LisaList2

Quote from: berskyboy on September 09, 2026, 11:38:35 PMThis is a picture of my video/picture of my setup

The PSU won't "turn on" unless pins 20 and "X" are connected to each other. In my photo above, they are connected via the little toggle switch on the right (note the orange text) so one can turn it on and off like the I/O Board does (aka "Soft Power") using the switch.

If you purchased the switch (or have one), wire it to pins 20 and "X". If you don't have a switch, connect the two pins with a wire and the PSU will turn on as soon as you plug it in.

Having the on/off switch will make it easier to troubleshoot logic problems on the daughterboard, as that allows you to connect the voltmeter to various points and observe what happens the moment the PSU "turns on" (and then decides to turn itself off).

(In this case, "turn on" is not the same as connected to the mains/110V via the power strip)

(I don't think you need to order any more parts, my guess is that the broken solder joint on the potentiometer was the problem, and once you get the test setup going, the rest will be quick... could be wrong though.)

edit: The pictures in your last post don't all show the same wiring, which makes it hard to point out what you need to change (since it is different for each picture). In summary, for the PSU to turn on (so you get +5V and +12V), you need to have all of these connections at the card edge at the same time:

• Connect a load from +5 to ground
• Connect a load from +12 to ground
• Connect +5 Sense to +5V
• Connect the "On" signal Pin "X" to +5Standby Pin 20 (perhaps through the little switch)

If any of those are missing, the PSU won't turn on (except momentarily perhaps).

edit: Your wires are not too small for a quick test... when they are too small and you run it too long, they will get hot and perhaps the insulation will start to melt or smoke. If you are concerned, you can double them up or use bigger wire.

It is very unlikely the new parts are bad if purchased from a reputable supplier such as DigiKey, Mouser, Newark/Element14. Once you get the PSU turned on, if there is a problem with a capacitor, we can diagnose that issue.

Quote from: sigma7 on September 09, 2026, 11:52:27 PMedit: The pictures in your last post don't all show the same wiring, which makes it hard to point out what you need to change (since it is different for each picture). In summary, for the PSU to turn on (so you get +5V and +12V), you need to have all of these connections at the card edge at the same time:

• Connect a load from +5 to ground
• Connect a load from +12 to ground
• Connect +5 Sense to +5V
• Connect the "On" signal Pin "X" to +5Standby Pin 20 (perhaps through the little switch)

[/list]

Morning,  so I put the switch onto pin 20 (before I get my protype board today) and it showed 9V or something like that, and it stayed on.  I turned it off still showing about 9V.  So it appears to stay on.  But as time s goes on it goes down.  Then measure 5V, then 4V.  But doesn't drop immediately.  I was playing with the potentiometer (both of them on the daughter board).  So it seems from a cold start, rest overnight, it has power, then less and less as you test?  I will test it again in couple of hours to see. 

Here's the photos of my edge connector

https://www.dropbox.com/scl/fi/q1xcrm974er25hmfr8kb9/PSU-2026Sep1001.jpeg?rlkey=5kbp30viejm71ktbfbtx944eu&st=mpp7fi9t&dl=0

https://www.dropbox.com/scl/fi/kt1z55x3sf6nd0kgp8l1n/PSU-2026Sep1002.jpeg?rlkey=zarq5ak8fhy521mmi7v4ut9qs&st=qo788y97&dl=0

thanks for all your help troubleshooting!

UPDATE:  I got it working after 2 hours and it stayed at 11v on the 5v rail, nothing on the 12V rail.  Then a flash by the daughter board, and a blowed fuse.

https://www.dropbox.com/scl/fi/i2vnj5lw3wmkqbwxt4vgg/Fuse01.jpeg?rlkey=jo950szucveuo72xsbre0xhfq&st=13h71jph&dl=0
https://www.dropbox.com/scl/fi/nydhs975oo1m3i1gwfvtf/Fuse02.jpeg?rlkey=pwuehcrdv4ex6h158o0qhpjrx&st=ni6eyth7&dl=0

Video

https://youtube.com/shorts/66k2mwcBKhw

I'm off to get a 3A 250V fuse.  But will check for shorts, maybe do some soldering touchups before trying to turn it on again.  (And I will assemble my prototype board)

Quote from: berskyboy on Yesterday at 09:30:00 AMHere's the photos of my edge connector

UPDATE:  I got it working after 2 hours and it stayed at 11v on the 5v rail, nothing on the 12V rail.  Then a flash by the daughter board, and a blowed fuse.

Hopefully the flash was from the fuse. If you replace the fuse without locating the reason it blew, then it will just blow again, so don't apply power until the reason is found and corrected.

The 5V output should never be above ~ 5.5V as that is a fault that could damage much of the computer (if the PSU had been installed in one). If you see that happen again, remove power immediately.

In the picture, it isn't clear if the wire attached to pin 12 is actually touching/soldered to the pin "N" below it. Rather than just fixing it, check the resistance between the pins with the DMM. If the resistance is high, then that confirms they aren't properly connected and that is likely the cause of the excessively high 5V and possibly blowing the fuse. If it is low resistance then you need to look for another fault...

Do not connect to power until a fault is found and corrected otherwise more damage may result.

Examine the board carefully, looking for signs that a capacitor has physically deformed, or signs of burns/scorch marks on the circuit board, resistors, and other components. Signs of overheating include browned or charred components, and you may be able to localize such damage by odour as well as appearance. Your recent pictures may be helpful to identify changes. If you find a damaged component, assume that is another consequence of the fault, not the fault itself. Locating damaged components will help determine where the fault is.

Check the bottom of the board too. See if the position of the board may have shifted, causing a short circuit to whatever is beside or under it. (In an earlier picture, part of the metal enclosure looked like it was partially under the circuit board without insulation or constraint, but one can't easily tell in a 2D picture if they are too close.)

When the PSU is operating, even briefly, the filter capacitors are charged and once power is removed they slowly discharge. This is why the voltage you measured was decaying over time. The voltage across capacitors that have bleed resistors (such as R33 & R2) will decay more rapidly, but most capacitors do not have bleed resistors.

edit: I now realize that the slow decay in the high voltage measured at the 5V output indicates the load resistor was not attached to it. This probably confirms that pin 12 was not electrically connected to pin "N" even though they look very close in the picture. ie. the load resistor was connected to +5Sense but they were not connected to +5. You should still look for signs of other faults.

Quote from: sigma7 on Yesterday at 02:56:31 PMHopefully the flash was from the fuse. If you replace the fuse without locating the reason it blew, then it will just blow again, so don't apply power until the reason is found and corrected.

The 5V output should never be above ~ 5.5V as that is a fault that could damage much of the computer (if the PSU had been installed in one). If you see that happen again, remove power immediately.

In the picture, it isn't clear if the wire attached to pin 12 is actually touching/soldered to the pin "N" below it. Rather than just fixing it, check the resistance between the pins with the DMM. If the resistance is high, then that confirms they aren't properly connected and that is likely the cause of the excessively high 5V and possibly blowing the fuse. If it is low resistance then you need to look for another fault...

Do not connect to power until a fault is found and corrected otherwise more damage may result.

Unfortunately I think I killed it.  Here's my nice edge connector board.  . But There is no life at all.  So likely I damaged the daughter board?  The fuse looks good.  3A 250V.

https://www.dropbox.com/scl/fi/pxtk578se98ln0w3dnlas/Dead01.jpeg?rlkey=0ber935eie7uetr3dmaee4av3&st=de62r74c&dl=0

https://www.dropbox.com/scl/fi/0p9xnjmjm7wji766sm9zj/Dead02.jpeg?rlkey=j1x75dld2ilzkn142skkxicot&st=rfhgipxc&dl=0

https://www.dropbox.com/scl/fi/t8djuejbej3v4dcgm6n00/Dead03.jpeg?rlkey=cdfekgjprfsho4krflpd6rsro&st=e6ihrnlo&dl=0

The only thing left would be to re-flow the components on the bottom of the PSU and daughter board?  Flash looked like it came from behind the daughter board. 

Quote from: berskyboy on Today at 12:46:29 AM

Quote from: sigma7 on Yesterday at 02:56:31 PMDo not connect to power until a fault is found and corrected otherwise more damage may result.

There is no life at all.  So likely I damaged the daughter board?  The fuse looks good.  3A 250V.

The only thing left would be to re-flow the components on the bottom of the PSU and daughter board?  Flash looked like it came from behind the daughter board. 

There should be no discernible activity in the PSU without a fuse installed.

The components on the daughterboard don't handle significant amounts of power, so a flash is more likely to come from the main board. Still, there could be faults on both boards, especially since the original fault may exist as well as a more recent one.

Good that you've got a fresh fuse to install once the boards have been checked carefully and the card edge jig wiring fixed.

I suggest you post pictures of both sides of all 3 boards from enough angles that all the components can be visually examined.

Reflowing a board is only helpful when bad solder joints are suspected. This particular PSU doesn't have a reputation of developing bad solder joints. In addition, reflowing solder joints may hide evidence of where the failure(s) occurred, thus making it more work to troubleshoot. It would be best to minimize any changes while the current condition is assessed.

A lot of energy is available from the power cord, so proceeding to test before confirming things are good-to-go has the potential to cause much more damage compared to working on a low energy circuit.

So my suggestion is:

Provide pictures, then wait for further suggestions on what to check, report on the results of those checks, then wait for further suggestions on what to check and repeat. (Or when the time comes, explicit confirmation that it is time to briefly apply power). As you've no doubt noticed, doing this sort of non-interactive troubleshooting takes more patience than ordinarily expected.