Power source KPS3005D

[elebot]

I have recently bought this Chinese power source on eBay and thought it would be useful to share some info.

It is a fairly standard 150W switch mode power supply providing 30V / 5A (actually goes a little bit higher 32V / 5.1A), with "digital control".
In Europe it is being sold for about 40 Euros on eBay delivered and for example in Germany they have local stock so no problems with customs and quick delivery (in my case anyway).

There are several other models in this line offering higher power and/or serial interface. Unfortunately by omitting the serial interface they introduced issue of missing grounding which leads to a somewhat erratic behaviour straight out of the box...
This is (together with the fix) documented on these videos (not mine):



Below are some pictures of the internals and complete schematic. As you can see almost half of the box is empty :-)

Looks like standard PC power supply based on TL494.

There is no PFC (in Europe from 80W should be).

There is no real output power switch. The CPU just disables/enables the control IC. Which together with the big filtering capacitors leads to "soft start" and stop (depending on the set current limit).
Personally I would prefer to internally let it run all the time and use for MOSFET to switch the output (like for exmaple in Manson NSP3630).
But other similar PSUs (e.g. Manson SSP-7080) does it the same way so I guess why not.

Control/ergonomoics could be better (in general the FW could be better). For example to set the current limit you have to press "A/V" twice, set the value with the encoder and push enter.

The meters have resolution of 10mV and 1mA, the same applies for setting steps. The refresh rate is fast (which I like :-)).

There is a thermally (NTC, controlled by the CPU) controlled fan inside. Its hysteresis seems 45C on and 40C off. I would expect/prefer higher and bigger difference - this way the fan can go frequently on and off.
Interestingly enough the schema seems to be ready for contious FAN control using PWM (even galvanically isolated) but in my case it is just being switched on (60% duty cycle from 12V) and off.

The CV/CC control is standard (analog control loop, CPU just provides the control voltage using PWM). But the CV/CC indication (CV/CC LED control on the front panel) is purelly done in SW by the CPU.

The whole power supply is controlled by STM8 micro (12b ADC/DACc, 16b PWM). Internally there is a programming header and the FW is not locked.
Attached is also the FW I was able to read from it (would be glad if anybody could decompile or at least disassemble).