This is actually very smart, I was pondering an alternate route. When relay is off, use the potential, when it's turned on it also runs the line through a transformer letting it parasitically live off the power being consumed by the attached through a secondary coil
I got one of these switches from another company and was dumbfounded by how they work without neutral. This video explained it very very well, thank you!
well having watched the video , brilliant as always and so cathartic , im not the biggest fan of sonoffs but the mrs loves your accent so i get to watch truly geeky stuff without getting my earhole bent by her who must be obeyed
IMPORTANT UPDATES. 1) Many people still have trouble with the touch switch. The reason mine works is probably due to replacing R10 after I damaged it. I don't have a second unit to test with, so DrZzs and others are working on verifying this right now. I'll update the web page with info about this. 2) YES, IT WORKS WITH LEDs. That seems to be the #1 question about this video! Controlling LED lighting is actually the most common use-case, and it's what this is primarily designed to do. I've tested it with an LED downlight and it works just fine, but needs the anti-flicker unit installed so that the Sonoff can get enough power to run itself. I didn't bother showing this on the video because I thought it would be the most obvious use for it, so there'd be no need to show it. Sorry! I should have included it anyway.
Three switches successfully flashed and working correctly. Soldering to the end of R3 is not for the faint-hearted, and I had to reflow it on my first switch. I would not advise using flux as it tends to obscure one's view. Thank you for making a very clear video!
Sonof are very late to the party. The Chinese have been making 433 MHz light switches which don't need a neutral wire for years. www.amazon.co.uk/gp/aw/d/B075YQ5JQL/ref=sspa_mw_detail_2?ie=UTF8&psc=1&th=1
Sorry if this is a really stupid question. I don't own one of these devices but I'm interested in buying one and have limited knowledge of electronics. You mentioned that the GPIO 0 pin has to be soldered via the resistor to get to the bootloading mode, but you also mentioned that the GPIO 0 pin is controlled by the touch sensor. Can you just use the touch sensor (pressing the capacitive "button" with your finger) to get into bootloading mode?
Thank you very much. As an electrician I have installed a few different smart switches and was very confused by no-neutral. Thanks for clearing that up.
That procedure worked perfectly fine for me, thanks a bunch ! Note that taking the ground on that small capacitor ended in shorting VCC to ground for some reason, that was solved by reflowing it with hot air. The easiest way to get ground is probably on the connector on the back :)
Thanks for a great explanation Jonathan. I flashed it without the reset. It worked. I got the Wifi AP, loaded the UI, was able to control the light through the Tasmota UI, but I could not get the Touch button to control the light. I don't know what could be different. Still has be stumped.
thank god for this video. my electrician told me he couldnt bring a neutral down to the sonoff switch i got but i got new sonoff TX that requires no nuetral and thank god its not a gimmick and is actually legit. thanks man!
Epic effort, and well-timed too! I've just been considering the very real problem of what to do about the 'lighting conundrum' in my home renovations. I've built my own ESP8266 shield for power and RGB lighting control for the kitchen and other places so this rounds off the project nicely!
As someone who's got a 1920s house with no neutrals, I very much appreciate this video and in depth explanation!! Now if there was only a DIY version of this to power a Wemos...
Almost! The 3.3V regulator is on the logic board, the 12V regulator is on the mains board. So you get 12V and the relay pin, and you'd need to put your own 3.3V regulator on your board
brilliant video... you reminded me of stuff I learned decades ago with electronics and forgot about! Saw a different video and it was so aweful watching someone try to explain how the switch worked without having a clue! Interestingly, I built my own switched socket using a sonoff mini... needed to build it this way as I need the socket to be covered and waterproof! BUT the neon indicator didnt work and I realised why because of what you reminded me of in the explination of the voltage potential. thanks
It's worth noting that he's using a battery powered oscilloscope for the analysis. To do this with a mains powered oscilloscope differential probes should be used for safety. (I've learned this by watching UA-cam so please correct me if I'm wrong. I don't own either differential probes or a battery powered scope.)
A differential probe would be ideal. You can use an isolation transformer on the circuit under test but you really need to know what your doing... Also make sure the front end of your scope is happy to sink mains voltage!!!
That's exactly why I used the isolated scope, instead of the much nicer main-powered scope sitting on the bench in front of me! I forgot to mention it in the video. Also it's easier to get it in shot on the camera when showing the stuff connected on the bench
@@SuperHouseTV I knew it but it felt so unusual for you to not mention the fact as you tend to always be very thorough on your videos, especially when it comes to safety. So I fixed it for you.
This video is GOLD, very good explaination if the L wire only trick, super nice tipp for cool terms baudrates and also to the last issue: I have experienced that many times and also came up with a very similar way of entering bootloader mode on the touch switches! I just always thought I was alone with the issue 😅 Thanks for the great video
I thought it was going to use the voltage drop across a small value resistor in series with the switch. But their idea is clever, it works well regardless of the load size.
Very well explained, thank you. Just one niggle in my mind though. My whole house uses LED (Mains) bulbs and one dimmer for LED bulb, so how would this device perform, bearing in mind the triac may interfere?
Great walk through. 70W power wastage (if I understand correctly) is a little too much for me though especially as the average light only uses 10-20W these days. Glad of the explanation though.
@@CyprienLaplace I'm not sure. The cap is in parallel with the lamp so I guess in not 70 W but it must be fairly high to provide a load larger than CFL or led bulbs. I would have liked to see so power measurements taken with a LED bulb as that would be the most common arrangement. Not sure heat is too much of a issue with this arrangement the caps store energy in the electric field. The resister is only for discharging the caps when power is off.
@@GlennPierce when the switch is off, I suppose that only the necessary current to power the circuit goes through, and that is certainly smaller than the 300mA. While on, well, it will depend on the load. I hope it will be far from the max, but still being waste. I wonder if it could pick the current in serie when the switch is on: it would only require a minimal voltage drop...
@Cyprien Laplace yes, that's correct. The capacitors act as a current limiter: they will allow up to 300mA through, but it will be less than that if the Sonoff doesn't draw much power. I didn't show it in this video (which was already epic) but I put a Current Ranger in series with the 3.3V rail and found the logic board pulls 70 to 80mA at 3.3V. There's also power used on the 12V rail for the relay etc but it's still far less than the 300mA/240V would indicate
@@SuperHouseTV yes indeed great epic video! If you have the opportunity to measure the current through the anti flickering module while powering the fluo compact bulb, i'll be interested by the result :)
just like magic, now i only need to wait for itead to put out a sonoff mini no-neutral. So i can "smart" all my house existing lights while using the old switch panels. I don't know if anyone like me, hate the feeling of touch switch panels. It just don't fit my house's style
Brilliant video! I was scratching my head for days wondering how they did it. One question, why use the mechanical relay when the triac alone can cut the power to the light? And secondly, if the anti flicker device can flow 300mA at 240V, that's 72 Watts, which is more than twice the consumption of an LED lamp. Does it flow this much in practice? How does it not overheat?
Me too, I am curious why they have a mechanical relay when a TRIAC alone can do the job. They do this too for the ZBMINI-L. I wonder if it is for troubleshooting purposes - making it easier to rule out the switch if the load (LED bulb, etc.) is dead.
I think I know why now: TRIACs are not ON/OFF, they are ON/HIGH RESISTANCE, so when a TRIAC is "off" there is still a small current flow, which at mains voltage is dangerous and can deliver a shock. Therefore a relay is also put in series to cut off that flow.
Yes, this is one of the downsides of smart switches! That's how much power the anti-flicker device will allow to pass, but luckily the switch won't pull that much power. I didn't show it in the video, but I put a Current Ranger in the 3V rail and measured that it draws about 70 to 80mA most of the time. That's just the MCU though, it's not counting power used in the power supply part of the circuit, the relay, etc
@@SuperHouseTV I see a lot of people come to the conclusion that the 300mA at 220V WASTES about 60Watt (60VA would be more precise) but that is not what happens. The anti flicker device does not significantly increase the power use. The regulator chip on the mains board will ALWAYS ONLY allow as much current to flow as it needs to maintain the 12V and/or 3.3V regulated. (It is on a feedback loop) It doesn't matter if it pulls that current through a resistive load (the bulb) or through a parallel inductor. There is some power loss over the lamp and/or capacitors, but the majority of the power is probably used by the ESP and the LED.
I've been pretty successful with just using bluetack to attach wires for flashing. Not sure if it'd work at this scale though. Maybe a drop of superglue in case one wants to avoid soldering. Another tip I have is to use wire wrapping wire for this type of wiring. It's much less intrusive as its extremely thin and requires less heat to get the solder joint done.
I bought the T1s when they first came out with the idea of re-wiring to fit them. I never got around to that - instead I used a battery powered servo to flick the switch. Now it seems I can swap those out for these! I guess I'll find out if they work for the 2 way switch arrangement (hall lights upstairs & down controlled from a downstairs switch & a an upstairs switch!). £12 delivered seems about the going rate!
Great Video Jon. Always look forward to seeing what you come up with next, you always cover things in good an clear detail. I did figure the anti flicker was a given for leds/ cfls etc. Also Nice UAV Futures intro ;)
The wasted power with the anti-flicker unit is only when the load is active, with the load inactive the power used should be basically the same as with an incandescent bulb (e.g. dependent primarily on the Sonoff's power draw). Given that the loads that would need it are already much lower power than the incandescent - it seems a reasonable trade-off.
Hi Jonathan, Dont suppose you have figured out using two of these switches with one light like in an upstairs downstairs 2way switching arrangement. Great vids and many thanks for your help
@ 27:20. If you are using a regular incandescent (without the series capacitors) you are ALSO using extra power. You are using this extra power to power the ESP. My guess is that it is not much different, because the smart power control chip in the smart switch will pull as much power as it needs to keep the 3.3V and/or 12V regulated.
Hi, great video.. I have sonoff touch no-neutral 2ch. On one channel i have connected LED light, on second FAN.. but once i switch fan on, switch loose power.. I will try replace LED bulb with normal bulb, and try it again with and without "load capacitor"
Hello, thank you for the great video full of helpfull information! For me only one question left: how much power it cunsumes via antiflicker device? As I understood, 0.3A is a maximum, but what about real consumption?
Excellent Thanks or that Jonathan - now, I'll still waiting for my Itead sample - the original Itead T1 UK version definitely does not work without neutral - and now you;ve sparked me off, I have a oair of ZemiSmart equivalents - one works perfectly - the other comes on and stays on, with or without the supplied capacitor (without it runs at half-power). I tested 6w LED E14 lamps - in the UK of course, on 249v mains power.
I tested my one - it uses 300mA when on (without a bulb) - but only 1.6W of power. It uses only 1 to 3 mA when off (with the wifi indicator off) with a power consumption of 0.2W. The PF for the switch circuit with the anti flicker device must be around 0.023 when on.
so, you can have a notification when a bulb is burned, setting the LWT on tasmota mqtt: if a bulb fails, device will loose connectivity, and that's it :D
Nice video ! Also played with this switch like a month ago and I like it a bit more than Minitiger switches U may also check those the power logic on them is different, but they are much easier to flash and also have the RF433 chip onboard so u may use that one too (useless for me though) But the big plus for this one is that I ripped off the front plate and cover and put instead an LCD with touch the power board here can produce enough mA to drive that, so all in all I like it but it only comes in 1 way gang and Minitigers can also be found in 2 or 3 way that's a bit sad but fine for now as I have the best of both worlds. DO recommend this 1 way switch if u need some more mA to power up some other logic and u have only one bulb to control :)
The trick of using the capacitor as anti flicker is nice, but in long term it is pretty expensive... A simple calculation shows that it consumes another 70 Watt when the load is on: 0.3[A]*240[V]=72[W], and this will usually be added to a bulb that consumes only 12-15W (standard LED / Florescent bulb) - so it increase the electricity usage by ~5 times (!!!) Please correct me if you think that I am wrong :)
Thanks for great vide @superhouse. Can I ask you if you would do a video about how to make a low-pass filter for Sonoff Touch ? As some of mines are ghostswitching (and it's not because of mqtt) I would like to give it a go, but not really sure how to do it on that board . If you could do that, that would be absolutely terrific! Thanks a lot for your time!
@@peterfitzpatrick7032 probably, since each switch consumes near 60VA, this can easily add up to huge numbers and cause over-voltage specially when the network load is low
Hey Jon, I know we've been saying 240V for ever, but I believe the standard for Australia/NZ changed to 230V in 2000. (Except Qld which is transitioning and WA)
Yes, technically the Australian standard says 230V, but Victoria still uses 240V as the nominal voltage I think. I should get in the habit of saying 230. When I do that, I get people saying "but it's 240V!" and vice versa
When the chip is pre-flashed, it does make sense to not have easy access to gpio 0, but then I wonder why gnd/tx/rx/3v3 headers are still there, as these should not be needed in normal operation. If the reason is to be at least somewhat tinker-friendly, adding gpio 0 and reset would be much preferable.
Good point.. I suspect the chip also needs to be flashed in factory. They could do that before they mount the chip on the board, but i think its common for manufacturers to give themselves the option to re-flash just in case.
@@dozog - Well, even the device manufacturer would find it hard to re-flash without access to the gpio 0 pin. And when cheap devices are in mass manufacturing, reflashing is highly unusual, especially as cloud-connected devices (like the Sonoffs with eWelink) are expected to receive OTA software updates at the consumer location. As long as there is room on the PCB, providing soldering points or holes is not much of a cost, and therefore I'm surprised if penny-pinching causes removal of gpio 0 but not gnd/tx/rx/3v3. Of course, the latter 4 pins may be useful for debugging (like connecting a serial terminal), but debugging is not much of a real-world use case with these cheap devices. While reflashing thru soldered wires is still feasible (I've done this in a smartplug), it does become a lot harder than when accessing the needed pins is easy. Therefore I wonder what the use case for providing gnd/tx/rx/3v3 without gpio 0 is. Speculatively, I suppose that it might be for attaching some sort of test rig on the assembly line. Yes, flashing before mounting was exactly the scenario talked about in the video here.
@@JohnnieHougaardNielsen I would guess it is for the test bench. with 3.3v/Tx/Rx/Gnd, you can put it on a nail bed and verify the function, operator can touch the switch to verify it works well. I am not sure the ewelink firmware post something through serial when working though but that may be an option.
Given that GPIO0 is used on the capacitive output of the touch switch, could you not just press the capacitive touch switch whilst performing the reset?
This is a great question, and I should have answered that in the video! The answer is no, because the timing doesn't work out. The ESP8285 has to see GPIO0 low at the moment it boots, and the touch sensor IC takes a little while to start up. It needs to calibrate its own input, and then it takes a while to detect the finger. So by the time it starts asserting GPIO0 it's way too late, and the MCU has already booted up
So, couldn't you delay the reset long enough to let the capacitive sensor to boot or alternatively a switch to CPU VCC or reset while powering the capacitive sensor, to slow CPU rebooting when you want to program it? Guessing VCC and maybe reset are easier to solder to than GPIO00. Ah, is the capacitive thing an I2C device? Then delayed boot won't work.
Hi, I noticed that after flashing the tasmota image the wifi LED works only when the circuit is closed (the bulb is ON)(see @40:00). Why is that? and how to fix it, I mean the led helps as a night indicator.
Ok please forgive my limited electronics knowledge but I have a question: why wouldn’t they have just placed a transformer in line with the active. When the switch is closed the transformer coil would induce a magnetic field across the secondary that they could use to run the circuitry? Is this not an easier option than pulsing the output? Cheers MB, 16 Perth
That would create a voltage division which would result in a dimming of the light. Same as putting two resistors in series. That's why they have to turn on/off the lamp very fast and briefly, or else the sonoff would turn off it self everytime it turns on the lamp which would also result in the lamp turning off immediately.
Correct, but since it's a capacitor not a resistor, it's reactive power (also known as imaginary power), not real power. You'll have to read up on power factor to understand it, but to explain it briefly, AC voltage is shaped like a sine wave. During each sinusoidal cycle, at the times when the AC voltage is higher than capacitor voltage the capacitor charges, and when AC voltage is less than capacitor voltage the capacitor discharges into the grid. So for a purely capacitive load, no actual power is consumed by the capacitor because theoretically it feeds back exactly as much power as it takes in. Of course, no real-world capacitor is purely capacitive, there's always some parasitic resistance that consumes power within the capacitor itself and makes it heat up. The advantages of this method is that the capacitor acts as a resistance to the AC current (whose "AC resistance" value is known as its impedance, and is calculated using the formula on the paper, 1/2*pi*f*c) without consuming as much power as a resistor of the same value, and some current is still flowing back and forth to charge and discharge the capacitor (so the sonoff can use it to power itself). Interestingly, utility companies will charge large consumers (like factories) in kVA (kilo volt-amperes) - which includes both real and reactive power - because the wiring and distribution system has to be built large enough to supply even the reactive power, which they have a lot of, due to inductive/capacitive loads (like big motors). For small consumers like you and me, they charge us in kW (kilo watts) - which is the REAL component of the power only.
@@krzysztofkujawski6513 For a purely capacitive capacitor, zero. For a real-world capacitor, I believe you can get a very rough approximation by calculating RMS current squared times Equivalent Series Resistance (ESR - you get that value from the capacitor's data sheet) - basically I2R. If you want a more accurate answer, start by reading this (since capacitor characteristics change with frequency) - passive-components.eu/capacitors-losses-esrimpdfq/
Perfect explain !!! Very good job !!! Well done !!! Does anyone know where can we found the ic lp3669 or what else can I use ? I have already 2. Switch with burn that ic.
---- 24:35 ----- Shouldn't that V be replaced with the RMS voltage there? Not the peak to peak voltage? Because if that 0.3A figure is correct, wouldn't that mean that there is a constant consumption of (240 x 0.3 =) 66Watts per each such device you install??? That's a very considerable consumption to add to your house. It negates any LED replacement bulbs you might have!
Has anybody who has installed a "no-neutral" switch noticed any noise from the capacitor when the light is switched on? I have both the Sonoff switch as in this video and some Lonsonho-branded Tuya switches and both of the capacitor units (both entirely different) "buzz" when the lights are on - tried with both CFL candle bulbs and low-power LED G9 bulbs to the same effect. The capacitors are silent when the light fitting is off.
I was thinking about getting into automation of lights and instead of automating bulbs and having physical switches turned on all the time smart switches make more sense and lights can still be turned on or off manually. But I would like to keep the switches physical (as in rockers, not touch points) because it works even with gloves, when hitting it with something else than hand etc. Aqara makes switches like that but they blink the buls every couple of seconds because they don't have the antiflicker capacitors (and I have no idea how to calculate those). So I was thinking if Sonoff switches could be hacked with a physical button instead of a touch button. From this video it seems the only necessary part is tripping the GPIO0 to turn on or off the relay (I don't know if momentary or latch is needed but I can find that out). But you mentioned soldering to GPIO0 is difficult so now I'm kinda undecided. By any means, do you think it could work? P.S.: I would need those switches without neutral, my wiring is old and I have concrete walls in an apartment so pulling more wires through is not an option unless I want to mess up the whole apartment.
Great video just getting into smart wiring etc after being an electrician for 30 years, we never used to run a neutral to the switch as this was bad practice(to stop your diyer switching the neutral instead of the line conductor) however in the past 5-10 years wiring a neutral to the switch is the norm,it's also alot easier when it comes to testing the installation .,could these sonoff be wired in series to a 230v coil controlling a contactor for larger loads?
Any reason you didn't scrape the via that's on the other side of your resistor instead of the middle of the track? Would give you a slightly better pad for soldering the wire to .
Great in depth stuff Jonathan, nice to see we have a bleed resister over those caps, I have in the last few days flashed my first sonoff basic with the latest Tasmotta firmware, however no matter what I do I can not get it to work with Alexa, someone has said that it no longer works, do you know anything about the matter ? thanks Stuart
If there is a triac, why the need of relay?! If the triac is not continuously on ( while the lamp is on) but instead switches on/off as you have shown with the scope, that means the lamp will not light at the rated power but less. To fully understand what is going on, there have to be an oscilloscope measurement at the lamp terminals ( not only the gate of the triac) .
I am wondering how the triac power stealing approach works for dumb LED lamps? They of course have an internal power supply of their own that tend to mess with old style triac dimmers.. different situation, I know. I was watching intently thinking a LED bulb would get plugged and tested after the CFL... but left hanging
Good video .. Please tell me what is R12 and C14 in board? How many ohms? Why not connect directly pad touch to pin IC ? ( just like ttp223 Schematic!!)
Regarding that antiflicker add on: So in a AC circuit the caps act as a resistor so the sonoff can get some power. But why not use just a plain old resistor instead of the caps? (I'm assuming there is a reason for that!)
AC and DC circuits behave in very different ways, so the resistor analogy may have been misleading. In this configuration the caps act as a current limiter, but they won't get hot like a resistor would. Resistors emit extra power as heat, while the caps accept power while charging and then simply stop accepting power when their plate voltage rises to match the supply voltage. They can sit in that state while using zero power (and emitting zero power) until the source voltage starts to fall and they begin discharging. Short answer: this is more efficient.
i have been thinking... why does ITEAD need the relay at all ?? The triac is used to interrupt the LINE in order to steal power. Does that not imply the triac can be used to switch the load off entirely ? In fact, since it is synchronized to the 50 Hz zero crossing it could be used to phase delay the leading edge of the voltage waveform, thus it could also function as a dimmer (at least for incandescent and some dimmable LED's). What am I missing ?
My question is regarding the capacitor and resistor. Those capacitors are connected to the bulb, wouldn't the discharge over the bulb and actually, there is no need for a resistor?
hi thank you for the video. i have a question about triggering the triac. Since its triggering the triac after mains voltage is above some treshold. How the circuit knows when to trigger. Is there a zero cross detector.
Great explanation. I love your videos! One question: Why did you call it "active in" and "active out"? Shouldn't it be "*live*" -- that's what the "L" stands for on the label: L = Live; N = Neutral! Is there a different terminology where you live?
Can I clarify how the anti flicker device is connected to the load? Is it as simple as connecting one end to the live and the other the neutral? Also does the direction matter? Thank you.
Sounds like you have worked out my old problem with Version one of the units. Now the mini will it work the same? I did the same with the button a couple of years ago. I ended up going to Espurna instead of Tasmota.
How much power can the high voltage side supply? Could I power a Pi Zero with a small touch screen for example? What I want to do is build a small unit that would replace a standard light switch and hack something like the Sonoff to power the Pi and the screen for a Home Assistant dashboard.
BTW The easiest way to put this thing into bootloader mode just connect it to your USB to serial power it up and tap the button while holding it just take with the other hand a pair of small tweezers or a small screw driver and touch the small resistor R3 the one u've soldered the reset wire to. NO SOLDERING NOR pcb torturing is required! I'm crying when I see what you've done to that PCB, why??? if u need a video with the steps I could upload one, but it should be easy enough just try it ;) works like a charm every time and I can not understand what was so hard that all others had troubles putting this device into bootloader mode.
I got a Sonoff SA-025-2.4G-1C with a single capacitor without the resistor. Does the stored capacitance on the resister not discharge itself across the load of the lamp over time? Thank you.
Very good explanation ! Is it so that the triac is in series with the relay contact ? What is the purpose of the relay if the switch can control output power ? I missed something probably...
Nice video! About five years ago, I tried to build a non-neutral circuit like that, and I did. But it did not work for me because this triac technique only gave me 100ma maximum and my circuit with wifi required more than this. So I would like to know how this circuit can delivery more current. Is there any other video or information source that explains this magic? Is it possible to record a new video about this theme? Thanks in advance, the video is really very useful.
Do you think two of these in series would work? L-out from the first into the L-in of the second and then to the light? Is there enough power left through the 50hz triac to power a second t4eu1c?
This video perfectly shows the difference between a UA-camr and an engineer on UA-cam.
I thought I’d just watch a few mins, next thing I know it’s over. Well put together and interesting, thanks.
Same!
This is actually very smart, I was pondering an alternate route.
When relay is off, use the potential, when it's turned on it also runs the line through a transformer letting it parasitically live off the power being consumed by the attached through a secondary coil
I got one of these switches from another company and was dumbfounded by how they work without neutral. This video explained it very very well, thank you!
oh yes a 40 minute video from my fave youtuber , cant wait to finish work tonight to settle down to watch this
well having watched the video , brilliant as always and so cathartic , im not the biggest fan of sonoffs but the mrs loves your accent so i get to watch truly geeky stuff without getting my earhole bent by her who must be obeyed
40 min, darn. Didn't notice at all
Was glued to the screen all the time
IMPORTANT UPDATES. 1) Many people still have trouble with the touch switch. The reason mine works is probably due to replacing R10 after I damaged it. I don't have a second unit to test with, so DrZzs and others are working on verifying this right now. I'll update the web page with info about this. 2) YES, IT WORKS WITH LEDs. That seems to be the #1 question about this video! Controlling LED lighting is actually the most common use-case, and it's what this is primarily designed to do. I've tested it with an LED downlight and it works just fine, but needs the anti-flicker unit installed so that the Sonoff can get enough power to run itself. I didn't bother showing this on the video because I thought it would be the most obvious use for it, so there'd be no need to show it. Sorry! I should have included it anyway.
Three switches successfully flashed and working correctly. Soldering to the end of R3 is not for the faint-hearted, and I had to reflow it on my first switch. I would not advise using flux as it tends to obscure one's view. Thank you for making a very clear video!
Sonof are very late to the party. The Chinese have been making 433 MHz light switches which don't need a neutral wire for years.
www.amazon.co.uk/gp/aw/d/B075YQ5JQL/ref=sspa_mw_detail_2?ie=UTF8&psc=1&th=1
@@trollobite1629 apart from this works with Google home and Amazon
Sorry if this is a really stupid question. I don't own one of these devices but I'm interested in buying one and have limited knowledge of electronics. You mentioned that the GPIO 0 pin has to be soldered via the resistor to get to the bootloading mode, but you also mentioned that the GPIO 0 pin is controlled by the touch sensor. Can you just use the touch sensor (pressing the capacitive "button" with your finger) to get into bootloading mode?
Thank you very much. As an electrician I have installed a few different smart switches and was very confused by no-neutral. Thanks for clearing that up.
That procedure worked perfectly fine for me, thanks a bunch ! Note that taking the ground on that small capacitor ended in shorting VCC to ground for some reason, that was solved by reflowing it with hot air. The easiest way to get ground is probably on the connector on the back :)
Thanks for a great explanation Jonathan.
I flashed it without the reset. It worked. I got the Wifi AP, loaded the UI, was able to control the light through the Tasmota UI, but I could not get the Touch button to control the light. I don't know what could be different.
Still has be stumped.
Any luck with your setup DrZzs?
Thanks, I've been wondering about this new sonoff for a week now.
Great vid.
I particularly like the nod to Stu @ UAV Futures!
I was trying to find the time myself to figure out inner workings of the T4EU1C, so thank you very much for the brilliant explination!
thank god for this video. my electrician told me he couldnt bring a neutral down to the sonoff switch i got but i got new sonoff TX that requires no nuetral and thank god its not a gimmick and is actually legit. thanks man!
Epic effort, and well-timed too! I've just been considering the very real problem of what to do about the 'lighting conundrum' in my home renovations. I've built my own ESP8266 shield for power and RGB lighting control for the kitchen and other places so this rounds off the project nicely!
You explain things really well. Thanks for that. It helps me make decisions on what to use when and where. 👍
Can't believe I have only just got round to watching this one! Great write up on how the magic works!!
As someone who's got a 1920s house with no neutrals, I very much appreciate this video and in depth explanation!!
Now if there was only a DIY version of this to power a Wemos...
you could just buy this and only use the power side pcb. The 8 pin header gives you 3.3V, ground and a relay pin.
Almost! The 3.3V regulator is on the logic board, the 12V regulator is on the mains board. So you get 12V and the relay pin, and you'd need to put your own 3.3V regulator on your board
brilliant video... you reminded me of stuff I learned decades ago with electronics and forgot about!
Saw a different video and it was so aweful watching someone try to explain how the switch worked without having a clue!
Interestingly, I built my own switched socket using a sonoff mini... needed to build it this way as I need the socket to be covered and waterproof! BUT the neon indicator didnt work and I realised why because of what you reminded me of in the explination of the voltage potential.
thanks
It's worth noting that he's using a battery powered oscilloscope for the analysis. To do this with a mains powered oscilloscope differential probes should be used for safety. (I've learned this by watching UA-cam so please correct me if I'm wrong. I don't own either differential probes or a battery powered scope.)
A differential probe would be ideal. You can use an isolation transformer on the circuit under test but you really need to know what your doing... Also make sure the front end of your scope is happy to sink mains voltage!!!
That's exactly why I used the isolated scope, instead of the much nicer main-powered scope sitting on the bench in front of me! I forgot to mention it in the video. Also it's easier to get it in shot on the camera when showing the stuff connected on the bench
@@SuperHouseTV I knew it but it felt so unusual for you to not mention the fact as you tend to always be very thorough on your videos, especially when it comes to safety. So I fixed it for you.
This video is GOLD, very good explaination if the L wire only trick, super nice tipp for cool terms baudrates and also to the last issue: I have experienced that many times and also came up with a very similar way of entering bootloader mode on the touch switches! I just always thought I was alone with the issue 😅 Thanks for the great video
I thought it was going to use the voltage drop across a small value resistor in series with the switch. But their idea is clever, it works well regardless of the load size.
Great work. I'm so glad I don't have any switches that need this beast. Flashing is a real chore.
Very well explained, thank you.
Just one niggle in my mind though. My whole house uses LED (Mains) bulbs and one dimmer for LED bulb, so how would this device perform, bearing in mind the triac may interfere?
Great walk through. 70W power wastage (if I understand correctly) is a little too much for me though especially as the average light only uses 10-20W these days. Glad of the explanation though.
Yeah, it's hard to believe how much waste it is. What about overheating?
@@CyprienLaplace I'm not sure. The cap is in parallel with the lamp so I guess in not 70 W but it must be fairly high to provide a load larger than CFL or led bulbs. I would have liked to see so power measurements taken with a LED bulb as that would be the most common arrangement. Not sure heat is too much of a issue with this arrangement the caps store energy in the electric field. The resister is only for discharging the caps when power is off.
@@GlennPierce when the switch is off, I suppose that only the necessary current to power the circuit goes through, and that is certainly smaller than the 300mA. While on, well, it will depend on the load. I hope it will be far from the max, but still being waste.
I wonder if it could pick the current in serie when the switch is on: it would only require a minimal voltage drop...
@Cyprien Laplace yes, that's correct. The capacitors act as a current limiter: they will allow up to 300mA through, but it will be less than that if the Sonoff doesn't draw much power. I didn't show it in this video (which was already epic) but I put a Current Ranger in series with the 3.3V rail and found the logic board pulls 70 to 80mA at 3.3V. There's also power used on the 12V rail for the relay etc but it's still far less than the 300mA/240V would indicate
@@SuperHouseTV yes indeed great epic video! If you have the opportunity to measure the current through the anti flickering module while powering the fluo compact bulb, i'll be interested by the result :)
just like magic, now i only need to wait for itead to put out a sonoff mini no-neutral. So i can "smart" all my house existing lights while using the old switch panels. I don't know if anyone like me, hate the feeling of touch switch panels. It just don't fit my house's style
Brilliant video! I was scratching my head for days wondering how they did it.
One question, why use the mechanical relay when the triac alone can cut the power to the light?
And secondly, if the anti flicker device can flow 300mA at 240V, that's 72 Watts, which is more than twice the consumption of an LED lamp. Does it flow this much in practice? How does it not overheat?
Me too, I am curious why they have a mechanical relay when a TRIAC alone can do the job. They do this too for the ZBMINI-L. I wonder if it is for troubleshooting purposes - making it easier to rule out the switch if the load (LED bulb, etc.) is dead.
I think I know why now: TRIACs are not ON/OFF, they are ON/HIGH RESISTANCE, so when a TRIAC is "off" there is still a small current flow, which at mains voltage is dangerous and can deliver a shock. Therefore a relay is also put in series to cut off that flow.
0.3amps at 240v is 1.7kw per day, 670kw per year, at about £120. That could be per light. Ouch. Great video on how they work.
Yes, this is one of the downsides of smart switches! That's how much power the anti-flicker device will allow to pass, but luckily the switch won't pull that much power. I didn't show it in the video, but I put a Current Ranger in the 3V rail and measured that it draws about 70 to 80mA most of the time. That's just the MCU though, it's not counting power used in the power supply part of the circuit, the relay, etc
It's not power unity so you cannot just calculate power like that. If it used 1.7kW it would be on fire!
@@SuperHouseTV I see a lot of people come to the conclusion that the 300mA at 220V WASTES about 60Watt (60VA would be more precise) but that is not what happens.
The anti flicker device does not significantly increase the power use.
The regulator chip on the mains board will ALWAYS ONLY allow as much current to flow as it needs to maintain the 12V and/or 3.3V regulated. (It is on a feedback loop)
It doesn't matter if it pulls that current through a resistive load (the bulb) or through a parallel inductor.
There is some power loss over the lamp and/or capacitors, but the majority of the power is probably used by the ESP and the LED.
I've been pretty successful with just using bluetack to attach wires for flashing. Not sure if it'd work at this scale though. Maybe a drop of superglue in case one wants to avoid soldering.
Another tip I have is to use wire wrapping wire for this type of wiring. It's much less intrusive as its extremely thin and requires less heat to get the solder joint done.
Great idea about blu-tack for temporary connections. This was wire-wrap wire shown in the video
@@SuperHouseTV Oh crap, so the parts were a lot smaller than the video made it look like. The wires seemed huge!
A super analysis and tutorial. Thank you!
Jon has his Louis Rossmann style cavity search gloves on!!!!!!!!!!!
I bought the T1s when they first came out with the idea of re-wiring to fit them. I never got around to that - instead I used a battery powered servo to flick the switch. Now it seems I can swap those out for these! I guess I'll find out if they work for the 2 way switch arrangement (hall lights upstairs & down controlled from a downstairs switch & a an upstairs switch!).
£12 delivered seems about the going rate!
Great Video Jon. Always look forward to seeing what you come up with next, you always cover things in good an clear detail. I did figure the anti flicker was a given for leds/ cfls etc. Also Nice UAV Futures intro ;)
Very interesting explanation. Thank you.
Great explanation! Learn so much watching your videos. Thank you for putting together great educational videos so we can understand all of this.
The wasted power with the anti-flicker unit is only when the load is active, with the load inactive the power used should be basically the same as with an incandescent bulb (e.g. dependent primarily on the Sonoff's power draw). Given that the loads that would need it are already much lower power than the incandescent - it seems a reasonable trade-off.
Hi Jonathan, Dont suppose you have figured out using two of these switches with one light like in an upstairs downstairs 2way switching arrangement. Great vids and many thanks for your help
@ 27:20. If you are using a regular incandescent (without the series capacitors) you are ALSO using extra power.
You are using this extra power to power the ESP.
My guess is that it is not much different, because the smart power control chip in the smart switch will pull as much power as it needs to keep the 3.3V and/or 12V regulated.
Hi, great video.. I have sonoff touch no-neutral 2ch. On one channel i have connected LED light, on second FAN.. but once i switch fan on, switch loose power..
I will try replace LED bulb with normal bulb, and try it again with and without "load capacitor"
Hello, thank you for the great video full of helpfull information! For me only one question left: how much power it cunsumes via antiflicker device? As I understood, 0.3A is a maximum, but what about real consumption?
Excellent
Thanks or that Jonathan - now, I'll still waiting for my Itead sample - the original Itead T1 UK version definitely does not work without neutral - and now you;ve sparked me off, I have a oair of ZemiSmart equivalents - one works perfectly - the other comes on and stays on, with or without the supplied capacitor (without it runs at half-power). I tested 6w LED E14 lamps - in the UK of course, on 249v mains power.
Very good explanation. Nifty technique.
300mA x 240V=80Watt extra use. That's substantial
I tested my one - it uses 300mA when on (without a bulb) - but only 1.6W of power. It uses only 1 to 3 mA when off (with the wifi indicator off) with a power consumption of 0.2W. The PF for the switch circuit with the anti flicker device must be around 0.023 when on.
so, you can have a notification when a bulb is burned, setting the LWT on tasmota mqtt: if a bulb fails, device will loose connectivity, and that's it :D
Oh, that's clever! It will only work if you don't use the anti-flicker device though
Nice video ! Also played with this switch like a month ago and I like it a bit more than Minitiger switches U may also check those the power logic on them is different, but they are much easier to flash and also have the RF433 chip onboard so u may use that one too (useless for me though) But the big plus for this one is that I ripped off the front plate and cover and put instead an LCD with touch the power board here can produce enough mA to drive that, so all in all I like it but it only comes in 1 way gang and Minitigers can also be found in 2 or 3 way that's a bit sad but fine for now as I have the best of both worlds. DO recommend this 1 way switch if u need some more mA to power up some other logic and u have only one bulb to control :)
I haven't seen the minitiger switch before. Thanks for the tip
The trick of using the capacitor as anti flicker is nice, but in long term it is pretty expensive...
A simple calculation shows that it consumes another 70 Watt when the load is on: 0.3[A]*240[V]=72[W], and this will usually be added to a bulb that consumes only 12-15W (standard LED / Florescent bulb) - so it increase the electricity usage by ~5 times (!!!)
Please correct me if you think that I am wrong :)
Thanks for great vide @superhouse. Can I ask you if you would do a video about how to make a low-pass filter for Sonoff Touch ? As some of mines are ghostswitching (and it's not because of mqtt) I would like to give it a go, but not really sure how to do it on that board .
If you could do that, that would be absolutely terrific!
Thanks a lot for your time!
the power passing through the capacitors is mostly capacitive which is NOT charged by Power Utility companies for Home Usage
So if everyone was using these switches , would the sub-station power factor need correction ? 🤔
@@peterfitzpatrick7032 probably, since each switch consumes near 60VA, this can easily add up to huge numbers and cause over-voltage specially when the network load is low
Great video, wonder if sonoff will bring out rectangular versions of this.
Hey Jon, I know we've been saying 240V for ever, but I believe the standard for Australia/NZ changed to 230V in 2000. (Except Qld which is transitioning and WA)
Yes, technically the Australian standard says 230V, but Victoria still uses 240V as the nominal voltage I think. I should get in the habit of saying 230. When I do that, I get people saying "but it's 240V!" and vice versa
SuperHouseTV fair call! 😀👍
When the chip is pre-flashed, it does make sense to not have easy access to gpio 0, but then I wonder why gnd/tx/rx/3v3 headers are still there, as these should not be needed in normal operation. If the reason is to be at least somewhat tinker-friendly, adding gpio 0 and reset would be much preferable.
Good point.. I suspect the chip also needs to be flashed in factory. They could do that before they mount the chip on the board, but i think its common for manufacturers to give themselves the option to re-flash just in case.
@@dozog - Well, even the device manufacturer would find it hard to re-flash without access to the gpio 0 pin. And when cheap devices are in mass manufacturing, reflashing is highly unusual, especially as cloud-connected devices (like the Sonoffs with eWelink) are expected to receive OTA software updates at the consumer location. As long as there is room on the PCB, providing soldering points or holes is not much of a cost, and therefore I'm surprised if penny-pinching causes removal of gpio 0 but not gnd/tx/rx/3v3. Of course, the latter 4 pins may be useful for debugging (like connecting a serial terminal), but debugging is not much of a real-world use case with these cheap devices. While reflashing thru soldered wires is still feasible (I've done this in a smartplug), it does become a lot harder than when accessing the needed pins is easy. Therefore I wonder what the use case for providing gnd/tx/rx/3v3 without gpio 0 is. Speculatively, I suppose that it might be for attaching some sort of test rig on the assembly line.
Yes, flashing before mounting was exactly the scenario talked about in the video here.
@@JohnnieHougaardNielsen Yeah, i got that :-)
Thanks for the response.
@@JohnnieHougaardNielsen I would guess it is for the test bench. with 3.3v/Tx/Rx/Gnd, you can put it on a nail bed and verify the function, operator can touch the switch to verify it works well. I am not sure the ewelink firmware post something through serial when working though but that may be an option.
Was that soldering iron a High School graduation gift? 😁🇦🇺
stumbled across this YT channel. i just discovered a gold mine! :D
Given that GPIO0 is used on the capacitive output of the touch switch, could you not just press the capacitive touch switch whilst performing the reset?
This is a great question, and I should have answered that in the video! The answer is no, because the timing doesn't work out. The ESP8285 has to see GPIO0 low at the moment it boots, and the touch sensor IC takes a little while to start up. It needs to calibrate its own input, and then it takes a while to detect the finger. So by the time it starts asserting GPIO0 it's way too late, and the MCU has already booted up
So, couldn't you delay the reset long enough to let the capacitive sensor to boot or alternatively a switch to CPU VCC or reset while powering the capacitive sensor, to slow CPU rebooting when you want to program it? Guessing VCC and maybe reset are easier to solder to than GPIO00. Ah, is the capacitive thing an I2C device? Then delayed boot won't work.
Hi, I noticed that after flashing the tasmota image the wifi LED works only when the circuit is closed (the bulb is ON)(see @40:00). Why is that? and how to fix it, I mean the led helps as a night indicator.
Ok please forgive my limited electronics knowledge but I have a question: why wouldn’t they have just placed a transformer in line with the active. When the switch is closed the transformer coil would induce a magnetic field across the secondary that they could use to run the circuitry? Is this not an easier option than pulsing the output?
Cheers MB,
16 Perth
That would create a voltage division which would result in a dimming of the light. Same as putting two resistors in series. That's why they have to turn on/off the lamp very fast and briefly, or else the sonoff would turn off it self everytime it turns on the lamp which would also result in the lamp turning off immediately.
Amazing video !
Clever ! Thanks
did i not understand? @27:25, upto 300ma of wasted current? at 240v isnt that 60w!
Correct, but since it's a capacitor not a resistor, it's reactive power (also known as imaginary power), not real power. You'll have to read up on power factor to understand it, but to explain it briefly, AC voltage is shaped like a sine wave. During each sinusoidal cycle, at the times when the AC voltage is higher than capacitor voltage the capacitor charges, and when AC voltage is less than capacitor voltage the capacitor discharges into the grid.
So for a purely capacitive load, no actual power is consumed by the capacitor because theoretically it feeds back exactly as much power as it takes in. Of course, no real-world capacitor is purely capacitive, there's always some parasitic resistance that consumes power within the capacitor itself and makes it heat up.
The advantages of this method is that the capacitor acts as a resistance to the AC current (whose "AC resistance" value is known as its impedance, and is calculated using the formula on the paper, 1/2*pi*f*c) without consuming as much power as a resistor of the same value, and some current is still flowing back and forth to charge and discharge the capacitor (so the sonoff can use it to power itself).
Interestingly, utility companies will charge large consumers (like factories) in kVA (kilo volt-amperes) - which includes both real and reactive power - because the wiring and distribution system has to be built large enough to supply even the reactive power, which they have a lot of, due to inductive/capacitive loads (like big motors). For small consumers like you and me, they charge us in kW (kilo watts) - which is the REAL component of the power only.
@@LMF5000 so what is amount of waste in watts approximately?
@@krzysztofkujawski6513 For a purely capacitive capacitor, zero. For a real-world capacitor, I believe you can get a very rough approximation by calculating RMS current squared times Equivalent Series Resistance (ESR - you get that value from the capacitor's data sheet) - basically I2R.
If you want a more accurate answer, start by reading this (since capacitor characteristics change with frequency) - passive-components.eu/capacitors-losses-esrimpdfq/
Perfect explain !!! Very good job !!! Well done !!! Does anyone know where can we found the ic lp3669 or what else can I use ? I have already 2. Switch with burn that ic.
flashed 3 of these I took my zero volts from the connector on the back of the board and touched the other end on R10 while booting all 3 working 100%
---- 24:35 ----- Shouldn't that V be replaced with the RMS voltage there? Not the peak to peak voltage? Because if that 0.3A figure is correct, wouldn't that mean that there is a constant consumption of (240 x 0.3 =) 66Watts per each such device you install??? That's a very considerable consumption to add to your house. It negates any LED replacement bulbs you might have!
Great video! Really interesting content.
Looks like the perfect nightmare for any radio amateur
That's a great point. I didn't even consider that! Now I wish I had a spectrum analyser handy
Jens. I agree 100%. A very relevant question. Too much RF pollution already. Does it claim to comply with a ny EMC standards ? I doubt it.
@@daveh7914 an unshielded unfiltered triac like that are going to transmit a lot of RF noise all over yer property
@@SuperHouseTV if you got a small pocket AM radio you can tune it between stations and use it as a crude RF "meter".
Excellent video! Very understandable and thorough. Thanks for making it!
Has anybody who has installed a "no-neutral" switch noticed any noise from the capacitor when the light is switched on? I have both the Sonoff switch as in this video and some Lonsonho-branded Tuya switches and both of the capacitor units (both entirely different) "buzz" when the lights are on - tried with both CFL candle bulbs and low-power LED G9 bulbs to the same effect. The capacitors are silent when the light fitting is off.
I was thinking about getting into automation of lights and instead of automating bulbs and having physical switches turned on all the time smart switches make more sense and lights can still be turned on or off manually. But I would like to keep the switches physical (as in rockers, not touch points) because it works even with gloves, when hitting it with something else than hand etc. Aqara makes switches like that but they blink the buls every couple of seconds because they don't have the antiflicker capacitors (and I have no idea how to calculate those). So I was thinking if Sonoff switches could be hacked with a physical button instead of a touch button. From this video it seems the only necessary part is tripping the GPIO0 to turn on or off the relay (I don't know if momentary or latch is needed but I can find that out). But you mentioned soldering to GPIO0 is difficult so now I'm kinda undecided.
By any means, do you think it could work?
P.S.: I would need those switches without neutral, my wiring is old and I have concrete walls in an apartment so pulling more wires through is not an option unless I want to mess up the whole apartment.
Great video thanks. But what a big soldering iron you have
This guy is really good
Just curious, could you use the cap sensor to bring gpio0 low and just solder onto reset?
That's exactly what I wanted to ask. Could it be that simple?
Hopefully they come out with an AU version. Please post a video if they do 😊
Great video just getting into smart wiring etc after being an electrician for 30 years, we never used to run a neutral to the switch as this was bad practice(to stop your diyer switching the neutral instead of the line conductor) however in the past 5-10 years wiring a neutral to the switch is the norm,it's also alot easier when it comes to testing the installation .,could these sonoff be wired in series to a 230v coil controlling a contactor for larger loads?
Any reason you didn't scrape the via that's on the other side of your resistor instead of the middle of the track? Would give you a slightly better pad for soldering the wire to .
I did, and then the whole pad came off and the resistor fell off, which is why I had to add the bodge wire!
4:16 with the Tasmota firmware the LED is not ON when the light is OFF, is that configurable? (7:49 normal operation)
Great in depth stuff Jonathan, nice to see we have a bleed resister over those caps, I have in the last few days flashed my first sonoff basic with the latest Tasmotta firmware, however no matter what I do I can not get it to work with Alexa, someone has said that it no longer works, do you know anything about the matter ? thanks Stuart
If there is a triac, why the need of relay?! If the triac is not continuously on ( while the lamp is on) but instead switches on/off as you have shown with the scope, that means the lamp will not light at the rated power but less. To fully understand what is going on, there have to be an oscilloscope measurement at the lamp terminals ( not only the gate of the triac) .
hey mate, great work!
just curious if you have made any videos on ceiling fan automation?
I am wondering how the triac power stealing approach works for dumb LED lamps? They of course have an internal power supply of their own that tend to mess with old style triac dimmers.. different situation, I know. I was watching intently thinking a LED bulb would get plugged and tested after the CFL... but left hanging
It works fine with LEDs
Brilliant.
Good video .. Please tell me what is R12 and C14 in board? How many ohms?
Why not connect directly pad touch to pin IC ? ( just like ttp223 Schematic!!)
Regarding that antiflicker add on: So in a AC circuit the caps act as a resistor so the sonoff can get some power. But why not use just a plain old resistor instead of the caps? (I'm assuming there is a reason for that!)
AC and DC circuits behave in very different ways, so the resistor analogy may have been misleading. In this configuration the caps act as a current limiter, but they won't get hot like a resistor would. Resistors emit extra power as heat, while the caps accept power while charging and then simply stop accepting power when their plate voltage rises to match the supply voltage. They can sit in that state while using zero power (and emitting zero power) until the source voltage starts to fall and they begin discharging. Short answer: this is more efficient.
i have been thinking... why does ITEAD need the relay at all ?? The triac is used to interrupt the LINE in order to steal power. Does that not imply the triac can be used to switch the load off entirely ? In fact, since it is synchronized to the 50 Hz zero crossing it could be used to phase delay the leading edge of the voltage waveform, thus it could also function as a dimmer (at least for incandescent and some dimmable LED's). What am I missing ?
Wheres Part II of your rock solid wifi system got to?
David Scrimgeour yes I bought a load of Unifi stuff since then, still be good to see how it’s setup etc
Partly filmed, but a few things happened that made me put it on hold. It'll still be done, and I've already mostly filmed part 3 as well
My question is regarding the capacitor and resistor. Those capacitors are connected to the bulb, wouldn't the discharge over the bulb and actually, there is no need for a resistor?
Hi! Great video! thanks... What is relay type used in main board?
@SuperHouseTv: Which valuedoes the R10 have? I unfortunately damaged mine too.
hi thank you for the video. i have a question about triggering the triac. Since its triggering the triac after mains voltage is above some treshold. How the circuit knows when to trigger. Is there a zero cross detector.
Hi can you make a video about the Sonoff iFan. I’ve read mixed things about fan control and being bad for the fan motor / potential fire hazard.
Great Video!!! Question: does the 50 hz oscillation dims the amount of light that the bulb outputs? Doesn't it work like a PMP diming the light?
Great explanation. I love your videos!
One question: Why did you call it "active in" and "active out"?
Shouldn't it be "*live*" -- that's what the "L" stands for on the label: L = Live; N = Neutral!
Is there a different terminology where you live?
“Full bridge rectifier!!” #electroboom
Can I clarify how the anti flicker device is connected to the load? Is it as simple as connecting one end to the live and the other the neutral? Also does the direction matter? Thank you.
Sounds like you have worked out my old problem with Version one of the units. Now the mini will it work the same? I did the same with the button a couple of years ago. I ended up going to Espurna instead of Tasmota.
How much power can the high voltage side supply? Could I power a Pi Zero with a small touch screen for example? What I want to do is build a small unit that would replace a standard light switch and hack something like the Sonoff to power the Pi and the screen for a Home Assistant dashboard.
tin the wires before soldering to tracks will make your life easier
BTW The easiest way to put this thing into bootloader mode just connect it to your USB to serial power it up and tap the button while holding it just take with the other hand a pair of small tweezers or a small screw driver and touch the small resistor R3 the one u've soldered the reset wire to. NO SOLDERING NOR pcb torturing is required! I'm crying when I see what you've done to that PCB, why??? if u need a video with the steps I could upload one, but it should be easy enough just try it ;) works like a charm every time and I can not understand what was so hard that all others had troubles putting this device into bootloader mode.
I got a Sonoff SA-025-2.4G-1C with a single capacitor without the resistor. Does the stored capacitance on the resister not discharge itself across the load of the lamp over time? Thank you.
Hi, question,
From when it's trying to take power in case of 3 gang module? All bulbs or just the one in L1?
Very good explanation !
Is it so that the triac is in series with the relay contact ?
What is the purpose of the relay if the switch can control output power ?
I missed something probably...
The switch is no longer a switch. It's a toggle. It is not hooked to the mains.
Nice video! About five years ago, I tried to build a non-neutral circuit like that, and I did. But it did not work for me because this triac technique only gave me 100ma maximum and my circuit with wifi required more than this. So I would like to know how this circuit can delivery more current. Is there any other video or information source that explains this magic? Is it possible to record a new video about this theme? Thanks in advance, the video is really very useful.
Do you think two of these in series would work? L-out from the first into the L-in of the second and then to the light? Is there enough power left through the 50hz triac to power a second t4eu1c?
Do you need the anti flicker device with led bulbs?