Korg Forums

Hmm. Sounds like the wire that the sense wire was connected to isn't designated as 0V. So, when the sense wire is shorted, the wire it's shorted to comes off being at 0V, to whatever supply voltage that wire is supposed to carry, and so the cycle repeats.

.

Since you have the Kronos open, you may want to connect a monitor to the motherboard. Or did you already do this? And could you confirm you have +5V SB running to the Boot Support board and this board is seated properly? I recall having some issues with this board but not sure exactly what it was, think it was that this board didn't fully seat onto the motherboard front panel header pins or something.

It's frustrating since I used the original pin header and simply soldered the corresponding cables onto it. Nothing 'should' have gone wrong...

Maybe the soldering is not solid after all...
Measuring the various voltages while the PSU is pulsating, all of them go up and down, from e.g. just under 3V back to 0V, and up again.

I still have the gut-feeling that the no-load protection (NLP) kicks in for some reason.
Need to find/create a proper dummy load, probably on a few cable connectors.

So, whoever is planning to replace the PSU I would definitely get one that has no NLP, just to rule out this possible issue.

EDIT: could it be that the max. amps are too low on this new PSU?
new one: +5 Vsb < 2A / original ENO = 3.5 A
-12V < 0.3A / ENO = 0.5 A

and that is why it is acting as it is?
I might unsolder everything again, try to measure the new PSU, if it is working at all...

EDIT: could a Cmos battery low voltage be the culprit?

PSU voltages should never be pulsating. One issue I came across was that when I pushed in the reconfigured connector into the PSU, some of the pin connectors got pushed out of the connector and didn't make contact with some of the PSU header pins. This could be happening to you and the reason for the off then on voltages. I'd inspect the each pin connector to ensure none got pushed out while you were seating the connector to the PSU.

I only have fixed cables coming out of the PSU. I took off their pins and soldered them to the existing pin connector from the Kronos.
So this is the only source of errors. But I just checked the soldering of each pin for contact/proper soldering, and all are OK, no open connections (in German "cold soldering").

I wanted to avoid opening all soldered cables again, and check the new PSU for functionality, without anything connected to it... Might be another long evening it seems...

One orange cable from the PSU which I didn't use is one that is connected to a brown one. Brown is maybe "remote sense" 3.3V. I didn't specifically look out for this but could it be that I should be using this one?

I'll drop a picture into my shared folder.

To be safe and given you updated the Motherboard connector, going forward, it would make sense to first try the PSU with the updated connector against a totally different motherboard to troubleshoot, hearing about fluctuating voltages and unconnected wires raises concerns and hate to have you blow up your Kronos motherboard.

Also, for the remote sensing wire, this site notes this:

“Classic" ATX Power Supply
The 3.3V sense wire needs to be connected to +3.3V. This is a brown or orange wire in pin 11 of the connector. It should be connected to 3.3V (orange).

and goes on to state that it should be connected to pin 11

Source: https://reprap.org/wiki/PC_Power_Supply ... 20(orange)

Thanks folks for helping so far.

When I disconnect the 4 pin plug (purple, green, grey, black) and only bridge green with black coming from the PSU, the Kronos starts up about 20% of the red bar and gives me a system error (can't read it since the keyboard is upside down).

When connecting any of these coming from Kronos it goes back to not working and pulsing. Hm..

The error you are seeing on the Kronos should be related to the "Gray" wire not being connected. This gray wire is the, "PWR-OK" line and should register +5V indicating the power supply's voltage checking logic reported a good working state. The reason I had updated my Kronos' PSU was because this 'PWR-OK' wasn't registering +V5 and the issue I was seeing is what you described during boot. Once I had a new PSU with PWR-OK at +5V, it booted up correctly. So I sense we are getting closer. Try ensuring the gray wire is connected. and go from there.

Am turning In a circle. The not-at-all-starting issue always occurs when purple is connected.
When green and black are bridged and gray is connected, it starts half way up as written before.
Am close to connecting the old PSU again... Also to see if the Kronos is still working at all...

Seeing how Purple is the "+5V SB", that leads me to believe that some issue with the Boot Support board and there are two different versions of that board, could you please take a picture of that board and post it? It's that board that connects to the motherboard header pins, think it is noted as KLM-3004 or KLM-3008, it seems one version has the +5V SB going to that board and another version doesn't.

It could be that during the boot up, Kronos is detecting some of its components are not powered on or receiving the correct voltages. Using the original PSU to confirm expected voltages and that it boots still would be a great test.

It's the KLM 3008.

Edit: and it's booting up OK back to the original PSU.

With the original PSU following voltages I can measure:

Purple: when switched on = 5v and stays so
Green: when switched on = 5V for 1 sec, then 0V and stays 0V
Gray: opposite to green, switch on = 0V, after 1 sec 5V and stays

Seems like normal behaviour.

I took the other PSU off so I can't figure out what is happening to those voltages there. It seems that the new one doesn't get beyond the 1 sec step.

Well, the good news is that your Kronos is still functional.

I take it that when the new PSU was taken out that you manually turned it on (green to black) and measured each pin for correct/expected voltages?

Yes, I did this. All voltages are in range. Purple is 5V, green too, and gray is 0V

All continuously, no 1 sec switching.