60V/20A bench supply for $130? Bought and measured!

[mcdn]

I've been after a better bench supply for a while, as my usual one runs out of amps and/or volts when testing power amp modules. When I saw this 60V/20A supply on sale I thought I had to try it. We are talking US$130 shipped for 1200W of power. Too good to be true? Let's find out!

I have a 2ohm dummy load that can sink that power, so I hooked it up and used a multimeter to check the voltage from 5V-40V. Accuracy was within 1mV even when delivering ~800W, which astounded me:



What about noise and ripple? Here's a composite graph of the output at various levels, with baselines for powered off and powered on but not engaged. I've shown it in Volts as I think that's most useful in this context, but the REW mdat file is attached if you want to play around. There's a broad noise spectrum but it's less than 1mV total until we get to 40V output.

It's also well built, with very hefty screw terminals for high current outputs and matching 4mm^2/12AWG cables supplied. The control buttons have nice detents. Looking at the inside there's a large heatsink and the high current outputs have suitably large wires internally. The fan didn't come on during my testing, but then I did only run it at 800W for maybe 2 minutes. There's even an RS232 control input for automation but I didn't test it

All in all I am amazed by the value and quality for not much more than a hundred dollars.

 

[Hayabusa]

SNR > 80dB !!

 

[mcdn]

SNR > 80dB !!

To be fair it only has to generate DC, but yes I was astonished by the result. I initially measured using my standalone digital scope but couldn't see anything as it only has 8 bits of resolution. Then I tried with my PicoScope oversampling to 12 bits and still couldn't really see anything. Finally I put the Cosmos ADC in there.

 

[SSS]

Well built power supply. Since it is a switched mode one did you measure the dip voltage and recovery time after an instanenious current demand like a amplier module would need when a big bass pulse occurs?

 

[mcdn]

Well built power supply. Since it is a switched mode one did you measure the dip voltage and recovery time after an instanenious current demand like a amplier module would need when a big bass pulse occurs?

As a regulated switched mode supply it will deliver exactly what it says. You want 90V, not going to happen. You want 30A for a millisecond, maybe but not very interesting. You want 60V and 20A forever? Yep, and with incredibly low noise and ripple.

 

[Hayabusa]

As a regulated switched mode supply it will deliver exactly what it says.

No, it will take time, if the current demand is going instantaneous from 1 to 10 amp , there will be a short dip and some regulation ripple towards the regulated value.
This is what SSS was asking about.

 

[AnalogSteph]

Being a bench power supply, I would assume it has a floating output. What about common-mode noise?

 

[mcdn]

Being a bench power supply, I would assume it has a floating output. What about common-mode noise?

How would I measure that? I've got some digital scopes and a good ADC.

 

[mcdn]

No, it will take time, if the current demand is going instantaneous from 1 to 10 amp , there will be a short dip and some regulation ripple towards the regulated value.
This is what SSS was asking about.

Sure, if the load resistance changes suddenly then any supply will have some delay before the voltage changes to match. What specific test would you like to see performed? Is there a standard to aim for?

 

[mcdn]

For those interested, I knocked up a command line tool to control the unit over its serial interface: https://github.com/mattclarkdotnet/nicepower-serial - It's basic but does the job of setting and getting volts and amps.

 

[AnalogSteph]

How would I measure that? I've got some digital scopes and a good ADC.

 

[mcdn]

@AnalogSteph sure, you buy me the R&S scopes, I'll do the testing Seriously, my scopes are good enough for messing around but not at all useful when the noise levels are so low to begin with. I can make an AC-coupled measurement using my PicoScope down to around 12*6=72 dB, would that be useful?

 

[mhardy6647]

As a regulated switched mode supply it will deliver exactly what it says. You want 90V, not going to happen. You want 30A for a millisecond, maybe but not very interesting. You want 60V and 20A forever? Yep, and with incredibly low noise and ripple.

EDIT: Ahem - where are my manners?! Thank you very much for testing and sharing your data and photo of the inards!

Define forever.
It does look nicely made, but the whole package smacks of too good to be true.
1) I wonder how long one will last in real-world use?
2) Did you try running a nice-ish, modern 2-channel power amplifier with it (i.e., the kind that needs a DC power supply) for, say, 6 hours listening to interesting music on interesting loudspeakers? By interesting music, something along the lines, e.g., of Holst's The Planets. Interesting loudspeakers? You know, some nominal 4 ohm loudspeakers, low 80s dB sensitivity, sadistic phase curves.
I realize this isn't the intended use for a bench power supply -- but it's the kind of stuff I'd do with one.

 

[Hayabusa]

How would I measure that? I've got some digital scopes and a good ADC.

Ok, just hook up a digital scope and capture the moment you attach the load to ramp up the current.
You can do the same when disconnecting the load.
It would be best to (dis)connect the load by a switch or a relay.

 

[mcdn]

EDIT: Ahem - where are my manners?! Thank you very much for testing and sharing your data and photo of the inards!

Define forever.
It does look nicely made, but the whole package smacks of too good to be true.
1) I wonder how long one will last in real-world use?
2) Did you try running a nice-ish, modern 2-channel power amplifier with it (i.e., the kind that needs a DC power supply) for, say, 6 hours listening to interesting music on interesting loudspeakers? By interesting music, something along the lines, e.g., of Holst's The Planets. Interesting loudspeakers? You know, some nominal 4 ohm loudspeakers, low 80s dB sensitivity, sadistic phase curves.
I realize this isn't the intended use for a bench power supply -- but it's the kind of stuff I'd do with one.

Haha, thx for the humour we need more of that around here! It does indeed fall into "too good to be true" territory, which is why I tested it and shared here. Get one while you can!

And I realise you were joking but I actually did run it at 100W into a 4ohm load for a few hours (so 20V/5A). The fan didn't even come on and the case was barely warm. To be fair some air was circulating nearby from the fans on the dummy load, but still very nice to see.

 

[mcdn]

Ok, just hook up a digital scope and capture the moment you attach the load to ramp up the current.
You can do the same when disconnecting the load.
It would be best to (dis)connect the load by a switch or a relay.

Can do. The load will be resistive, 2 ohms. Let's say 30V, or in other words 450W. What would you consider to be a good recovery time between the load being applied and the output getting back to 30V?

 

[Hayabusa]

Can do. The load will be resistive, 2 ohms. Let's say 30V, or in other words 450W. What would you consider to be a good recovery time between the load being applied and the output getting back to 30V?

https://www.electronicdesign.com/te...easuring-power-supply-transient-recovery-time

 

[amirm]

Nice find. One Amazon reviewer is saying it doesn't have proper current limiting:

"This power supply does not limit current correctly. When testing 60V with a 50 ohm resistor the output can spike above 100 amps on first connection. The power supply does not have an easy output on/off. When testing a dead short condition the power supply immediately failed, smoked, and shut down. Other bench power supplies will either stop their output or be current limited and drop the voltage. A bench top power supply is for testing and should not fail instantly in the event of a short."

 

[Hayabusa]

Nice find. One Amazon reviewer is saying it doesn't have proper current limiting:

"This power supply does not limit current correctly. When testing 60V with a 50 ohm resistor the output can spike above 100 amps on first connection. The power supply does not have an easy output on/off. When testing a dead short condition the power supply immediately failed, smoked, and shut down. Other bench power supplies will either stop their output or be current limited and drop the voltage. A bench top power supply is for testing and should not fail instantly in the event of a short."

There is a current control knob, so that does not work then?

 

[mcdn]

Nice find. One Amazon reviewer is saying it doesn't have proper current limiting:

"This power supply does not limit current correctly. When testing 60V with a 50 ohm resistor the output can spike above 100 amps on first connection. The power supply does not have an easy output on/off. When testing a dead short condition the power supply immediately failed, smoked, and shut down. Other bench power supplies will either stop their output or be current limited and drop the voltage. A bench top power supply is for testing and should not fail instantly in the event of a short."

Hey Amir, thanks for checking in!

I think with a lot of these items it's hard to know if you are getting the original or a knockoff or factory seconds that have failed QA, or something random Amazon has mixed into their supply chain because it's got a similar SKU. In this case the vendor "Precision Instruments Tool Store" does seem to have a good reputation on AliExpress. I was also sceptical hence the testing!

With shorted outputs my unit first switches to constant current mode as the voltage goes out of range, then shuts down and restarts once it can't control the current into the short. It also does have "an easy output on/off" - it's the big button marked "output" on the front panel, which is separate from the input power switch. As for output spiking above 100A into 50ohms, I think that person needs to go take electronics 101 again - that would be 5000V!