How [NOT] to MAX test a power supply!
Вставка
- Опубліковано 20 гру 2024
- Keep exploring at www.brilliant.... . Get started for free, and hurry-the first 200 people get 20% off an annual premium subscription.
How hard can it be to max power test a power supply? After finding out my laboratory power supply will deliver more volts than advertised, I have decided to test how many watts it will deliver. The issue is that I don't know how? In this video, I will give it some attempts and learn from the mistakes... Will the power supply deliver the watts it is supposed to?
My Patreon-page: / brainiac75
Did you miss one of my videos?: / brainiac75
FULL MUSIC CREDITS
Time codes: 0:01 + 7:30 + 11:03
"Long Note Two" Kevin MacLeod (incompetech.com)
ISRC: USUAN1100420
Time codes: 0:44 + 8:44
"Energizing" Kevin MacLeod (incompetech.com)
ISRC: USUAN1900040
Time codes: 2:52 + 6:18
"Perspectives" Kevin MacLeod (incompetech.com)
ISRC: USUAN1300027
Time codes: 3:58 + 10:42
"Adding the Sun" Kevin MacLeod (incompetech.com)
ISRC: USUAN1900041
Time code: 9:43
"Peace of Mind" Kevin MacLeod (incompetech.com)
ISRC: USUAN1200099
All music above licensed under Creative Commons: By Attribution 3.0 License
creativecommons...
Time code: 5:50
Mix of two tracks:
1) The Shimmering by fran_ky (freesound.org/s/237363)
Licensed under Creative Commons 0 license
2) "Spacial Harvest" Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 3.0 License
creativecommons...
ISRC: USUAN1100653
#Max #Watts #PowerSupply
time to miserably fail my exams by using 90% of my brain cells, powering off when realising the paper has a back page
Bro that feels like dying 😭
Genuinely a question. What's the difference?
*Windows XP shutdown noise*
1:52 “I can even watercool it” this made me laugh more than it should have
Mike's electric stuff has an amusingly simple means of load testing, just put some enameled copper wire in a bucket of water. The enamel is thin enough it can conduct heat from the wire into the water, but prevents it from corroding from being a dc potential in water. Tune the amount of wire to the desired resistance, and then if you need a longer runtime, use a bigger bucket of water.
its exactlly how i tested my DIY 2000w labbench powersupply with those chines dc to dc setp down transformers and server powersupplies in series i think they are called RD DPS5020 modules 50v at 20 Amps and 2 of those for 50v 40 Amps, and what can i say, i boiled that water as fast as a kettle and those modules actually were handeling it quite well.
which is just crazy for how small they are.
no consumer electronic load would have been able to handle that amount of power, only if i would use like 5 of these shown in the video which just wasnt viable for just a simple extreme load test.
Nice idea! So simple, yet functional.
As long as the wire isn't all too thin compared to the current running thru it and the water temperature is kept constant, the resistance will stay constant. It's more important to keep the temperature constant for copper though. Wire wound resistors and metal film resistors are usually using high resistance metal alloys that is made to have small variations with temperature - even though the resistance can still change a little bit, if they get very hot. But copper may have something like a 5% increase just going from for example 20 to 50°C (for example if the water temperature is allowed to rise).
But it also depends on your application how large deviations in resistance that can be accepted. If some variation can be accepted, heating wire can also be a good option if a lot of power need to be dumped - since those wires can get red hot for long periods without damage 😃Many of those alloys can go from room temperature to red hot with only 10% resistance change - some like Kanthal A1 is as low as 5 % increase going from room temp to over 1200°C (which is like yellow hot).
i wouldn't call it amusingly easy, as it requires a fair bit of experimentation to get the right resistance with a reasonable length of wire. But it definitely is amusingly cheap and accessible.
How does using wire in mineral oil compare to using water for the heat sink?
@@FindLiberty water has higher thermal capacity, will boil if overheated, is cheaper, and is less of a mess to clean up. Oil is an insulator, so it is better for high-voltage application. It may be a fire hazard if overheated.
It is normal for lab power supplies to go a bit above. That way they make sure you can get up to 30V and not 29.99V
Yeah, mine is listed as 30V 5A, but it's actually 31V 5,1A.
I really like the electronic load unit. Never seen one before. I also note that when using the 3 ohm resistor it still hit 31.7V @ 9.68A which is over the rated 300W. So even though its cheap, its pretty damn good too.
You can't get one any more because the company CONRAD is shutting down
I would love to see the benchtop power supply's output on a scope. Wonder how noisy it is.
it will be tricky for him to measure. Even as an EE: You need to specify a test scenario like the resistor, because it does not create its own artefacts, but also the resistor does not care about PSU quality.
With the switching design, there might be "a lot" of noisy HF AC hash riding on the DC and even some 50/60 Hz ripple at high current loads, especially when set near max voltage. In most cases, it won't be a problem with less sensitive circuits. The specs should give the numbers you'll need to decide price/performance sweet spot. The little, low cost switching boost reg boards (not stepdown/buck benchtop P.S.) can be hideous sources of noise.
I have a linear power supply from UNIT-T and tested it. It is mostly fine, except for a 50Hz base from the grid. But for a switching PSU, the noise could get much higher. The common solution would be to add capacitors to the outlet, but it hurts the PSU's ability to respond to really fast changing loads and may cause a power loss for sensitive systems that sometimes spike in current.
@@yuxuanhuang3523 Long live linear - 9 out of 10 capacitors agree! lol
We had one of these supplies at work. They are the most noisy lab power supply i have ever seen. You get what you pay for
Crazy to think the first time I saw one of your videos was when I was a sophmore in highschool extremely excited to show my physics teacher your content. Fast foward 3 years, I go to school for physics, 4 more years, I graduate, and 2 more years since then working in the field :) (wow insane, 9 years of your videos) thank you so much for your inspirations and learnings, this type of experimental and hands on content feels so much like a consequence free (for me haha) lab in university and i just love it
very proud to be a patron after all this time
That a “reverse power supply” is a thing is so amazing 🤩
8:17 I love the 3210. Used to play Snake on it all the time as a kid.
This is a critical teaching method that some people do not get right: you must show failures. Human simply do not learn properly without failures being involved in the lesson.
Yep. Failures and understanding why it failed is often more enlightening and memorable than just seeing something that works. Thanks for the early watch!
absolutely. I am worried about electronic loads because Dave Jones created an oscillating circuit using the constant power.
Here, we saw the constant resistance mode and also saw oscillation.
Oscillation can kill both devices.
Meanwhile honestly: Neither the cattle nor the resistor where a "failure": that was just approaching the limit in an easy to understand and fail safe way.
Another quality video! Keep it up, we enjoy these!
Thank you very much! I do put a lot of time into these videos - glad is appreciated. Much more to come!
Interresting. I rember Dave Jones (EEVBlog) torture testing a PSU with an electronic load set to constant power.
Electronic Loads are awesome, but as you have already experienced: It is very very easy to get them to oscillate and from there, you can kill power supplies.
6:49 notice the lego model of the International Space Station!!!
This was a rather interesting video, and not what I had expected just by seeing the thumbnail. Keep it up!
Really interesting video, i was genuinely surpised about the instability issue. It would be really nice to see with a scope the ripple on the voltage waveform.
I was concerned for a moment, because that is nasty use case.
Dave Jones toured a PSU with an Electronic Load in Watt mode.
Using Resistor mode is a bit less concerning, but still:
As said here: Both devices try to regulate/limit current and voltage, but the electronic load also tries to regulate ratio of voltage to current as secondary controll loop, using the current as primary control loop.
That can oscillate. Oscillation can kill the PSU.
I would not blame the PSU for dying under such a test.
Collab between Braniac and Elctroboom? Blow electronics up with lasers!
Could be fun, though I prefer to keep my electronics in a functioning state ;) Thanks for the early watch!
@@brainiac75 agreed in hindsight, stuffs expensive 😂
@ElectroBOOM That would be so fire , literally 🤣
When testing linear PSUs (not sure if this one was linear or an SMPS) maximum output does not equal maximum stress. That is because the PSU has an electronic load in it self to dissipate the rest of the power. For example if you dial in 10V@10A, but connect a load with 0.5Ohms resistance, the current limiter in the PSU has to dissipate half of the power. With a 1Ohm load, it has to dissipate nothing.
This is certainly not linear. I have this exact PSU and it is quite light. It is obviously a relatively cheaply built SMPS. (And why wouldn't it be an SMPS? Sometimes lab power supplies are dual-stage switching+linear, but I've never heard of one being pure linear)
No matter what happens in that day, if Brainiac uploads, it’s a good day.
Thank you! It is also a good day for me when a video is finally released. After a month of thinking and working on the same video, it is a relief to publish it :)
@@brainiac75 aww dude, i cant imagine the stresses of deadlines for yt and stuff, especially if you have no ideas lol. I love the content dude, keep it up!
every day is good when real madrid loses
@@brainiac75
I need an affordable 240v to 120v transformer that will handle a 1400 watt power tool. I am in Italy. Can you suggest a supplier? Ebay has many, but the quality?
This is a harsh test for the transformer, particularly the primary windings as it needs to deliver max current.
However, high power linear power supplies usually have multiple secondary winding taps at different voltages, and then use a BJT transistor circuit to curtail the last few Volts of that tap to reach your desired output.
Say the transformer has taps at 9V, 18V, 27V and 36V. The BJT circuit always needs a couple of Volt to operate. Hence the 9V would be used for outputs up to say 5V, the 18V tap up to e.g. 14V, etc. This could be a reason why the PSU reaches 32V instead of 30V, as the transformer could output less V with a low line voltage, and the BJT circuit could heat up and have a higher voltage drop.
However, notice that in order to properly stress test (e.g. "burn in" or "burnout" test) the PSU, you would need to get the maximum voltage drop over the BJT. In my example figures, the PSU outputing 32V@10A=320W from a 36V tap, means the BJT circuit is only dissipating (36-32)*10=40W.
But note that if you shortcircuit the output, that the PSU would likely use the 9V tap, "curtail" 9V to reach 0V/10A output, and internally dissipate 90W(!).
If you want to grossly overload the primary side of the transformer, then don't use a shortcircuit but a voltage at which it just about needs to use the highest tap, at max current. So say it changes taps over for Vout
... The two devices fighting because of ohm's law has to be the most unintuitive but obvious error. I want an episode just on various caveats like this where obvious things are easy to miss.
this was fun, but i like my Raiden RD6018, it's a 1080w max power supply at 60v and 18a but limited to 600w because its powered by an old 600w power supply from an old decommissioned pc that can luckily provide 12vand 50a with all of its rails tied together and a boost converter in-between that can handle upping the voltage from 12v and 50a to 72v and 20a all for 60USD from a direct from china app on a sale. no fire so far especially considering i used plywood and 3d printed PLA to build the housing and mounting brackets.
You can probably modify the PSU to output more than 12V but will have to replace the capacitors
ohms law simply describes a relationship. The symbology doesn't matter. You can rearrange the equation to put whatever properties you want on whatever side. For example I=V/R
Haha, the Nokia 3210 reference, loved it.
I just immediately thought of the Nokia when I realized it could have been model 3210. Such a classic :D
The oscillation has nothing to do about on which side of ohm's law equation they are. it's all about how the feedback loops of the load and the power supply are tuned. It's complicated, we'd have to dive into all of the internal circuitry to properly explain it. The simple explanation might be, both are optimized for constant-voltage operation, because it is the more common application.
Exactly
Ohm's law can be V/R=I, making R and I on different sides of the ecuation and it will still wooble
Honestly, this is exactly what I thought of, after I saw this video. Ohm’s law can be arranged in three forms, so provided explanation sounds weird.
Now I got an electronic load, and I can confirm, that in my setup, power supply in CC and load in CR are stable.
8:15 As a Finn I always appreciate the mentioning of our lost Art of mobile phone industry.
Sticking with the crude and cheap method, the "massive cooling" would be as simple as a couple of canned air dusters. If you turn them upside down, you get liquid air to help cool your resistors. I've used this method on CPU dies when troubleshooting without having to keep removing a cooler. (Tin shot glass set directly on the die and dribble in liquid air).
Alternatively, the average capacity of a standard refrigerator is 350 watts and the capacity of higher end consumer grade chest/upright freezers is 700-900 watts. Drop the temp on your refrigerator to the minimum, pop in the capacitor with the wires running through different sides of the door seal and you've got sub-ambient cooling perfect for your experiment.
As for your added radiator, the air gap would make it more or less completely useless. Maybe you didn't show this step, but using some thermal paste or thermal pads would greatly improve the cooling capacity, as would a small fan.
Just be careful with the canned "air" dusters in electronics and related experiments :)
Those cans are typically filled with a refrigerant like butane, not regular air (compressed air requires very thick and heavy tanks). Therefore they tend to be highly flammable or even explosive.
Having a connector/switch spark shortly after using one of those cans would certainly make the day more interesting ;)
This isn't just a theoretical risk, either - I've done it myself. I used one of those cans to blow dust and crap out of a 4-way trailing mains socket on the floor, then plugged something into that socket around 10 seconds later. The plug pins arced when they made contact, igniting the butane and engulfing my hand in a flash of flame with a loud pop. I was completely unharmed, thankfully, but I hadn't used the can very much.
The thing is, _I already knew it was butane!_ I just absent-mindedly forgot for a minute 🙄
It's actually really common for lab bench power supplies to have a higher max voltage and/or max current than advertised. I'm not entirely sure why, but I guess it's just to be able to guarantee it can handle the rated power.
It is just to make sure they get get to exactly 30V and not 29.9V. They are cheep analog things inside so they are not very precise (they rely on you turning the knob to the exact value)
Generally it is to compensate for wire losses at higher currents, you may need the extra headroom to get 30v at the other end of the connection
I actually have a power supply similar to that one, and it is overspec as well. i think it is because if you turn up the course voltage adjustment only, that gives 30 volts. But if you also turn the fine voltage adjustment up it adds an additional 2 volts to that, giving 32 volts.
I can recommend keeping a fire extinguisher nearby. I have one at my desk, never had to use it but it's always there just in case. I also have one in the kitchen too, since cooking involves hot stuff.
I actually have 2 of thease power supplys and they are amazing for what they cost me, my ones where actually able to output almost 11a at 32.1v!
for how much did you buy them?
I had a Wanptek KPS3010 PSU, but it failed shortly after getting it. I don't know if I overloaded it or the design was really crappy, but I kept getting major voltage fluctuations (like you), that would eventually stress my projects.
And one time, I got a 15V spike while powering a 5V load, and as I was concerned, I turned off the supply, only for it to go bang a while later when I tried turning it on.
After days of troubleshooting I found out that the main IC (TNY277) was overheating, and this caused the switching MOSFETs to short out and blow the fuse, rectifier, thermistor and other parts.. I was powering a powerbank circuit at that time, and it completely fried it. A major power surge went through, as the regulation failed and the switching transformer let the surge through.
I'm not trusting Chinese stuff ever again... especially after paying 90$ for it...
Chineseum crap in that case 😠
That power surge wasn't from the mains. The power supply made it, along with making a loud bang!
I'm gonna have to make myself a new power supply, as I can't tinker without a reliable power source.
@@samuelfellows6923 And my unit had weird "seizures" at times since the start. Maybe I had a defective unit? And the major PSU "seizure" happened while adjusting the current. I had 3 minor PSU "seizures" before.
Of course, I couldn't return it, as I was curious and I voided my warranty and stripped the screws....
I have the same one but a 30/05. 24 bucks on eBay new 8 years ago. I noticed the internal fan only came on for a second to test at power on and never other times when it ran for a week straight. I lined-up the fan from a timer to blow straight into the front grill of it. They run together so why not let one fan cool 2 devices. If you're wondering why a timer would ever need its own cooling fan its b/c it was a one rpm synchro motor turning a metal disk with numerous contacts. Every 8 seconds it was off for 2 seconds. I couldn't find a simple timer that could switch like that so I made one. But it runs hot.
I wonder how many years I've been watching this channel and it's still too damn entertaining ..
EDIT: just checked .. ever since 12 years ago huh .. have had to change accounts 3 times 😂 nostalgic
So I'm not the only one who uses a kettle as a load dump! Sometimes you get so much solar production that you have to put it somewhere so the kettle it is, my coffee consumption has gone up dramatically since I started this LoL 🤣
8:00 This is because your Load has a regulation circuit wich handle than not that good.
May try ohter brands ;)
I thought, that everybody had forgotten Nokia 3210 long ago, untill now😅
I came here for science, but get much more. Thank You😁
Nice video and presentation.
For power supply we need to account for max Watt to derive max loaded VxA
Electronic loads are usually switch mode and can easily cause power supplies to be unstable at certain settings.
A few hundreds of low-esr uF across the supply usually fixes this problem.
Dude, you helpt me a lot 😅 had the same issues thermal stressing a buck converter
I didn’t see it in my brief peruse of the comments, so I’ll say I am interested in the efficiency video
You should try powering a hair dryer with that power supply
Because hair dryers use a DC motor and a rectifier
The thing where the power supply actually delivers more volts that advertised also happens to a 60v 15a power supply my dad borrowed from a friend. It can deliver 16a and 65v! I (=my dad) also own a old 32v 4a analog power supply, therefore it also has analog displays. Their markings go from 0 to the advertised limits, however its display arrow things go beyond that by a little bit too! (They also have a mode to double the shown numbers, as a way to see small currents easier)
Messing around with cheap power supply! Hope to see some magic smoke!
Would be sad to blow up my only lab power supply ;) Thanks for the early watch and comment!
Very cool video and great explaination of basics electric laws. btw. It would be nice to make a video about efficiency of that power supply, for instance make a graph every few volts and see how much power does it consumes vs how much power it gives.
Not me saying hi back after he enthusiastically said “HI!”
Ooooh, that ISS model is something I'd definitely like to see more of! I also follow your facebook page, would you post more about it there?
I’m interested in the efficiency of the plugs and in products, my class is currently doing a topic about it so I think it would really help
watercooling the load? genius!
Note: You can actually overdrive these cheap PSUs. They have internal calibration potentiometers that can be "tuned" to whatever value, within a particular range. So, a 30V/5A supply could theoretically get up to about 37V/6A if lucky.
I like your dragon electrical clips, Shirley must have been the result of a Karen! Nice touch!
My Rigol DP712 (50V/3A) will go up to 53V/3.2A and provide that reliably. I've always wondered why it's like that.
the voltage is obvious: if you need 50V on the load, you need more on the source. If you can not get 50V at your load when the source has 52V, then it is your problem.
If the psu barely makes 50V under no load, then it can not provide 50V to the load.
Normally when I test load the amperage, I would use car (12V)/ truck (24V) halogen bulbs either in parallel or series rather than using load resistor/s because from the brightness of the bulb I can however estimate it's voltage and wattage.
Great Video as always! 👍😀
10:56 - I wouldn't mind seeing some magic smoke being released.
It's been years since I've seen a video from you. Looks like your channel was unsubbed by that UA-cam bug s few years ago.
Glad you're still around. Missed your Scandinavian voice 😂
There are bonus points for the Nokia 3210 joke, to be collected! 😄
Great video!
No flames, I am disappointed!
Clever way to ask for a like. You got it.
You should have put the 2 resistances in parallel... 😁
very interested in efficiency!
Have you planned to make a video about laser safety glasses ? 🤨
Hmm, how did I miss this Super Thanks? Thank you very much, mnikpro! I am hesistant about making videos about laser safety glasses. When is enough, enough to be safe? In general, I would only trust expensive glasses from trusted brands and sellers.
Working at the limits = Letting the magic smoke out. FUN!!!
Awesome video.. I love electrical related videos
"A resistor is usually boring to me. In my mind it is just a badly-conducting wire."
Yes please, efficiency vs loading.
I've got a very similar variable power supply, so I'm pretty interested to see what the outcome will be
Good info
11:02 yes please maje a video about it!
How does this electronic load control resistance electronically?
probably with a transistor, that is shut on and off very quickly, in series with a power resistor. The on off rate determines, how high the average resistance is. (how much average current can flow, with a set voltage). Probably way more complicated, but i think this might be the basic conceptm
i have a cheap power supply like this aswell, works fine tbh!
brian in the hazard warning going 'syke'
It is a switching power supply, it being able to supply the rated power is expected almost regardless of price or size.
LOL WHAT?
@@johncoops6897 If this was linear power supply then it would be worth making a video on.
@@toalan - either kind of supply can "exceed" or "not meet" ratings.
Shouldn't a labbench power supply be safe, even at the edge of its powercapabilities?
I mean I don't expect it to supply the power for long but I expect it to cut the power once it heats up to much.
Also another thought, I live in Germany and was able to buy a psu from the same brand at conrad, shouldn't the unit be safe regardless of setting according to consumer safety laws?
The extinguisher trigger looks like a arrow in the thumbnail 🤣
good equipment.interesting
Pretty cool!
When I tested a power resistor rated 100W 2 Ohms with a lab power supply 30V x6A=180W I cooled the resistor in water.
Could you use 2 electrodes places in a solution of salt water as a good load?
creating hydrogen gas next to a fully load power supply?
using table salt and wondering about the health effect of the gases created?
Running into issues with the non-linearity of an salt water bath?
It will also corrode the electrodes
7:25 this part cracked me up
can someone explain to me
why connecting dc to ac rated device ?
Very cool!
I think you should test some longwei/topshak psu's, they're even cheaper and done very well in my usage, despite their lightness and size
They might have worse noise though.
@@Scyth3934 might be true, but that would not matter for this channel.
An electro magnet does not care. ... but the electromagnet will kick the input protection circuit to hell and back when switching improperly, so... maybe stay with better power supplies?
@@sarowie anything i ran on them ran fine, and they're cheap as hell when discounted
that includes Arduino's and other MCU's, even on the NPS605W (60V 5A model)
it's best to avoid the 2 dial ones tho, those are pretty finnicky to get the right voltage and even will often temporarily peak far higher while turning it up (K3010 for example), but the NPS (course and fine tune knobs) models are perfect for my usage and don't have that issue
since i don't really trust the 2 knob models too much i mainly use them to either charge something like a lead acid battery or one is in use at a local slot car club where they had 1 PSU down and 10A for a single track was enough for all classes they run there
Add a waterblock and replace the heatsink inside with it
3.00 ohms for 77.7 feet of 5/8 inch wide, 0.015 inch thick, stainless steel banding material, at temperature of 140F.
WHAT?! No fire? No POP! ..ooh wait.. This isn't PhotonicInduction!
I think for you to build one power resistor,you only need a lot of cheap tiny resistors and a heat sink.
Please leave that running for an hour ;)
BASETech eh? I have this exact power supply.
I'm sure it's a cheap rebrand of something else though because the brand BASETech is only sold at CONRAD Electronics, which is now shutting down.
Way more interesting is how long it can do 10 amps at 1 Volt dumping 310 Watts in its internal heatsink...
Really interesting video. 2x👍
Big Clive love Ohms law
Let's max it out.
nice Video
At 32V 10A maybe the ripple is excessive. Something to test.
You find 500W resistors to be cool, I find them to be hot, however.
I have exactly the same power supply. It‘s actually not that cheap. There are much, much cheaper ones on amazon. Anyway, i use it to charge my batteries and the only Limitation ist, that the terminals get so hot, that the terminals literally melt the plastic around them which leads to a bad connection. So dont overdo the amperage. 8 amps is the limit.
I'm using the rock bottom 30V 10A switchmode lab power supply, the one that comes without knobs for fine tuning and mine has already starting doing funny things at around 10 to 15V (probably the most used range) when voltage sometimes spikes to around 20V. I guess the potentiometer is going bad.
Interested
There are variable resistors for lorry trailer brakelight flash speed adjustment too. 12v but maybe a few in series? They've got a screw down contact that slides up and down a coil bent in a "u" shape. Heavy duty chromed wire, might just be a spring. Not sure, haven't looked at mine in a while. Got a different vehicle and didn't need it.
to avoid being blamed for false advertizement they say 300 watts and 30 volts at 10 amp then if it can do any more that is a bonus.
I’m guessing why they said 30V and 300W becuase if the power supply blows up they can’t get sued or in trobble
lol - sueing a cheap Chinese manufacturer?
The 32V is a margin so that you can deliver 30V / 300W at the load.
@@sarowie Wouldent be surprised if someone tired
Lad festen begynde og bare rolig jeg er brandmand