's heat for power! We could also include a small wind turbine to harness the energy in the convection currents.
@@johncoops6897 Thats pretty crappy smoke detector. When the peltier is heated up enough to produce power, there is already fire mayhem. Smoke detectors detect smoke, not fire.
@@kapu4369 - whooooosh! Over your head, huh?
6:23 Dave needs a Cliff Quick Test....
I'm sure BigClive could send one. Although it will have to be the southern-hemisphere version, these can be harder to source.
@@Brian_Of_Melbourne Yeah, they need to work right-way-up, since they're mounted upside-down on a ceiling which is itself already upside-down. If you get the wrong model for your hemisphere the smoke falls out before it can be detected.
I'm very fortunate: I'm an ex-pat Aussie at the moment living in Singapore, and I've found that we're so close to the equator that both types work fine here.
Here in France with our new connected smart meters we now pay for VA and our contracts were subtly altered to kVA from kW! So it hits our pocket!
So big clive's prediction is coming true, then.
@Dingo Nates Yikes, do you have a source?
The same meters are used in Denmark.
Not nice. Here in Germany, we still pay for KW. But therefor, our KW-price is about 0,30€.
AFAIK contracts in France have been in kVA for quite a while (if not forever).
However old meters were only able to measure kW so that’s what it was in practice.
The new meters tripping on kVA instead of kW would actually be a good thing (force individuals to correct their PF) *if users were notified when they trip because of bad PF*
If you stop and really think about how much energy is wasted every day, you'll probably feel sick in the stomach!
There is NO electricity wasted here. The capacitive dropper in these creates a LEADING power factor, which improves the inherently lagging power factor on the supply grid. Dave is wrong.
@@johncoops6897 No, he gives Watt and VA rating .. and even the Watt rating is way too high (because of the Zener diode solution).
Additionally the ohmic loss (think supply line) increases if the current increases, which it also does if only the reactive power increases.
I'm typing this reply while on night shift at a production facility with no other employees in for the weekend, yet heat and light is on full for almost a square kilometer of buildings. A 2W smoke detector is the least of my worries.
Hmmm... I'll pass this onto my company, we're building a new campus and we're interested in total CO2 emissions and TCO of all operations. We put in a bunch of stuff that's supposed to be "energy efficient" it might actually be a good idea for us to audit the hardware the site planners have selected.
You should check the facts first. The use of capacitive droppers like these will IMPROVE your site's power factor. That will take the load of the expensive on-site Power Factor Correction devices that you will be installing to add capacitance and reduce your kVA loading onto the grid. Capacitance is capacitance... add it with these or pay thousands to add it anyway. Don't listen to Dave, he is clueless about this stuff, despite a squeaky voice and massive fanboi base.
If the current THD is high there wont be the perfect inductive power factor compensation that is expected from capacitor load.
@@lucaskendc - Of course loads with lots of distortion and harmonics (THD) will not be perfectly inductive or capacitive. However, in order to be significant to the power grid, those "distortion components" have to be associated with a large load. These smoke detectors are a tiny tiny load, so the THD is literally "a drop in the ocean" compared to a (say) a 150W switchmode PSU, even if it has higher PF and lower THD..
Except the zener PS wastes in heat almost all the power the device draws from the grid... _(independent of PF)_
This is a case when low Iq is a good thing.
WAHAHAHAHAHAAAAAAA. Good one :)
That's why the humans won't get more intelligent, because the intelligent ones kill them self because the bad world
@Pete In other words, smart people are shoving themselves out of existence.
I think you're over-blowing the situations. You'd have a point if this was a grid of nothing but smoke detectors, but the truth is they share the grid with things like fans, A/C compressors, refrigerators and other devices that don't have a power factor of "1". In the case of residential power, the power company takes care of it by adding capacitors to the power line (since most of the problem is inductive loads, caps are the "anti-inductive" fix.). All that these smoke detectors do is shift the amount of correction by a small about. (Small because their current is small compared to motors.)
Yeah, but I'm trying to point out the potential seriousness of such poor design. See the pinned comment above for the discussion on this.
@@EEVblog Ok, but good engineering is the art of compromise. How much are you saving compared to the additional parts (and their impact on the environment)? I still think you picked the wrong poster child. I'd worry more about bigger electronics and modern LED lighting systems. Big Clive routinely talks about how bad the PF is on the bulbs he takes apart.
I wonder if the electricity companies are seeing a shift from the mostly-inductive load as it was in the past (requiring addition of capacitors) to a mostly-capacitive load today.
Or maybe it is still mostly inductive, but far less than it was in the past.
After all, many loads like mains transformers and small electric motors have been replaced by new designs that use high-frequency switching and present a capacitive load if not corrected well (in which case it will be resistive).
Great video. Addresses EXACTLY my concerns about power consumption of late.
So i'm building my own switching power supplies with standby circuits for stuff more and more lately, its not rocket science, you go to a semiconductor catalogue and find a chip that allows the features and there is a plethora of stuff in the 'ù' amp range for hibernate/standby. Build that chip into the design.
Its time people started lobbying the politicians: demand very strict energy efficiency. Revise it every 18 months as the stuff that keeps coming out is getting better and better. Another gripe: Don't want E-waste? Then mandate a 10 year manufacturers warranty for EVERYTHING.
That'll elliminate dodgy power supplies, cheap electrolytic caps etc in no time ;-) Solutions literally sitting on the shelf!
Manufacturers will conveniently go out of business right when their products start failing. And news ones will start up with the exact same products
There is no such thing as "generation" of reactive power. Apparent power just causes an increase in current in the transmission cables, transformers and generators. Of course the increased current causes losses due to resistances and ultimately some additional active power to be produced but nowhere near the total apparent power.
I can tell who has a degree in electronics!
That thing uses almost as much power as having 4-5 odroids or raspberry pis running.
You could keep your bitcoin wallet up-to date with that kind of power... And they say Bitcoin consumes too much power... (yeah, i know mining will always take more.)
In VA yes, but not in W
I've thought for some time that it would make sense to standardize around some DC voltage (12v?) to extra pins on the household mains sockets, and have a central AC-DC converter(s).
You'd have to deal with volt drops around the house, so you might need more than one AC-DC converter in the house to keep the cable runs short, but you wouldn't need one per socket. If you had specified a nominal 12v it would actually be between 8 and 16v, and a max of 500mA maybe. That would be enough for clocks, smoke alarms, and the standby circuits in lots of the consumer electronics (TVs, CD players, microwaves clocks, oven clocks... etc).
2 main advantages:
1) The central converters could be much more efficient, maybe even specified in building codes.
2) It would reduce the cost and complexity of all devices that can take advantage of it, especially those that only need low current DC - a cheap small switcher and you're done.
Unlikely to happen, but I think it's an idea with some merit...
the same here, most new lamps are low voltage LED connected to main power, millions are installed every year for example. If you have PV even better, avoiding losses converting to DC to AC
I would want something more like several amps max. So you can charge phones and stuff on it
the voltage drop would be a big problem since people would not like to invest in the correct wire gauge to lower the voltage drop and it would be another wire running around your house making the wiring more complex and creating more confusion for people who can barely wire the line and neutral wires the correct way around.
power saving? in straya? Why, when there's so much coal!
Real eye opener here. So much waste. I wonder how many other crappy mains devices are out there. Perhaps make a section about this dave? We should all inform our local representatives to take notice so products like these come with a “waste” warning, or better still, are banned from the market unless they do it right.
There are TONS of other devices!
The Energy Star legislation for example was bought in to reduce standby power consumption on plugpacks.
I remember Apple Powermac G5 computer having very very high ''off state'' current.
Here in Europe we have energy labels per product. But I haven’t seen those on these “low” consumption devices. They only use it for big devices such as refrigerators. And I’m not even sure if that label incorporates the power factor to measure the actual consumption.
Either way, I’ll try to ask for more info on this. Thanks a lot for reporting this! I always suspected that there was a lot of waste with these rectifiers, but never knew it was so massive.
Psssh, photoelectric .. pull apart ionization ones and become Hulk
it does get cheaper!, that aint 55 yanky cents, thats 55 centsarydoos, ya switched to the aussie digikey site
Instantaneous power is P = V*I. In an AC system with a resistive load (PF=1.0), the V*I is always positive, thus the load is always consuming power. As soon as the PF is not 1, then there are parts in the AC wave where P = V*I that is negative. Because of the phase difference between I and V. Thus, I will be positive while V is negative for some instances of time, and vice versa. That means the load is supplying power for those negative instances. You still have the real I^2*R losses, as those are PF 1 loads. The generator, and all the infrastructure, must be able to supply the V and the I. So yes, if I becomes larger than what the generator is able to supply, you would need another generator (or tranformer). The reactive power on the other hand appears as a torque ripple on the generator shaft. Not ideal either. The real power, as seen by fuel usage, is only P = V*I*PF, PF the power factor as seen by the generator. The I^2*R losses will change the PF at the generator from the PF at the load. So, bad PF wastes extra I capacity that could have been used to perform actual work.
Hmm... I was thinking that for non-linear electronics load, negative V*I does not happen. Load is not sinusoidal. So V*I is either positive or zero. That will create torque ripple and generally mess things up for sure, but no extra I^2*R losses in down time. I think PF is somewhat misleading in those cases as "apparent power" is kinda BS in non-sinusoidal situations.
@@otherbasis8505 Non-linear loads does not preclude negative V*I. It just means that given a single frequency sinusoidal voltage, the current will have different frequency components than just the supply frequency. Also means that superposition does not apply. I have seen some add a PF based on the current total harmonic distortion to the normal lead/lag PF. Though that part I'm not too familiar with.
I simply cant thank you enough for this video. It is an important one. Governments need to take action. I have a modem in my house. It costs me almost $ 200 a year. The designers should care and it needs to be a topic.
You should pay more for these alarms! They’re a combined fire alarm and space heater.
Yet another reason to implement a DC circuit in your house?
30min video, just uploaded, already has dislike... people should check their brains.
This one will get a combination of the serial haters plus the anti-global warming crowd.
This is youtube. Even a video of a puppy playing with a kitten will get a bunch of downvotes.
Hey DAVE check the Power factor again...its leading ( capacitive ).....and it is actually better solution because its compensate our home power factor which is generally lagging in nature due to inductive load and nonlinear load like bridge rectifier ...so our generating station will get little help from each of this tiny capacitor....
Which is not enough to compensate for high consumption by the Zener diode(s).
@@absalomdraconis do you think that those zener can dip the power factor to .069 with 1 watt active power consumption . ??? It's connected in series with the mains capacitor and its main function to protect the output capacitor ... Its can't draw reactive power ...
100%! I just wrote a comment explaining the same thing, but it was "magically deleted".
If @EEVblog wants to post a follow up, there is a very simple experiment he can do. Hook up the scope to show both mains voltage and current going into those smoke alarms (not easy - high common mode but small signal).
Then compare that to a small linear power supply, or a traditional linear plug pack, with no load on it. (ie just the input transformer).
In the tranny of the linear plug pack, the voltage will peak before the current (that's what inductors do), and in the smoke alarm the current will lead the voltage (you guessed it...capacitor...).
When both are supplied from the same mains plug, these 2 effects will (partially) cancel out, ie the overall power factor of the 2 devices in parallel will be closer to one (current closer to being in phase with voltage) than either one of them individually.
That's pF correction in a nutshell. And that is what the smoke alarm is providing. pF correction for the rest of your house.
Note pf = cosine(phase angle between voltage and current) .... but that angle can be positive (inductive loads) or negative (capacitive loads).
I lived in a house that had Live and Neutral swapped on a few outlets. It caused a significant amount of voltage on the ground contacts of the coaxial cable. Never did figure out which one was the problem.
I discovered that the live and neutral were swapped on a light switch in my house when the socket went POP and showered sparks in the kitchen when I was trying to remove a broken light bulb base with a pair of needle nose pliers. I was glad that I'd decided to put on gloves at the last minute before starting work.
Here in western Europe we have symmetrical mains outlets - no way to tell which pin is live and which is neutral. So they cannot be swapped either :-)
But what you describe there can occur nonetheless. It is usually not caused by "swapped live and neutral" but by a symmetrical mains filter in the equipment (small caps from either mains lead to chassis) causing the chassis to float to 1/2 the mains voltage when it is not grounded.
Indeed, this can cause nasty surprises e.g. when you grab a grounded antenna cable and try to connect it to a device touching both the cable shield and the device...
6:21 Almost became an episode of ElectroBoom!
The products are lacking a FULL BRIDGE RECTUMFRYER!
Dave (as others have done) has just driven off of the road into the ditch. Homes and industry present a largely inductive (positive reactance) load on the power suppliers. Items like this and LED lamps with capacitive droppers (negative reactance) partly counter the existing positive reactance, they do NOT create a bad reactance. This is review is a dud. Ron W4BIN
Power factor correction really only applies to the device in question, and PF can even vary how the device is used. You can't simply slap a negative reactance into your system and expect it to help. Besides, with more and more devices that have negative reactance on your house mains (LED bulbs being more common), you can no longer guarantee that most reactance presented to power suppliers are positive.
@@wyrmofvt True. Our house is mostly capacitive. The few motor loads we have are sporadic (fridge, boiler pump, clothes dryer). Most of the time it is electronic stuff with switching PSUs or capacitive droppers.
9:28 "That's 65,000 times the power consumption!!"
Hmmm, maybe a 16-bit overflow problem in the microcontroller? /s
Pepperidge farm remembers when all smoke detectors needed was just a simple 9v battery and it lasted.
And ran out at 3AM and started beeping all night
These smoke alarms should bring those "chinese power energy savers" in the box.
Wish they would make an ionizing detector that also produced it's own power :)
Should we mention all the unnecessary ITO devices, cheap TV's and video boxes, computers that stay idle instead of powering off (my laptop refuses to turn off completely when powered from main, it draws a continuous 12W)...
Edit: oh! And I forgot, Wi-Fi connected lightbulbs...
I'm glad I'm running mostly on off-grid solar. As long as the sun shines, I can waste energy.
5:41 - These are those 240-volt kind of watts too! None of that 120-volt Yankee watts rubbish :)
Morbo voice: WATTAGE DOES NOT WORK THAT WAY! GOOD NIGHT!
110V just a bit stingy, 240V is the baseball bat :P
I often wonder if LED bulbs suffer from similar issues with low power factor and high peak currents on the input, not to mention the environmental impacts of the built materials (plastic and components etc). It's like we've traded lighting power consumption in return for having more e-waste to deal with, I'm not sure how LED bulb failure rate is now but a few years ago it was a pretty high.
Bigclive often tests LED bulbs on their power consumption and power factor. You get what you pay for usually.
The problem with the smoke detector is that it has to use a capacitor dropper that is sized for the max current draw (when the alarm is sounding) When the alarm isn't sounding, it still draws that power and just dissipates it as heat. A similar power supply on an LED light bulb will be sized for the light bulb itself- there is no appreciable excess. Also, it's only drawing power when the light is in use. The smoke detector draws power constantly. A capacitor dropper is fine for something like an LED bulb that is either on or off. But in a device that only needs a tiny amount of current most of the time, and occasionally needs a lot more, there should be a switching power supply. But that's more expensive than a dropper, which always draws the same amount of power whether it's needed or not.
Could it be that they chose such a "simple" power supply design because it's simplicity makes it more robust and reliable which is an important consideration in what is basically equipment used in a life safety application?
Whilst the argument might have some validity, in the real corporate world of "products built down to minimum BOM cost for maximum profit margin" the decision was probably made because capacitive dropper supplies are as cheap as it can get. A passive transformer-based linear supply would use minimal extra components, would be just as reliable as a cap dropper (if not more so), and could easily be designed to have better power factor. But such a design choice would add 20-30 cents to the BOM cost per unit.
29 years ago i went to a training for reliability , the closing speaker ( from rocky mountain institute) was addressing the issue of designing for energy economy , telling that it our responsibility as engineer to design energy efficient product especially when they are mass produced . he was given the exemple of instant start crt based TV that use 6 watts standby to keep the tv ready to turn on rapidly , he was saying that represent in USA the power of 3 nuclear power station just to keep TV ready to turn on rapidly !
Wow! You think the move to LCD screens has just been offset by the sheer numbers all consuming say 0.1w standby power?
A mains powered smoke detector? lol whats next a mains powered pacemaker?
Good one! I think its just so that you cant ignore the warning chirps as the battery goes flat. The mains guarantee you will be nagged to replace the battery.
It's always pissed me off how inefficient the mains versions of these things are, for exactly the reason you give. Never realised it was 65000 times - just knew it was 'really bad'.
You are not the first person in Oz to notice: there is even a govt report about it: reductionrevolution.com.au/blogs/news-reviews/5842566-99-waste-the-unexpected-energy-consumption-of-smoke-alarms
I just tried and failed to find a mains one that was actually efficient.
I also just complained to Which that their review of 'best smoke alarms' didn't even mention the power consumption. Without some consumer pressure they are not going to spend the extra dollar to make a decent PSU.
You could fly to alphacentari on half a watt
Imagine how far 20VA gets you!
EEVblog i could start a galactic empire with that!
@@EEVblog I'll get to the moon
and 1.21 GW will get you back in time
Interesting rant on the VA of the device, it may be overlooking that this is most likely Capacitive, and 180 from the Inductive power factor of the transformer supplies in the microwave, home router, and the inductance of the fridge, and other appliances with a motor, such as heat pump, AC, and furnace? Don't overlook the inductive load that is the utility transformer on the street feeding the house. This may actually reduce the total VA connected load of the house. Individually this looks bad, but it may fix some of the large amount of connected inductive load on the utility. Have you put current transformers in your main panel to find the total power factor of your house and if it is leading or lagging?
You could also have a look at the alarm clocks sold. Probably the same amount of units per house hold and it could be battery powered or even wound up ones. But there are lots and lots of them on the market which draw more power costs in a year than you have to pay when buying them. And as they have a "clock" in them, they don't need to be sub 0.5 Watt stand-by as other devices must comply to. (ever bothered to think why lots of devices have clocks these days? :) )
Can you point me somewhere where I can read more about this regulation? It sounds interesting
@@sasodoma As far as I know, it is at least an EU regulation. See for example: ec.europa.eu/info/energy-climate-change-environment/standards-tools-and-labels/products-labelling-rules-and-requirements/energy-label-and-ecodesign/energy-efficient-products/mode-standby-and-networked-standby_en
I bet the springy metal switch contacts on the Quell unit cost a lot more in terms of storage space and automation. I bet it's a lot easier to get production line fixtures for your surface-mount button than to fit those metal spring contacts at just the right angle.
You can actually use a high voltage depletion mode MOSFET to modify any "normal" regulator (7805,LD118,...) to support serveral hundret volts on the input.
Not so good for improving efficiency though.
@@absalomdraconis That really depends. In standby its efficiency is near perfect, with iq of a few microamps, under load it is as unefficient as you'd expect from a linear regulator =)
@@berniwa Don't forget that all idle current of your linear regulator is REAL power, as opposed to the APPARENT power that a capacitive dropper has. In the shown circuit only the power dissipated by the zenerdiode is real power, all other power is apparent power. But drawing even only 1mA in a linear regulator that operates on line voltage will cause a big real power usage. That effectively precludes any use of conventional devices like a 7805.
@@Rob2 Sure, with a 7805 the current draw will be problematic, but technology progressed since then and TIs 7805 cousins (like the TPS783xx) have an iq of 500nA. So the real power consumed by the regulator would be a moderate 115uW at 230V ;-)
@@berniwa But once the alarm starts beeping and draws 50mA or so, your solution draws over 10W of power and dissipates that all in the regulator/fet, while the capacitive dropper solution still only draws about 1W as it always does.
Not important from energy consumption perspective, but the design has to accomodate the dissipation of those 10W at least occasionally.
I'd be more pissed by the amount of power it really uses. That 1.3 W thing would cost 3.5 € per year in Germany. That's insane.
years ago I have seen a TV antenna rectifier which was rated for 150 watts - 1300 kWh per year, or almost 400 € in Germany (it was installed in Germany).
Chinese products are notorious for this, but they could be fixed really easily.
Kidde is an American company...
No way they make their PCBs in America, just sayin'
Capacitive power supplies help correct the low PF of inductive motors. The big waste here is not the large VA but the use of linear voltage regulation from 240V to 12V.
6:02 Yes, those electric smoke alarm numbers are truly *shocking*
I had some poorly designed battery operated fire alarms before, they'd chew through a 9v alkaline battery in under 2 months, whereas the good fire alarms i replaced them with can last a few years on a 9v battery.
A good lesson to learn, never cheap out on safety equipment.
I have seen the struggle to understand reactive power among engineers working at power plants !
There is this option to set pf from SCADA system in the power plant I worked, we used to set it 0.95 and that will increase the excitation current of the generator, meaning generators need to work more hard to make up the reactive power, thus needing more torque to rotate to maintain the same rpm (frequency) and more fuel.
I am talking about generation sides , not load sides.
Yes, it's seriously perplexing. See the pinned comment above for discussion. Would appreciate actual power generation engineers weighing in.
Where do they generate the exitatiion current from, and this is applied to the electromagnets on the rotor right?
@@ryanb1874 some of the current generated is fed back into the rotor through a rotating transformer i think. When you need to start the thing up there should be enough magnetism left in the rotor to start generating enough current to get the whole thing going again.
Smaller generators use slip rings and can be shunt wound/series wound and all that stuff I don't remember from uni...
@@ryanb1874 Well, it is but not directly !
The generators in the power plant I used to work had actually 2 generators in 1 genset shaft
(made by ABB).
Excitation controller will supply few Amps of DC to the stator of excitation generator G2,
G2's rotor will make 3 phase AC which will be converted back to large DC current to feed
G1's rotor with a 3 phase rotating rectifier, and finally the actual AC power comes from the stator of G1. Something like this -
diary-of-electric.blogspot.com/2019/05/wiring-diagram-avr-of-3-phase-generators.html
p.s. I am sorry if I explained something incorrect, it was almost a decade ago. Now, I live in embedded electronics world and care much less for electrical systems.
@@demoncloud6147 sounds complex, I was just watching a Don Smith video and the old fart could even get his Over unity, resonator to work. THANKS for the comment.
Why do you need a mains powered smoke alarm anyway? So you have a product there, with a backup battery, that you still need to check periodically and eventually have to replace. Might as well use it to power the device?
Out of curiosity, is there a benefit to overall reliability or certifiability of the circuit with the less-efficient design? I am just curious about other factors - apart from cost - that might have informed the design decision.
So in America, we have devices that just have (I think) a 24volt-ish line as well as a signal line to sound off the other alarms, as well as a 9v backup battery, but my home was built less than a decade ago, and older homes usually have just a 9v battery and the alarm line.
Most new homes have mains unswitched smoke detectors now. (At least in new builds in the last 10years in Canada)
Power factor is difficult to understand, if the current is 90 degrees out of phase with the voltage the wattage is zero. The only power consumed is from the resistance in the wires, transformer windings and generator windings. You need to understand the difference between true power and apparent power
17:15 "Ya know, sniff of an oily rag stuff" 😅😅😅
A video on PFC would be nice. A lot of misunderstanding regarding this topic.
Hey @EEVblog, I just thought of another one. Bathroom extractor fans, mine draws a couple watts all the time since it's just a big dropper resistor and a zener diode to run the fan timer control circuit (I've seen them with capacitive droppers too). Every bathroom gets them now which is another necessary vampiric load in every home.
i guess we need to download more electricity
Seems like you are in a position to start your own manufactoring company. Just get your youtube community to do the design ;p
I love the NOTE at the bottom of the page for the more efficient chip warning about blowing the ass out of your CRO . . . lol
Been there, done that. You do it once and never forget. Though in my case the scope was fine afterwards, but the probes didn’t fare too well with a dead short on the shield conductor.
I am somewhat surprised to see how many electronic engineers somehow struggle to get their heads around the phase lag between voltage and current. The bottom line is that the overall lag in domestic installations is inductive (think motors fridges, etc. ). This is even more so for industrial installations.
Adding a small capacitative lag will (slightly) improve the power factor!
I think Dave is getting at the additional I^R losses and extra carbon footprint required at the generation end to cater for low PF appliances.
uK8cvPAq
Yes, I know that’s what Dave is saying, but here’s the thing... we don’t have a separate grid connection for each appliance, just one to the household. In most cases the overall household load is inductive, so biasing the load with a small capacitive load will actually *improve* the overall household power factor.
@@mikelopez9893 I get what you're saying too, but isn't it better to fix poor pf at the source rather replying on loads to balance each other out?
uK8cvPAq yes, I’d agree that we should fix the power factor. So, how will we do this?
Again, the dominant problem is inductive load. This is prevalent in industrial applications, not so much in domestic. The standard “solution” to fixing an overly inductive load is to put a capacitor across the supply. That way, there’s a reserve of electrons to meet the out of phase inductive load. Oh dear, that’s where we started!
Unfortunately, there’s a lot of BS on you tube on power factor. The only youtuber I’m aware of who calls it out is Big Clive. We need the EE community to take a stronger stand on this. Reactive power is not something else that needs to be generated. The only possible implication is *current* for line distribution companies, not power. If we have to have higher rms current to meet the power needs, we may need beefier wires. That’s a big if though!
Here’s my challenge to any EE. Do the calculations. It might surprise you.
Maybe the next generation could run on 24Vdc and “network” together with the doorbell and also be a WiFi mesh network.
The trouble with zener-regulated supplies like this is that the series reactance has to supply the peak current demand, without dropping the voltage too much, which as you observe is in the ballpark of 15mA when the alarm is sounding.But even given that, as designed, it is drawing 70mA, so they could have used almost 4 times the reactance (reduce the capacitor value by a factor of around 4 to, say, 0.27μF) and get the same voltage drop with only 17.5mA of quiescent current, and it still would have been enough to run the alarm when sounding.
I have to wonder if they are using exactly the same circuit for 120V systems. this would give a slightly less feeble justification for a whole 1μF dropper capacitor, though that would still be about twice as large as necessary and the circuit would still draw twice as much current as necessary
"If the actual product itself takes less than 50uA, why does the product have to take orders of magnitude more than that?"
So the real reason for this is because this kind of power supply draws constant power (all the time, regardless of the actual useful load current consumption), and you have to design for the maximum load current that you'll ever draw. If you have a 30mA load to power a 2V LED, then that also results in a 30mA input requirement at the AC mains voltage as well--which is a lot more power than you need. If you also want to chirp the alarm while keeping the light on, and that takes another 30mA, you now have a 60mA design requirement, so you will always draw power equal to AC mains voltage * 60mA. Add another 10 to 15mA for the microcontroller, etc. Then When you also factor in that that a lot of the resources for transformerless power supply design don't even do the calculations correctly, you end up with overdesigned transformerless power supplies, or worse, ones that are reused from previous designs without optimization because "they just worked"
Also, at 240V, your losses for a given load are 2x that of a 120V mains system, so that's unfortunate. 50Hz vs 60Hz doesn't matter much in this type of application because it only affects the size of the DC capacitor you put on the output.
I've put together a number of helpful design resources on this very specific topic (mostly because I couldn't find anything good myself a few years ago) at www.designercircuits.com/transformerless.htm, including a document that walks through the calculations and different topology options, spreadsheets and schematics for both capacitive (like the smoke detector) and resistive version of transformerless power supplies.
One of the conclusions I've come to is that this is really only a viable topology for loads in the 10s of milliamps, because beyond that, it is difficult to even reject the heat generated. But if you do need more current, a very clever approach that greatly improves the efficiency is to target an intermediate output voltage, like say 40VDC (meaning less energy is burned up in the input resistance), then use a switching power supply to efficiently step that down to lower voltages. This lets you get upwards of 100mA without much cost, because switch mode controllers in that voltage range are still pretty low-cost.
Beyond that, it really is a matter of going into the design with ultra-low power consumption components. And avoid LEDs.
A quick addendum to the above; at 0.5W of real power consumption, that maps to about 2mA of current that gets consumed/dissipated. Still like 40x what is needed for a 50uA average load. A far better solution economically for the manufacturer and consumer would be to eliminate the capacitor (short across it), replace the 100ohm resistor with 1Megaohm resistor, and still have 100% margin in your load current (support for 100uA). Beef up the output capacitor to handle peak transients (beeping, LEDs, etc.)
So it looks like the most clever approach here will be to use a resistive transformerless power supply to run the logic needing 50uA constantly, and whenever it is time to beep, since 15mA is needed for that, have a separate transformerless supply that you turn on with an SCR/thyristor controlling that second transformerless supply from the low side. You still save money because you can still avoid the capacitor for 15mA, and the diodes, resistor, and SCR is still cheaper than the 1uF AC mains-rated capacitor.
He didn't even factor in hotels that have 100's of these sucking away power. Here in the US you go to any hotel and they have a smoke detector in every room I'm assuming it's the same in other countries. The amount of wasted power is insane considering the smoke detectors here and everywhere else are most likely nearly identical designs to those down under.
As an electrical engineer in the USA, there are so many reasons why 50hz is rubbish compared to 60hz. I think it would be interesting if you did a video comparing the two, and seeing your reasoning on why 60hz is rubbish.
Could you give a few reasons why 50Hz is rubbish?
(To give an example of a disadvantage: 60Hz creates a bigger resistance and bigger losses in transmission lines).
I prefer higher refresh rates. Lol
Whoosh!
Maybe a discussion about "what is beter, 115V AC or 230/240V AC for consumer mains outlets"...?
Americans write that 115 is "safer" but I really question that. 115 is high enough of a voltage to get a dangerous shock, and with 115 the current of the average appliance will be double of what it is in the 230/240 system. The higher current contributes to fires due to bad connections, so 115 may be more dangerous as a whole.
In Japan we have both 50Hz and 60Hz power grids!
We were incorporating switchmode power factor correcting IC's in designs where I worked 35 years ago
Mobile phone chargers. Many people leave them plugged in without any load, I bet the quiescent current is awful on many of those.
They've spent a bit more money per unit and use switching supplies in those, though.
There is the energy star legislation for plugpacks.
Ive got one plugged in nearby but its cold to the touch so cant be worth fussing over.
The quiescent current in phone chargers is very low. They use proper buck converters to drop the voltage, so when there is no load on them, they are basically turned off. Power factor will be horrible because they are only loading the peak of the sine wave, but a low power factor on a ~zero load isn't a problem.
Good ones use circuitry that follows the sine curve to give them a good power factor under load.
TVs have outstanding standby current. My dad is a service tech for Philips. 20 years ago, the training tech was HYPED about the low power consumption on standby on those sets. It was so low, the standby LED would remain powered for a good half hour after you unplugged it. They used a big capacitor so in standby the TV would just "burst" the PSU to charge the capacitor and go back to sleep at milliwatts again. My old-ish LG LED TV has a relay to completely disconnect (not just standby) the main supplies. Only the standby PSU remains powered in standby. The rest of the TV is OFF. And then the PACE cable boxes just blank the HDMI output when "off". The tuner and decoder circuitry runs 24/7 and the device is a nice space heater...
Hey that device has a dollar more of components. That's gotta cost and extra $40 on the purchase price. 😅
It’s winter time where I live (middle of US). The waste is just heating my house this time of year, right?
More like it is heating the wires to your house :)
@@TD-er Don't give solar freakin roadways any ideas!
@@uK8cvPAq well I do live in the Netherlands, where we apparently have those, although I had to learn about it on some Australian vlog channel ;)
Yes...but a heat exchanger (your A/C unit) will do it more efficiently. But if you're just looking at financial cost, at even 1W, and estimate 10 cents per kWhr, then 1W * (1kW/1000W) * ($0.10/kWhr) * (24 hr/day) * (365.25 days/year) = $0.88 per year to operate a 1W smoke detector and incidentally heat your home.
@@davidhoelscher True. I have a heat pump and high efficiency gas furnace. I should probably work on shutting off electronics in the summer when I am cooling. In that case, the electronics are using electricity and I am then using more electricity to remove the heat from the house.
If it's mostly *harmonic* distortion as opposed to voltage-current-phase shift, can you power-factor-compensate that on a grid-level?
Gotta love how these environmental nuts whine about the craziest shit when things like THIS are being totally IGNORED as if it's all peaches and cream. Thanks Dave for highlighting this issue again!👍👍👍👍👍
Thank you Dave for showing us the problem and fix
What about the perils of mixing full wave and half wave rectifiers on the same incoming ac feed? If one smoke alarm manufacturer was to use a non-isolated half wave rectifier in their product and another chose to use to use a non-isolated full wave rectifier if their DC commons are tied together you will also be wanting to have on hand a battery operated smoke detector to detect the smoke the two hardwired ones will produce if they are connected to the same incoming AC feed.
I think that may be why we dont see many non-isolated half wave and full wave rectified consumer products. If they have isolation transformers before the rectifiers then no problem but snce in most countries our neutrals are grounded back at the distribution panel keeping the DC commons from being connected is difficult without isolation of some kind before the rectifiers.
With things as low power as these, it's not an issue. And you'll notice that in his reverse engineered davecad circuit, the zeners simply short out one half of the mains cycle(through the capacitor and resistor), so the draw them is still reasonably balanced. Yup, toss out half the input current, then toss out most of the other half. Uugh, I hate bad design.
Dave's reverse engineered Davecad circuit is of the manufacturers current inefficient capacitive dropper circuit.
I am not questioning the manufacturers use of capacitive droppers but rather Daves suggested use of half wave and full wave rectifiers instead unless they are also isolated from the mains. If the proposed rectifiers dont have isolation from the mains this could lead to mixing full and half wave rectifiers together on the same AC feed which will not end well. Adding the required isolation before the rectifiers would increase the cost of these smoke dtectors even more than Dave's proposed 250V linear regulators. In the case of a hard wired smoke detector you would also want to be connecting them together in some way to allow for one detector to set off the others in the same building. and this could very well lead to interconnecting their DC commons.
Mixing a non isolated half wave rectified power supply with a non isolated full wave rectified power supply on the same incoming AC supply with their DC commons tied to ground will result in their diodes shorting each other out as they find a path through ground. Proper incoming isolation and grounding would need to be observed or the magic smoke will be released from the rectifiers diodes.
With an R-C voltage divider input to drop the mains voltage prior to rectification, won't the power factor be LEADING? Therefore, given that the vast majority of reactive loads are inductive, these smoke detectors would each serve to provide a small corrective effect on the grid. Or am I missing something here?
The assumption that there's lots of inductive loads may no longer hold true because so many products use capacitive droppers.
FireX is also made by Kidde.
I'm surprised you didn't get suspicious when the circuits were nearly identical, down to the same control ic.
Hi Steve, Ever wondered how much power is used by all the installed 'smart meters'. I know this is not paid by the customer, at least not directly. I never succeeded in finding in the specs of smart meters how much power they consume. I remember that once you investigated a smart meter. If you still have that meter you might investigate how much this meters draw or how the power supply internally works. Keep up the good work.
Sorry, i meant Dave..
The next public service announcement should be about Vampire Loads. "The More You Know."
its a CONSUMER product... made to CONSUME..... gettit ?
I have seen something really "nice" in that regards:
some sheds for bikes in a housing complex. To make it "more green" they included motion-sensors for the lights. The sensor takes ~2W continuous, the LED-Lamps installed are 3.5W .................. why not just a simple door contact switch with a timer.
Oh yeah - nearly forgot to mention - there is not just one sensor - there is one on the outside and one on the inside. And the sheds are built in such a way that small animals like cats can easily leave them, and of course there are many cats that constantly trigger the sensors.
Reminds me of the neon nightlights with the sensor to turn them off during the day. The sensor is just a phototransistor in parallel to the neon bulb, so they consume more current when off than when on.
But hey, it _looks_ like they're saving energy...
@@strehlow Oh yeah - for bigger installations it was always nice to see the little extra box that had a heatsink to not burn up XD
Dave is wrong about the reactive volt amps on the power grid with the low power factor, power companies use capacitors on the primary power lines to correct the power factor back to the generators, so the amps on the wires from the reactive power only affects the wires and transformers from the source to those capacitors.
What do you need the 15V for when the main chips in smoke detectors are designed to work with
Why not simply run a wire in entire home for 12V 10A smps which powers all alarm, Cameras and led lights and also other smart peripherals.
RJ45 connector -- PoE
if my math isn't wrong the extra 50 c cost to manufacture these it would pay itself trough power savings in 1-2 years
Sounds about right.
So, those worried about the environment are wasting energy just running smoke alarms. Wow! Hypocrites.
Something like that. Always wanted Thunderfoot or Dave here to do the math to see how much CO2 was pumped into the air in the production and viewing of the videos to plant 20 million trees (one time, rather than the 20 million trees, twice a day, every day, it would take to make just the US carbon neutral. Oh, yeah, also have to be magic trees that grow to full size in a day.)
Spring lever clamps are way more satisfactory for retaining tension on the wires in the long term. Old installations often have loose connections where the Copper has relaxed and the screw clamps aren't holding them tightly any more.
Totally agree. ( I don't want to make an advertisement, but wago 221 series are fantastic. They have PCB versions too.)
@@joelstienlet1641 - It's one of those rare occasions where it's cheaper to make them that way, ie without screws, and it's a better product.
Question, is there a better version you can recommend?
Speaking of which, that NBN modem I received gets HOT. Do you think anyone anywhere calculated how many extra power stations we need in Oz once we install them in every single household?
And I'm pretty sure it has no fundamental reason why it needs to use any of that power, especially when idle.
I wonder why to choose a linear regulator? My understanding of linear regulator is, that they just "burn" the input voltage minus the output voltage with the output current into heat. Like the old 78xx variants. I always thought you would need a switching regulator for such high dropouts.
In case of 450V input, 15V output and 80mA input = output => this results in (450V-15V)*80mA ~ 35W - way more than the original "bad" design.
No, with the high voltage regular the load is only 50uA + the 10uA or so quiescent current of the regulator. The zener is a shunt device that needs higher current to regulate it's voltage.
Because he's the co-host of the world's biggest electronics engineering podcast, and for him only the active power factor matters, not the deformant power factor, which is mainly because of semiconductor rectifiers which introduce a whole lot of odd, in-phase harmonics into the grid, and is the reason why his magnificent 3-phase power analyzer shows that aberrant ACTIVE power factor.
@@EEVblog Where do you take the 50µA for the NCP758A?
The reason for such a power supply is obvious to me: What do you need to run the alarm: If you think microamps, you are fooling yourself. Also, who wants to design two supplies. Better design is certainly possible though. Put a switch and an additional capacitor somewhere and you could drop the quiescent power most of the time.
I always thought that a capacitive dropper could be very efficient if matched perfectly to the load. Perhaps they did the calculations based on the current draw when the buzzer is operating continuously so it's extremely miss-matched when operating in sensing mode (99.99999% of the time) and almost all of the power is being wasted away by the zener diode. If that's the case could they have the microcontroller operate a MOSFET to switch in a different value resistor or capacitor when in alarm mode? That way it would change to being extremely miss-matched and inefficient only 0.00001% of the time.