Are the $20 bench power supplies ok?

[Peabody]

Dave did a video about them, and apparently thought they were ok for the money.  I'm thinking of the DP20V2A:

https://www.ebay.com/itm/153901898094

I thought I would power it from a laptop power supply, which I think is about 19V.  Anyway, I just wondered if anyone had tried them and had any major problems.

[ataradov]

They are fine for generic switching supplies. I would not power very sensitive circuits from them, but for general use they work.

Go fo the one with UI with big colorful numbers. That UI with a lot of small numbers is very confusing and harder to navigate. UI in general takes getting used to, but it is usable.

[wizard69]

Dave did a video about them, and apparently thought they were ok for the money.  I'm thinking of the DP20V2A:

https://www.ebay.com/itm/153901898094

Well it really depends upon what your intentions are.    These are not bad and they are not great from the standpoint of quality, however they are a bargain for DIY supplies of various types/uses.

I thought I would power it from a laptop power supply, which I think is about 19V.  Anyway, I just wondered if anyone had tried them and had any major problems.

Depending upon what you are doing it might make more sense to use/buy a fixed voltage supply.    For example if you are powering Arduino's or Raspberry PI's, a fixed supply suitable for the device might be the best route.   Other avenues into electronics likely could use a variable supply, often more than one supply.   So are those $20 supplies OK, it really depends upon your needs.

[bidrohini]

The power supplies are OK as far as I know. But I do not find ebay to be a dependable source.

[inse]

Get one with two knobs, otherwise clicking around for basic settings is a pita.

[Peabody]

Thanks very much for the replies.  I was going to breadboard my design for an IR remote control with two LEDs in series with constant current, and I have an LM317 supply with a 10-turn pot.  But at the last minute my 9V supply crapped out.  And it looks like I could get the cheapest bench supply for not much more than it would cost to replace the 9V wall wart, and would pick up the current limit function in the process.  But I would need something like this maybe once a year, so I don't want to spend a lot on it.

I have a 12V supply, but it's actually 16.3V with no load, and I could be drawing as much as 1A, and I don't want the LM317 to lose its smoke.  So maybe it's just time to get a bench supply.  I just wanted to be sure the cheap one isn't dangerous to me or my circuits, and it sounds like it's ok, with maybe a little more ripple than would be ideal.

[ledtester]

Couple of points...

1. For cheap wall warts I check out thrift stores like Goodwill, Salvation Army, etc. (since you are in the US). Also check if there is a local "Free Geek" near you -- I know there is one in Portland and Minneapolis.

2. For testing your IR remote I think I'd just use batteries. First since I suppose that is what will be powering it in "production"; and second, I'm not sure a power supply will respond in the same way to short, intensive current pulses that batteries would. Note that the response of a switching power supply is limited by it's switching frequency.

One more thing... to increase the current capacity of linear regulators you can always add a pass transistor. Some examples:

https://www.eevblog.com/forum/beginners/variable-regulator-pass-transistor/msg1059349/#msg1059349

https://www.maximintegrated.com/en/design/technical-documents/app-notes/3/3333.html

[Peabody]

Thanks for those suggestions, ledtester.  I understand that a power supply may not duplicate a battery's response, but for testing the constant current circuit, which is the big issue in doubt, I'll just be firing two 8us bursts 18us apart (38KHz, 30% duty cycle).  I think a honking big capacitor should be enough to prevent any sag.  Then if I get that working, I can do a reality test with a battery.

I decided to go ahead and order the cheap bench supply, mainly because I think it's time I had one.   But in the meantime, I think I can use the LM317 supply with the 16V feed.  It should tolerate the short test bursts ok.  And as you say, I can always use a pass transistor with the LM317 if I can find one in my junque box.

[BeBuLamar]

I have one of those and it works fine. There is some noise but not bad. It runs quite cool. The interface is small but is fine. There is the larger version made by the same Ruideng company but they are more expensive. I bought my from Banggood but I think the best place to buy those is from aliexpress. Ruideng selling their stuff there.

[IDEngineer]

My "road supply" uses one of those and a laptop DC supply. With 19VDC in, it can output up to ~18VDC. I designed and 3D printed a little enclosure for it that has banana jacks for test leads.

To quiet the output I experimented with various ferrite beads and caps on the output leads. Got it down to under 250mVpp noise at ~12VDC, definitely not lab supply territory but honestly quieter than our products typically see in the real world so more than adequate.

I've made a battery clamp input lead for it recently, after being out on a boat without an AC inverter. Am going to add a cigarette plug input lead too. The idea is to have controllable DC in nearly any environment.

Bottom line: With some tweaking these little units are pretty useful. I'd replace mine if it failed, hard to imagine more convenience in a smaller package.

[Peabody]

For those few cases where you need to reduce the ripple, I wonder if hooking up an LM317 regulator after the bench supply would eliminate most of the noise.  I'll try that when mine arrives.

[IDEngineer]

That's a pain if you want adjustability. Plus regulators don't usually have much high frequency response and the noise is switching noise with lots of high frequency content.

[Doctorandus_P]

You may also find this project interesting:
https://github.com/rin67630/Drok-Juntek-on-steroids

It basically takes the power supply with a crappy interface and adds a TFT lcd, Wifi and rotary encoders to it.

[BillyO]

Do you have a decent oscilloscope?

Switching power supplies are great until they're not.  What I mean by this is, when they are relatively free of significant load, cheap SMPS units can be just fine, but when pushed to their "specified" power rating can get cranky.  I have measured as much as 20% (p-p) HF noise on some "stellar" units when pushed.  Just be aware of that and if equipment or experiments that are attached and are pushing the limits of the supply suddenly start acting like rotten eggs then have a look at the noise before tossing the DUT out.  Especially circuits employing digital.

[Peabody]

I do have a scope, but most here would say it's not a decent one.  It's a DSO150 kit scope.  I paid $20 for it, then converted it to battery power.  Of course it would depend on the switching frequency of the power supply, but I think the scope would give me some idea of what's going on.

[BillyO]

I do have a scope, but most here would say it's not a decent one.  It's a DSO150 kit scope.  I paid $20 for it, then converted it to battery power.  Of course it would depend on the switching frequency of the power supply, but I think the scope would give me some idea of what's going on.

The DSO 150 should let you see the switching, but not all the dangerous stuff.  It may be enough though.

In the shots below, the first two are a cheap 3A 5V SMPS with a 100mA load on it.  Admittedly, this is a pretty horrendous example but it does serve to illustrate.  You can see the 67KHz switching noise @ 155mV in the fist one (15us period).  The 2nd shot is the same signal and conditions looking a little closer.  We can see there are components of that noise that go up to 23MHz!

In the next two shot we have the same PS but at a 67% load of 2 Amps.  It's still switching at 67KHz, but now the noise is 1.8V!!!  That's 40% of the output voltage.  Looking deeper in the 4th shot we see that there are significant components of the noise going into the 550KHz to 23MHz range.  The very high frequency noise is still there and at much larger amplitude, but it's being dwarfed by what you see in shot #3.



 

[kripton2035]

I have the DC32V/3A from the ebay auction.
here are the scope screenshots at 16V with no load and with a 700mA load (computer charging)
they are quite good for the regulation and the price IMHO.

[kripton2035]

and here the same at 16V 3A still computer charging.

[Peabody]

I'm not an engineer, and have no training in filters, but I wonder if there couldn't be some kind of multi-stage L/C filter module that you could insert inline when noise is important.  Well maybe it would have to be inserted within the feedback loop to be effective.  Anyway, as small as these bench supplies are, there would be plenty of room in the box for additional capacitors and coils if they would do any good.

And since I already have the LM317 regulator built, with a 10-turn pot, I will be experimenting with adding that, and setting the switching supply to maybe 3V above the LM317 output to help keep things cool.

[IDEngineer]

One thing to keep in mind is the real world is seldom as quiet as a lab supply. If you're designing products and not doing fundamental research, some noise on the rails isn't the worst thing.

EDIT: Speaking of filtering, here's what mine looks like inside. Couple of wraps through a ferrite on both leads combined with some additional capacitance.

Just looked at my notes from March 2018. Best performance of 350mVpp at 10V/2.5A was achieved with three turns (two visible wraps) of each wire through its own Kemet ESD-R-14S NiZn ferrite. Yes, 350mVpp ain't great but it's a field power supply and the real world isn't as quiet as a lab.