We all know that good power supplies are essential to an EE lab and we also know that the cheap ones are not the best power supplies out there, but the best don't come cheap, so I'm always on the lookout for affordable power supplies. I purchased this one several months ago, but couldn't find the time to review it until now.
Tenma is a brand owned by Farnell and this particular power supply seems to be a rebranded Korad KA3003D. It costs about $100 new, after tax, which is extremely cheap for a digital power supply. They have a mixed reputation so far, but they also appear to step up when issues are reported, so it might not be that bad.
|DC Output||Voltage||0-30 V|
|Line Regulation||Voltage||0.01% + 3 mV|
|Current||0.1% + 3 mA|
|Load Regulation||Voltage||0.01% + 2 mV|
|Current||0.1% +5 mA|
|Ripple and Noise||Voltage||< 1 mVrms|
|Current||< 3 mArms|
|Programming Accuracy at 25°C+/-5°C||Voltage||0.5% + 20 mV|
|Current||0.5% + 5 mA|
|Readback Accuracy at 25°C+/-5°C||Voltage||100 ppm + 10 mV|
|Current||100 ppm + 5 mA|
|Program Resolution||Voltage||10 mV|
|Readback Resolution||Voltage||10 mV|
|AC Input||220 V AC|
|Operating Environment||0-40°C, 80% RH|
|Dimensions||W 110mm x H 156mm x D 260mm|
This model doesn't have PC connectivity, but I think there are some that do have a USB to Serial adapter on the back.
- 0-30 V with 10 mV resolution
- 0-3 A with 1 mA resolution
Well, I'm not gonna lie to you, it looks cheap. Decent, but cheap. On a close examination of the front panel, you can see that there are 5 memory slots, but only 4 memory buttons, which is silly to say the least.
On the other hand, it feels very good and appears to have good mechanical construction. The power button is a real mains switch, the buttons are made from rubber but they have good tactile feedback and the rotary encoder seems to be one of those $1 cheappies - can't expect better than this.
The binding posts are of relatively good quality, however the hole that's destined for wires, is a bit too small for my liking. On the plus side they have a standard 19 mm spacing, so double banana plug adapters fit perfectly.
Adjustment to the voltage or current is done by hitting the voltage/current button. This puts the power supply in edit mode, which is indicated by one of the digits blinking. In this mode you can use the rotary encoder to change the value or the arrow buttons to change the digit you want to alter. After a short while of inactivity, it goes back to regular mode - meaning that if you rotate the encoder, nothing will happen. It's a bit of a drag, but I guess it's done for safety and it makes sense, because otherwise they'd have to keep one of the digits blinking.
What doesn't make sense, is that it doesn't remember which digit was previously selected, so next time you enter edit mode to adjust one of the values, it will default to the 1 V and 100 mA position.
For this test I used a Brymen 867 meter (0.03% + 2 for DCV and 0.1 to 0.5% + 20 for current) which should be sufficient, given the performance you expect from these power supplies.
Voltage was measured at the binding posts with and with out load. My electronic load is damaged and I was unable to test the current driving capability at the low volts, some other measurements were impossible as well, but everything was double checked with external multimeters. No calibration was done as I don't know how to do it or if it's at all possible from the front panel.
Most of the input values are picked at random, but in some cases I was constricted by my electronic load which couldn't do higher currents at very low or very high voltages.
|Set Voltage (V)||Load (mA)||Readback (V)||Readback (A)||Multimeter (V)|
The output voltage accuracy appears to be very good and it looks like it can drive the loads I tested it just fine - keep in mind they are pretty low loads tho - couldn't go higher because of the broken electronic load.
The constant current test
This is the same test I did on the Array PSU. It's purpose is to determine how accurate is the supply in constant current mode. I find myself using my supplies in this mode all the time and I bet other people do that too.
Knowing that current regulation is tricky and can get the electronic load to fluctuate together with the PSU, I only used the load for the 900mA test, while the other ones were made by shorting the output directly trough the meter.
First thing I noticed is that with the output off, the PSU is sourcing about 5 mA @ 8 mV. Those 8 mV are present even with no load connected.
|Set Current (mA)||Readback (A)||Multimeter (mA)|
|1||It stays in short, with out any load|
Aside from the poor regulation at less than 4 mA, it seems to do great. The display value appears to be a bit off, but the real value is much closer to the set value that I would have expected.
A closer look at the output
Ok, so far it looks very promising. Let's hook it up to a scope and see what we're really getting out of it.
First of all, I should note that it picks up a lot of high frequency noise from the environment, even when turned off. The same amount of noise I get from the Array PSU.
The first thing I noticed was that there are some repetitive spikes in the output voltage, even when the output is off, a big spike of about 200 mVpp that repeats with a frequency of 52 Hz and another one, of about 60 mV, that repeats with a frequency of 130 Hz.
Another thing I wanted to confirm was if there really was a voltage at the binding posts, with the output off. Turns out there, and it appears to be even more than the 8 mV we found earlier.
about 10 mV at the binding posts with the output off
Because of the high amount of high frequency external noise being picked up by the supply, current ripple was measured with a 20 MHz bandwidth and high resolution mode on.
Constant Current ripple @ 1 A
I think this qualifies as low enough, to say that it has clean output, both in CC and CV modes.
Ramp up and ramp down
For ramping up and down, I didn't want to use my electronic load so I wouldn't introduce oscillations that otherwise wouldn't be there, which is why I resorted to only using a 1 Ohm resistor whenever possible. Please note that because of the value of the resistor, wherever it's being used, the read voltage should coincide with the current (I = V / R, so I = V / 1 = V), but because of the resistance of the leads, that's not the case.
Having a fairly beefy capacitor on the output, when unloaded, the shape of the ramp will be a function of current limit and the value of that capacitor. The ramp up appears to be fairly clean when the current limit is set either very low, or very high, however for everything in between, it looks like the constant current mode kicks in and out, generating spikes in the ramp. The set voltage doesn't seem to alter this behavior.
2 V into 1 Ohm
2 V - no load
As I was saying, when there's no load, the ramp-up behavior seems to differ based on the set current limit, because the constant current mode kicks in and out while the output capacitor is being charged.
Constant Current Overshooting
Another thing worth looking into is how it's dealing with loads that will make it start in CC mode. I set the voltage to 5 V, the current limit to 1 A, connected it to the 1 Ohm resistor and the the output on. I found that it starts up fairly clean, with no overshooting.
starting in constant current - 1 A into 1 Ohm. 2 ms/div
Disconnecting the load, while in constant current mode seem to simply make it to ramp up in the usual way, so nothing interesting there.
For this test, I had the supply on, at 2 V in constant voltage mode. Then, I connected and disconnected the 1 Ohm load.
While it takes a bit longer to recover from removing the load, I think it behaves well enough in both cases.
Adjusting the level, while the output is on
Since it makes use of several taps on the transformer, I think it's important to know how it behaves when you're changing taps, while the output is on.
It turns out that there are no gaps and no ripples or anything weird going on when this happens. The ramp up/down in this case, seems to follow the same rules as the turn on and turn off ramps.
I was just playing around with it and I stumbled upon a very weird and dangerous feature. While powering some 12 V light bulbs, I entered "edit mode" and started changing the current using the knob. It was set to 12 V and less than 1 A as I was turning the knob up and down, when out of the sudden it jumped to 30 V @ 3 A and the power went off. I didn't realize what happened so I turned it on again and the light bulbs started shining really bright. I quickly turned it off - and started investigating.
As it turns out, when not doing anything and with M4 active, turning the knob clockwise will make it activate M5. Definitely a feature I would rather do without, especially since the "edit mode" times out, meaning that you might think you're still editing, but by the time you turn the knob, it's out of edit mode.
Considering how little it costs, I believe it's worth the money. Sure, it has some drawbacks, but It's definitely a decent power supply. Very accurate for what it is and with fewer bad surprises in stock than I would have expected.
If you're looking for a cheap, general purpose power supply, I think this little thing is worth considering.