Testing a cheap power supply
Вставка
- Опубліковано 1 жов 2024
- This cheap Aliexpress 12 volt power supply showed up in the last mailbag, and I wanted to run it through its paces to see how bad (or good) it is.
I hope to push it harder in the future to see where it maxes out, but I've go t to wait for more power resistors to arrive first.
-----
My tip jar: / pileofstuff
My other tip jar (channel membership) : / @pileofstuff
Buy me a coffee (or a beer): ko-fi.com/pile...
If you are OK with me getting a commission on your purchases, feel free to use my affiliate links:
Ebay: ebay.to/33IF7PA
Banggood: pileofstuff.ca/...
Some other links to products may be affiliate links
odysee.com/@pi...
#ElectronicsCreators
-----
This isn't the only way to do it.
It's probably not the best way to do it.
But it's how I did it.
As several people have helpfully pointed out, I got the NTC/PTC thermistor mixed up.
Mea culpa.
Also, I have some additional power resistors on order so I can test this thing further toward it's limits.
Not worries, great video anyway.
What I didn't see is proper rfi filter so better don't use nearby ham radio operators or general radio receivers. Ntc is OK, inrush protection. Hope that's an y3 cap
That NTC thermister is for inrush current limiting into the cap at startup
Nice teardown! Safety seems quite legit for me as a former SMPS developer. Of course you can not check the Y-capacitor but the spacing together with the slots in the PCB may be okay. My next point of interest would be the transformer itself. Primary and secondary windings need double or reinforced isolation, so enamalled wire of primary must never touch enamalled wire of the secondary. The standard transparent isolation of the wire itself counts only as one isolation, so either they must use specially qualified "triple insulated wire" that is clearly yellow or they need additional insulation sheets between the windings.
Transformer quality mostly is a quite good indicator for the security awareness of the producer.
I bought 3 of these a while ago and tested them with an electronic load. Each was barely able to deliver about 1/3th of the advertised power before breaking down. I got a refund from the seller.
The isolation is provided by the slot and the optocoupler, which looks ok. The other point is what is the rating of the cap, is it a safety cap or just a high voltage one.
I'm surprised that they didn't specify it for the full input voltage range so it could have a worldwide market, after all its just a lower voltage limit, maybe they had issues with the lower level. Or they have never tested it in the first place and expected it will work ok at 180V. The lack of components for conductive or radiated emissions hints at it having never been tested to any standards as well.
The only problem with the input voltage level being lower than rated is whether the components can handle the extra current in the input circuitry at full load. Its about an extra 0.5A flowing through the input and the transformer. Dont sound much but its about 50% more.
200W/180V =1.1A
200W/120A = 1.6A
So as long as it starts up ok at 120V and you keep the load to around 100W, change the isolation cap to the correct type it should be usable.
But always wear wellies if touching the output when its live.
likely due to the higher power units. The higher power units are a forward converter where you can't exceed 50% duty cycle. Lowering the mains voltage would require the higher duty cycle to reach the rated voltage at high currents, which is not possible.
Okay, those green and gold caps are known to not be what they claim to be. I don't know who makes them, but I think they are rebadged Xmas decorations... The temp rating is out by about 50%, had many dry out across smaller SMPS - better to replace with known good caps next voltage rating up.
The massive dents in one of them don't exactly inspire confidence either... and I'd bet my house that the primary cap is NOT a Nichicon. These are red flags for me, they indicate lack of attention to detail. But then it is cheap.
@@jasejj I think that many people would be tempted to choose such a cheap PSU, then use the purchase-cost saving to buy good quality caps and swap them in for the fakes.
I doubt I would trust this for anything much more than a low power bench supply or LED strips.
Looking forward to part 2 😃
I didn't quite see why the anti tracking distances would be insufficient, particularly as I see loads of other issues...
There a re several issues, but as long as you don't have too high expectations, it'll do what I need it for.
if it had noise filtering eg a common mode choke and poly cap, it would be an ok psu
Hmm, this is a 200w device, and you tested it with just 36w. For 200w the price would be great, not so much for 36w. Mine dropped to 11V at 5A and turned off at more than that.
Yeah, I ran out of load resistors to test it further. More are on order to give this thing more of a challenge.
@@pileofstuff awesome! I tested mine (apparently the same model) with an electronic load according to their claimed specs. As long as anyone is ok with using them at 15% max of claimed specs, they are ok. Thanks for your tear down and looking at their internals. Once you start drawing more than 15%, they fail. The initial voltage drop is worst because i.e. LED strips will still turn on, however at much less brightness. In general, with these Chinese gadgets, build quality isn’t necessarily bad but the sellers attach specs to them that are fictious. Same with Chinese batteries for example: for what they do and cost, they are not bad at all. However they almost never have the claimed capacities, typically 30-50% is closer to truth. When a Chinese device works well and is still a bargain at just 20-40% spec, then for me it’s a buy. :-)
The NTC resistor is actually a "Soft start" resistor. The resistance is initially high when it is cold and decreases as it warms up. So it limits inrush current. PTC resistors increase in resistance in response to temperature and in special cases are used like resettable fuses.
Nice.
I am always impressed about the tiny little converter ICs which are used to drive a single MOSFET for 200 W of conversion power. Lately I have ordered a bunch of synchronous buck converter SO8-Devices which are capable to drive directly 100 V at 5 A for 250 W of power through this very tiny package.
Seems to perform as advertised, good for LED's, and a reasonable buy for what you paid for it.
Yup. It's certainly not a high-end product, but it's better than I expected for only $9.
Gotta judge these tings in the proper context.
@@pileofstuff they sell this exact model at times for under USD 3 (including shipping). Don’t know how they do it. amazing to realize the price range at Ali goes from 1x to 10x for most items.
hi, you were saying about the NTC, isnt that inrush limiting?, it would be a PTC for operating current limiting?
Ptc is inrush protection/ limiter
NTC at input is definately inrush current limiter. When it is cold, resistance is higher (at device plug in) and after it warms up, resistance goes down.
In low power applications PTCs are used are current limiters, "resetting fuse".
the ntc is to prevent the cap's blowing fuses on startup, they charge up slower - common on psu's
It works! You could always add some parts if you think you need them later.
Non existent fuses are great. They never fail.
So glad I stumbled upon your channel recently being a ex-pegger I miss the accent please keep up the great work.
that supply looks OK I would trust it more or at least a much as the little plastic covered LED drivers. I use some really big power resistors built some into a kind of decade box totally passive to push thing till they start to sag
A very thorough test. 👍
It seems pretty solid to me. Definately mislabled on the input voltage. I'd trust it.
I would add a fuse holder or use a fuse on the input but otherwise good enough for LED strips.
Many of these cheap units have copper clad aluminium transformers and secondary side inductors (for Forward converters). They cripple these units in a thermal sense, as there's lots of losses. I bought a 400W 24V unit for shits & giggles, and it struggles to even reach 80W without dropping 5% at the output. It gets absolutely roasted at 50% load, and shuts down with time.
And yet my 24V unit looks better than yours. The assembly and design is nice.
I've got some additional power resistors on order so I can test this thing harder. We'll see what it can really do eventually.
I used a similar psu for a bench power supply. Put it in a box with a fused ac input. Output to a 12v auto socket. I have a fused 12v plug from radio shack that has banana jacks that I use for 12 volt stuff and a good quality car phone charger for 5 volts.
That's a cracking idea.
I think I may borrow it :)
@@mattsadventureswithart5764I’ll take that as a compliment. Thanks 😊
Beyond all this pile of stuff, this time an absence of a proper DC Load became apparent.
I've got one, but it doesn't have enough capacity to fully test this thing.
I have more resistors on order!
100 veiws in twenty minutes, doing good.
and the surface mount passives you were worried about were protected pretty well by the routed out slot?
Tha slot helps, but there are still some close distances that could cause issues if things went pear-shaped.
Good value
My bet is that is not a class Y capacitor in reality but justa standard 1KV blue capacitor 🙂
I think that DiodeGoneWild would give this power supply slightly dodgy rating, just for this one
Before I looked at it, I was expecting "super dodgy".
It surprised me.
There should be 2 output diodes. Only 1 has been installed. With only 1 output diode, i don't believe the 16. 67A current rating that the label says. You should replace it with a matched pair.
Possibly. We'll see what it can do after I get some additional resistors to load it down harder.
The 9mm dia. NTC does seem undersized...( physical ) .
Your test @ 3-5 amps load & it was " hot " on the th. camera..
I don't think it will sustain at the max. of 16 A dc load.....
doubt very much about the life of both i/p & o/p caps too...
Was the heatsink temp.rise with 3A load measured?.. that's another doubtful component.
@@analoghardwaretops3976 Once I receive my order of additional power resistors I plan to see how close to the claimed 16 amps it can get.
For something this cheap, I never expected it to be able to life up t that claim. As it is, I'm pleasantly surprised what it has been able to achieve.
You have to take these things in context of the cost.
@@pileofstuff Well it claims around 400W @16A load...and for it's size & cost ..the old 300-400 W , XT/AT computer smps heatsinks & cooling requirements..gives plenty of clues
on it's build capability..This I infer from the assumption that the switching freq. of then & now is about just 1 order of magnitude higher today..so the low cost here is really a big compromise.
Doesn't look like the voltage doubler for multi-voltage is there. Wouldn't there be 2 input bulk caps?
No. Its a flyback SMPS. Only half bridge SMPS have 2 input caps.
You wouldn't use it on any thing flammable? Unless your house is made of stone or steel that is a hard one.
I just wouldn't use it at all! Certainly not on the UK's 230 V mains. Just not worth the risk.
The NTC is not what you think, it is an inrush current device and is supposed to be hot when normally working.
I am aware. I simply misspoke. See the pinned comment.
What about 200W test, these devices likely will not do it sustained!
I ran out of power resistors to push it any further.
Moire are on order.