An investigation about ripple current of Kunkin KP184

[tautech]

Note that I live in the city, and there is a lot of RF noise from nearby radio stations, unfortunately, which you can see on my scope.

Turn your 20 MHz BW limiter on for these types of measurements. 

[PaWill68]

Sorry, I haven't been on the forum for months.
Insider information from KUNKIN:
Production of a new hardware version has begun, where the main changes are in two places.
1) Stabilization of the fan supply voltage from LM317.
2) Improved voltage and current measurement accuracy, temperature drift is much smaller.
3) The new version of the MCU is based on ARM-CORTEX-M3 GD32F130C8T6, which greatly increased the performance and protection response.
The new software is not compatible.
I don't have more detailed information.

[burkm]

I resolved some of the problems mentioned by rerouting the fan cables and resoldering some suspicious solder-connections. My mainboard version was V5.
Was actually going to order a 2nd one after this, but I am now waiting for the new hard- and firmware version announced in the post before...

[burkm]

Anyone seen anything of the KP184 version mentionend before ? I checked ebay i.e., but I don't find any mention of a new hardware version or else.

[seronday]

I recently obtained a KP184 which has a main board ver 06,  fitted with an LM317 to regulate the fan supply and a GD32F130C8T6 MCU

Regards.

[burkm]

Whom did You get it from ? Couldn't find any mentioning anywhere about the new version / new hardware ?

[seronday]

I purchased the KP184 from this Ebay seller:-    tooldealers.

https://www.ebay.com.au/itm/263281377583?hash=item3d4ccb012f:g:nqwAAOSw-qpg9R2~

The KP184 has a label on the rear showing a build date of September 2022

Regards.

[electr_peter]

Banggood has started offering KP184 recently again. BG only ships in batches till stock lasts. Before BG sold KP184 model till batch lasted.
Likely BG have the newest version, but you never can be certain for sure.

[JimBeam]

Hi everyone,

I bought a KP184 from eBay and the mainboard is the new Version 06 with the GD32F MCU. It was sold as new but arrived without any accessories - no cables, not even the original package, just wrapped in bubble foil... It was manufactured - according to the sticker - on 2022-09-14.

First tests were OK, I was just wondering why the fan would not spin up after sinking 300W for a minute and the case already becoming hot.

Then I tried to connect it to the PC without any success - and I tried a lot of adapters, cables, USB-to-RS232 converters, etc.

Finally I opened up the device, as I wanted to change the connectors and install a power switch on the front anyways. After some poking around I found that neither the fan nor the MAX232 for the serial port got any power. Looking for the fault I found the LM317T to be defective, between the ADJ and OUT pins where only 60 Ohms. After replacing the LM317T and also relocating the LM7805 hanging under the mainboard to the heatsink just beside the LM317T, I was able to connect the unit to the PC and - bingo - the fan now ran too.

Being euphorized I went ahead and installed the new 2128 firmware, but while the upload was still running my heart stopped beating for a moment, wasn't there something with the MCU being a GD32 instead of the "old" STM32? For a moment I was tempted to stop the upload but didn't as it already had erased the flash and was at about 30% upload, so I let it finish.

After the upload was finished the fan went to hyperspeed, the display flickered but it showed the 2128 version - for about 15s - then the normal screen appered (still flickering) and I could even use the menu, I was able to enter a new current value in CC mode too, but when I pressed SET, the unit froze. 

Now I had a half-bricked device, with fortunately the bootloader still intact. Of course I tried to upload again, in the hope something went wrong the first time, but nope...

Well, that's the current state. So my question to PaWill68: did you already have a source for a firmware for the Ver.06 mainboard? I did not note the version the unit had before I bricked it, but it was something like 2021.

Andreas

[JimBeam]

In reply to my own posting:

I contacted 64908476@qq.com (mentioned in msg#27 of this thread) , who sent me the firmware 
With permission of 64908476 (I don't know his/her real name...) I attach it to this post, for anyone who might need it, it's V2201 for the Ver.06 mainboard with GD32F processor.

And he/she wrote further, that there is already a Ver.07 on the way:

However, the current PCB version is 07, buzzer sound changed to chord mode, optimize fast over-power protection function, and voltage and current temperature drift are further optimized.
In addition, the problem of small current (<100mA) with power frequency ripple (100Hz) in CC/CR/CW mode has been improved (affected by PCB wiring has not been completely solved).

[burkm]

Any further News on the latest versions of KP184 ?

[torch]

I just got one today. Build date 2023/06/29. Main board version 7, dated 2022/11/??. Bottom board is version 6, dated 2022/12/12.

Very little noise compared to earlier reports. A 15uS ripple of 9-10mv at 400Hz.

Other observations: I like the sound. Instead of the standard beeping, it gives a soft chime on button presses -- and has different chimes for up/down, etc. They seem to have fixed the oversize banana plug issue -- mine all fit fine. The displayed values seem to all be in spec - very close when compared to a Fluke 175, a VC8145 and the Korad PS display. The large diameter wires all seemed adequately soldered.

Other noted problems still exist: the transformer mounting bolts were just finger tight so the transformer could wobble. The case screw holes still need to be scraped. The rear panel heat shrink screws are still connected to DC+. There's still no indication if the voltage sense source is internal or external. And while overall the internals look quite neat and professional, I found a blob of spilled solder lightly stuck to the the main board, between the DC+ connection and a little 3.3v regulator.  Mine came with a European power cord (I'm on teh other side of the Atlantic) and did not come with an RS232 cable. Nor did it come with any software.

But overall, I'm pleased. At that price point it's a pretty darn solid piece of kit.

[Everbrave]

I just got one today. Build date 2023/06/29. Main board version 7, dated 2022/11/??. Bottom board is version 6, dated 2022/12/12.

Very little noise compared to earlier reports. A 15uS ripple of 9-10mv at 400Hz.

Other observations: I like the sound. Instead of the standard beeping, it gives a soft chime on button presses -- and has different chimes for up/down, etc. They seem to have fixed the oversize banana plug issue -- mine all fit fine. The displayed values seem to all be in spec - very close when compared to a Fluke 175, a VC8145 and the Korad PS display. The large diameter wires all seemed adequately soldered.

Other noted problems still exist: the transformer mounting bolts were just finger tight so the transformer could wobble. The case screw holes still need to be scraped. The rear panel heat shrink screws are still connected to DC+. There's still no indication if the voltage sense source is internal or external. And while overall the internals look quite neat and professional, I found a blob of spilled solder lightly stuck to the the main board, between the DC+ connection and a little 3.3v regulator.  Mine came with a European power cord (I'm on teh other side of the Atlantic) and did not come with an RS232 cable. Nor did it come with any software.

But overall, I'm pleased. At that price point it's a pretty darn solid piece of kit.

Did you watch the change of the ripple value as the load current is changed?
Do you mind telling where you bought it from in Europe? Price if possible
Firmware version?
MOSFET type number?
Perhaps some photos of the new Board?
Thanks

[Everbrave]

I have a V6 Board and I can, interestingly, report that the ripple value at lower load current is higher than at higher load current!
Here are some pics …
At 100mA and 1.1A respectively (around 42V) …

EDIT: I can imagine that the ripple (noise) is inherent in the current control loop (bandwidth) since at lower currents the S/N ratio is lower.

[torch]

Did you watch the change of the ripple value as the load current is changed?
Do you mind telling where you bought it from in Europe? Price if possible
Firmware version?
MOSFET type number?
Perhaps some photos of the new Board?
Thanks

Sorry I've taken so long to get back to you, I've been too busy using it to take it apart again until today.

Ok, so change in ripple current. Judge for yourself. Power source is Korad KA3305P:

5v 1a


Huh, that's weird. The scope is showing a negative value for Delta-Y. Is that even mathematically possible? 

5v 5a


15v 3a


15v 5a


31v 1a


31v 3a


31v 5a


I'll get to the guts in the next post...

EDIT: I forgot to mention. Strange things happen if I inadvertently set the Kunkin draw to exceed the Korad CC limit! The two seem to fall into some sort of race condition thereafter with each fighting the other over control of the current. Wild voltage swings of well over 100mV can result. One must shut them both off and restart with the Korad set to supply more than the Kunkin is trying to draw.

[torch]

Ok, circuit board porn:

Rear half of the top board


Front half:


Can't really make it out in the pic, but it says IRFP250M on the chip:


Top board revision:


Bottom board revision:


Front half of the bottom board, as near as I can photograph without further disassembly:


Rear half of same:


The rear board is only 1/2 width anyway. Chips bolted direct to the heat sink:


And the little switch board attached to the rear panel. This becomes important in the next post.

[torch]

Like I said, I've been using it. Great addition to the bench, wish I had one years ago. But. That reaching around to the power switch is getting old fast. Time to do something about it:

As others have noted, the front panel is cast to hold something on the bottom right. I don't know if it was for a power switch or a USB port, but I thought this little guy would fit:





As it turns out, 13mm was perfect. 7mm was a hair too wide. I thinned the front panel plastic a hair with a Dremel tool, until it popped in snug and secure. After drilling a hole for the toggle, of course!



I had hoped to interrupt the supply to the transformer. However, they used a centre tap transformer, switched by the 120/140 switch at the rear. As I am on 120v mains, and as there is no way a wider switch would fit in that tiny pocket, I had to go for the point were Line met board. The socket's tab was actually just soldered to the side of that rear board so it was easy to desolder, bend slightly, solder one wire to the tab and the second to the board. Heat shrink all 'round, of course!



And here we go. No more reach-arounds for me!!! 



Oh, and as for your question about price: I'm not in Europe. I bought one off Banggood. About CDN$250 all in after shipping, insurance and taxes.

[Everbrave]

Thanks a lot for the great effort. However, the resolution of the pics is too low to read the MOSFET Typ Nr., do you mind writing it down?
Also, the Firmware version would be interesting.
As for the ripple form and magnitude, it’s quite different from my board V6. I can’t explain without schematics, which I couldn’t find anywhere.

[Hamelec]

see above:
"Can't really make it out in the pic, but it says IRFP250M on the chip:"

[torch]

Thanks a lot for the great effort. However, the resolution of the pics is too low to read the MOSFET Typ Nr., do you mind writing it down?
Also, the Firmware version would be interesting.
As for the ripple form and magnitude, it’s quite different from my board V6. I can’t explain without schematics, which I couldn’t find anywhere.

As noted, the Mosfets claim to be IRFP250M. They look pretty convincing, I don't suspect counterfeits, but who knows these days?

The second line on boot is "2303" I assume that's the firmware version?

[Everbrave]

Thanks a lot for the great effort. However, the resolution of the pics is too low to read the MOSFET Typ Nr., do you mind writing it down?
Also, the Firmware version would be interesting.
As for the ripple form and magnitude, it’s quite different from my board V6. I can’t explain without schematics, which I couldn’t find anywhere.

As noted, the Mosfets claim to be IRFP250M. They look pretty convincing, I don't suspect counterfeits, but who knows these days?

The second line on boot is "2303" I assume that's the firmware version?

Reverting to the older IRFP250M  from the newer IRFP264 (in V6 board) is a bit astonishing. I compared the specs thoroughly but can't think of a reason!
If I would care to have a max. junction temp of 175Deg Celsius, I would rather choose IRFP260N.
I wounder if we, V6-Board owner, can upgrade to FW2303; any ideas?
Thanks

[torch]

\
Reverting to the older IRFP250M  from the newer IRFP264 (in V6 board) is a bit astonishing. I compared the specs thoroughly but can't think of a reason!

It looks like the IRFP250 has faster rise/fall times, so maybe helps control ripple. I'd like to think it's something like that, but honestly, at this price point it probably has more to do with what was on sale that week.

I still don't understand why they used the wrong DB9 connector. Or if male connectors were on sale, why they didn't at least wire it for a null modem cable instead of straight-through?  Mine did not come with the cable, so I had to go hunt down a cable AND a gender changer.

[Hamelec]

RS232 was always weird over the last decades   

[Everbrave]

\
Reverting to the older IRFP250M  from the newer IRFP264 (in V6 board) is a bit astonishing. I compared the specs thoroughly but can't think of a reason!

It looks like the IRFP250 has faster rise/fall times, so maybe helps control ripple. I'd like to think it's something like that, but honestly, at this price point it probably has more to do with what was on sale that week.

I still don't understand why they used the wrong DB9 connector. Or if male connectors were on sale, why they didn't at least wire it for a null modem cable instead of straight-through?  Mine did not come with the cable, so I had to go hunt down a cable AND a gender changer.

IRFP250M has less capacitances in general. This could also be the reason to the high frequency contents (ringing) in the ripple; but again, I am only guessing here.
EDIT: I looked-up the prices: IRFP250N is around 1.5USD, IRFP264 is around 5.3USD

[torch]

I looked-up the prices: IRFP250N is around 1.5USD, IRFP264 is around 5.3USD

And there we go... 

Hey, at least they seem to have sprung for the genuine article.