EEVblog #404 - Korad PSU Followup

[xani]

Does anyone have experience with triple ( KORAD KA3005D-3S 2x(30V 5A)+5V ) version of that power supply ? I

[poundy]

Is the switch-on behaviour of the "new" PSU still showing the same transients and overshoot in the combinations the original test showed?

[uoficowboy]

I recently picked up the Korad KA3005D on Amazon. It seems to work well, except for a couple of quirks:

1. The power button doesn't work all the time. It gets stuck on. I just attempted a rework of the switch and I *think* that has fixed things, but I'm not positive of that. The switch appears to be a Chinese copy of the E-switch P197EESB. So if the problem reoccurs, I'll just replace it with that one.

2. Whatever is driving the display is not super smart. I occasionally use it to charge LiPo batteries. I'll set a current and a voltage (say, 50ma and 4.2V). For most of the time, it'll properly display the voltage and current (ie 50ma and 3.8V). But as it approaches the set voltage (4.2V) it'll switch over to displaying the set voltage, instead of the output voltage. So, this is what I just saw about 5 minutes ago:

50ma 3.8V CC (measured voltage: 3.8V)
50ma 3.9V CC (measured voltage: 3.9V)
50ma 4.0V CC (measured voltage: 4.0V)
50ma 4.2V CC (measured voltage 4.05V)
50ma 4.2V CC (measured voltage: 4.1V)
40ma 4.2V CV (measured voltage: 4.2V)
30ma 4.2V CV (measured voltage: 4.2V)
etc (I'm simplifying here, obviously)

The switchover seemed to be around the 4.01V mark. I even saw the display oscillating there - between 4.01 and 4.2V (where 4.01V was the correct voltage). Anybody else seen this?

3. The current measurement has a small offset to it - maybe 4ma. Anybody know of any clever way to calibrate this thing? I often operate at low currents, so it'd be nice to get rid of that offset!

Overall, this is a great supply for the money IMHO.

[Huluvu]

The switchover seemed to be around the 4.01V mark. I even saw the display oscillating there - between 4.01 and 4.2V (where 4.01V was the correct voltage). Anybody else seen this?

3. The current measurement has a small offset to it - maybe 4ma. Anybody know of any clever way to calibrate this thing? I often operate at low currents, so it'd be nice to get rid of that offset!

I never ever would rely on the displayed values shown by an Power Supply if I need really accurate readings.
Most power supply's are not calibrate able and the confidence level is very low.

[uoficowboy]

The switchover seemed to be around the 4.01V mark. I even saw the display oscillating there - between 4.01 and 4.2V (where 4.01V was the correct voltage). Anybody else seen this?

3. The current measurement has a small offset to it - maybe 4ma. Anybody know of any clever way to calibrate this thing? I often operate at low currents, so it'd be nice to get rid of that offset!

I never ever would rely on the displayed values shown by an Power Supply if I need really accurate readings.
Most power supply's are not calibrate able and the confidence level is very low.

My main concern is not that the displayed voltage is inaccurate - rather that the displayed voltage is the set voltage, not the output voltage, when in current limiting mode and your output voltage is close to the set voltage. This is almost for sure a software bug.

On a side note - I replaced the power switch and the device turns on and off reliably. A colleague later showed me a way of tweaking the contact inside the button to make it springier so if anybody else has this problem the original switch may still be salvageable.

[Rasz]

My main concern is not that the displayed voltage is inaccurate - rather that the displayed voltage is the set voltage, not the output voltage, when in current limiting mode and your output voltage is close to the set voltage. This is almost for sure a software bug.

Mmm sounds familiar. This is most probably not a bug, but a way chinese do it.
I have cheap Chinese knockoff RKC REX C-100 ($14) Controller and its exactly the same, as soon as it approaches SET temperature it starts to lie in your face about its super duper stability/accuracy. It almost makes me think its an Asian thing (mm sweet racial discrimination). Those products lie just like Japanese Businessman/Government official would - shaking your hand, nodding his head and lying straight in your face because its more honorable to lie than to admit defeat.

[MBY]

The switchover seemed to be around the 4.01V mark. I even saw the display oscillating there - between 4.01 and 4.2V (where 4.01V was the correct voltage). Anybody else seen this?

3. The current measurement has a small offset to it - maybe 4ma. Anybody know of any clever way to calibrate this thing? I often operate at low currents, so it'd be nice to get rid of that offset!

I never ever would rely on the displayed values shown by an Power Supply if I need really accurate readings.
Most power supply's are not calibrate able and the confidence level is very low.

My main concern is not that the displayed voltage is inaccurate - rather that the displayed voltage is the set voltage, not the output voltage, when in current limiting mode and your output voltage is close to the set voltage. This is almost for sure a software bug.

On a side note - I replaced the power switch and the device turns on and off reliably. A colleague later showed me a way of tweaking the contact inside the button to make it springier so if anybody else has this problem the original switch may still be salvageable.

Thats a serious issue and a complete show-stopper! Can more people please try to confirm this?

[Mdbelen]

Hi, i can still confirm this issue with my unit (KA3005P), just recently ordered from hotair.pl. (Btw they sadly put stickers on some case screws so I can't open it without losing warranty.)
I tried with 1 and 3S 1W LEDs = 3 and 10V and for both there was a gap of 200mV. This is a >5% error for low voltages. Pretty damn much for a claimed accuracy of 0.5%+20mV! What a retarded software """feature""".

To summarize misbehavior in few a words:
If in CC-mode V_real approaches V_set to (V_real = V_set - 0,2V) the display switches from displaying V_real to display V_set.
This does not happen for the displayed current in CV-mode though.

Another notsonice fact i discovered:
OVP does not trigger reliably for me: If OVP is enabled and output is open looped it does not trigger (though ofc V_real=V_set). Also if I slowly increase the current through a load until V_real reaches V_set it does not trigger, no matter how much further I increase C_set. It does only trigger if I increase the C_real pretty fast, thus "shoot" in the V_set limit pretty fast.
Don't know how much this matters in practical usage, as OVP won't be used often anyway I guess.

3S=3 in series | CC=constant current | CV=constant voltage | C=current | V=voltage