Clearance - 30v 3A TENMA PSU

I did some initial measurements, and it's very good : )

I did the measurements at 5V (since that's a popular voltage!) and at 30V.

5V set voltage:

At no load, the output was 5.0088V, and at 3A load it was 4.9969V

30V set voltage:

At no load, output was 30.039V, and at 3A load it was 30.029V

So, set voltage is spot-on, and it seems to handle the max current fine too. Noise was low, I couldn't observe any ripple.

This is noise at 30V with no load attached, and 3A load  (there is some faint garbage just visible on the trace background, but some of that will be due to scope probing, and some of it can be removed with a ferrite):

Switching the output on (i.e. power supply is already on, but output is off, and it is switched on via SCPI), with no load attached:

Switching it off:

Tests with a constant current load attached:

Next I tried the same thing as above but with a 3A constant current load attached:

And switching the output off:

Tests with a resistive load attached (home-made load, Building a Miniature 300W Speaker Simulator  ):

I set the supply to 24V and attached a 8 ohm load, so that the expected current was 3A. I switched the output on via SCPI:

As can be seem, there was an anomaly just after 15V or so. Maybe it is switching ranges there, since maybe the time period seems consistent with typical relays switching. Still, this looks great, no overshoot, and the output is flat within 20msec or so.

Here's the output switching off:

That anomaly at switch-on happens at different voltages too, down to about 15V. Here it is at 16V:

I then set the load to 4 ohms, and set the supply to 5V and switched the output on via SCPI. Again, no overshoot, and output was flat within milliseconds:

I tried lower popular voltages too, like 3.3V:

and 1.8V:

So, all this looks great. There were some negatives though:

Power off/on glitches

If the supply is powered on or off, sometimes there is some output visible.

Here it is powering off. This was the worst-case I could capture. Most of the time, it was less than this, but (say) 5% of the time switching off, I saw this:

Powering on produced junk for a shorter time:

Adding the largest ferrites I had (not very large - one was about 1 inch OD and 0.5inch long, and the other was about 1.5 inch long and 3/4 inch OD) to the output wires, improved things a lot.

With a load attached, it's better:

Other issues:

The main issue is that there's no easy way to switch the output on or off, without powering up or powering down the entire supply. However, if the output is switched off via SCPI, then it seems to retain that forever*, and the output can be switched on by holding the voltage knob down to lock it, and then holding it down to unlock it. But, there's no way I can see to easily switch the output off from the front panel : ( It requires SCPI to do that.

Maybe I'm missing some trick on the front panel. (EDIT - problem resolved, see Steve's comment regarding Mode 3).

The workaround is to always use SCPI to control the output on/off. That could be an annoyance for some, it's certainly not convenient. Worst-case, someone could write some Arduino code to control via the rear serial, to provide this functionality.

* I could be wrong here.. I've not spent enough time to understand the behaviour totally - these were just quick experiments to get to know the device.

To do:

I've not tried to examine behavior under overload, with CC mode nor the electronic fuse (OCP) mode, nor changing load. Also, I've not explored how accurate the current and voltage readback capability is. These latter two items are nice accoutrements, that I figured can be measured later.

Summary:

The performance seems to be great so far! It is very low noise, and practically no ripple even at 3A load. The voltage set is accurate. I'm really happy with it. The lack of front-panel button to easily switch the output on or off is a limitation, but I cannot help feeling I'm maybe missing some button trick to do that (EDIT - problem resolved, see Steve's comment regarding Mode 3).. The problem is not unsolvable, there are workarounds, but it would be nice for this to be easily controlled from the front panel, since the capability is accessible via SCPI using the rear-panel USB or serial connections.

I did some initial measurements, and it's very good : )

I did the measurements at 5V (since that's a popular voltage!) and at 30V.

5V set voltage:

At no load, the output was 5.0088V, and at 3A load it was 4.9969V

30V set voltage:

At no load, output was 30.039V, and at 3A load it was 30.029V

So, set voltage is spot-on, and it seems to handle the max current fine too. Noise was low, I couldn't observe any ripple.

This is noise at 30V with no load attached, and 3A load  (there is some faint garbage just visible on the trace background, but some of that will be due to scope probing, and some of it can be removed with a ferrite):

Switching the output on (i.e. power supply is already on, but output is off, and it is switched on via SCPI), with no load attached:

Switching it off:

Tests with a constant current load attached:

Next I tried the same thing as above but with a 3A constant current load attached:

And switching the output off:

Tests with a resistive load attached (home-made load, Building a Miniature 300W Speaker Simulator  ):

I set the supply to 24V and attached a 8 ohm load, so that the expected current was 3A. I switched the output on via SCPI:

As can be seem, there was an anomaly just after 15V or so. Maybe it is switching ranges there, since maybe the time period seems consistent with typical relays switching. Still, this looks great, no overshoot, and the output is flat within 20msec or so.

Here's the output switching off:

That anomaly at switch-on happens at different voltages too, down to about 15V. Here it is at 16V:

I then set the load to 4 ohms, and set the supply to 5V and switched the output on via SCPI. Again, no overshoot, and output was flat within milliseconds:

I tried lower popular voltages too, like 3.3V:

and 1.8V:

So, all this looks great. There were some negatives though:

Power off/on glitches

If the supply is powered on or off, sometimes there is some output visible.

Here it is powering off. This was the worst-case I could capture. Most of the time, it was less than this, but (say) 5% of the time switching off, I saw this:

Powering on produced junk for a shorter time:

Adding the largest ferrites I had (not very large - one was about 1 inch OD and 0.5inch long, and the other was about 1.5 inch long and 3/4 inch OD) to the output wires, improved things a lot.

With a load attached, it's better:

Other issues:

The main issue is that there's no easy way to switch the output on or off, without powering up or powering down the entire supply. However, if the output is switched off via SCPI, then it seems to retain that forever*, and the output can be switched on by holding the voltage knob down to lock it, and then holding it down to unlock it. But, there's no way I can see to easily switch the output off from the front panel : ( It requires SCPI to do that.

Maybe I'm missing some trick on the front panel. (EDIT - problem resolved, see Steve's comment regarding Mode 3).

The workaround is to always use SCPI to control the output on/off. That could be an annoyance for some, it's certainly not convenient. Worst-case, someone could write some Arduino code to control via the rear serial, to provide this functionality.

* I could be wrong here.. I've not spent enough time to understand the behaviour totally - these were just quick experiments to get to know the device.

To do:

I've not tried to examine behavior under overload, with CC mode nor the electronic fuse (OCP) mode, nor changing load. Also, I've not explored how accurate the current and voltage readback capability is. These latter two items are nice accoutrements, that I figured can be measured later.

Summary:

The performance seems to be great so far! It is very low noise, and practically no ripple even at 3A load. The voltage set is accurate. I'm really happy with it. The lack of front-panel button to easily switch the output on or off is a limitation, but I cannot help feeling I'm maybe missing some button trick to do that (EDIT - problem resolved, see Steve's comment regarding Mode 3).. The problem is not unsolvable, there are workarounds, but it would be nice for this to be easily controlled from the front panel, since the capability is accessible via SCPI using the rear-panel USB or serial connections.

Great work shabaz! Your tests show its performance to be better than I expected at the price.

I agree about not being able to switch the output off without powering down. I had already been searching around for a way to do that. It's a bit dangerous I think (for the components I mean).

Actually, there's Mode 3. I think that does what we want. [added edit] "At this time, the CURRENT adjustment knob becomes the output knob".

Hi Steve,

Thanks!

I've got a theoretical solution, but it would need a hardware mod : (

The idea is, to use any small chip (e.g. Arduino or MSP430) to act as a pass-through for serial transmit. In other words, it would have one input, and one output, and it would be inserted in the path for the data received by the serial port on the PSU. No need to implement dual UARTs, the pass-through would be done via interrupt or forever loop, since the baud rate is low. It would just repeat whatever logic level is received, with no regard to protocol.

But, if a button was pressed (connected to another input on the Arduino) then temporarily the pass-through would be broken, and the Arduino would send the output on/off command and then re-connect the pass-through. Like an ability to send an out-of-band message via button-press, but retain serial comms to any attached PC at all other times.

I like the PSU enough to want to buy another, it could be worth implementing this feature, if there really isn't a way to do it from the front panel today.

(EDIT - problem resolved, see Steve's comment regarding Mode 3).

Hi Shabaz,

Those are some great measurements you have undertaken, very useful reference to anyone looking at buying a supply and wondering how to compare it. I'm sure Tenma will be pleased to see their product looking so favourably.  I'm glad you and Steve seem to have picked up such a useful and well priced bench unit. It is always worth having a look through CPC Farnell and searching on 'clearance' to see what is available. I usually skim through the top items in a few categories that are 50%+ reduced. I picked up a reel of PLA (750g 2.85mm) for £11 (UK prices) incl VAT and postage plus some SMD basic diodes for 2p each!

Rod

Hi Steve,

Oh that's fantastic! Well spotted. Problem sorted! Now it seems there are no shortcomings to this PSU at all!

Hi Rod,

I'm definitely going to look out more often for these offers! This is great to hear that there's other interesting stuff too.

The Tenma supply is back in stock in a few weeks time, so that's good.

I'm very happy with the power supply (even more so, now that Shabaz has tested it), so thanks again for letting us know.

Yes, I lurk in Bargain Corner too! (But I don't normally look in the "Test Equipment" part of "Bargain Corner" which is why I missed this power supply). It's also worth glancing at the various printed CPC catalogues that seem to arrive on my doorstep every Saturday. Sometimes they have offer codes in there which give you a much lower price than if you search for the same item on the online catalogue (e.g. packs of 40 AA batteries are often half-price in the catalogues).

Hi Shabaz, that’s a nice set of measurements! I was wondering what you’re using for the 3A load? I’ve got a project driving a very inductive load at a constant current of around 2A using PWM at around 500Hz, and want an electronic load for testing with, so any recommendations would be much appreciated!

A

Hi!

The constant current was using a BK8600 load (review here), it's got an easy-to-use front-panel interface, and simple to automate if desired too. If it's for personal use, there's a lower-cost Tenma load, but I've not tried it.

It's my third BK Precision piece of equipment, it seems very usable to me. Most of the electronic loads (including the Tenma one it seems) use large binding posts, so normal 4mm banana test leads may not fit, but the end of the test lead could be cut off and a M6 lug crimped on.

Thanks Shabaz, I’ll take a look! I’ve got a BK Precision signal generator and am very pleased with that, so am likely to stick with the brand too.

A