Seasonic II+ clone PSU negative voltages problem.

4 posts / 0 new
Last post
Offline
Last seen: 1 week 6 days ago
Joined: Oct 8 2021 - 02:14
Posts: 3
Seasonic II+ clone PSU negative voltages problem.

Hi. Yes, this is my first post. I mostly joined because searches for various things I want to know tend to bring me back to here.

I've started a project to bring as many of my old machines back and make them healthy as possible. Oddly, most of my ailing ones are Apple.

I've been trying to tackle the Seasonic PSU that came with a categorically messed up II+ clone I picked up over 20 years ago. Long story short. The Motherboard works again. The keyboard won't work until I remove the chips and give it a new brain. The floppy controller needs a repair, along with one of the floppy drives. Not sure about the rest.

The PSU may have been what wrought the havoc in the first place. I vaguely remember it being super unstable before it died.

 

Skip to present. I've pulled and checked all the caps. Some of the resistors and diodes. Repaired a couple of scratched traces. It works. Or at least the +5 and +12v rails do. It can happily power my dummy load of two WD Caviar drives. The -5 and -12 are sad. -12 is around -4.x volts, and -5 is +2.x volts. My gut is telling me something about a ground reference but apparently I'm not smart enough to understand it.

In the attached photos, there is a solder bridge between one of the power transistor bolts and a track. There was a gouge there. It was the deepest part and there was some minor damage to a nearby track. After asking elsewhere online it was mentioned it could have just been a sloppy factory mod. So I removed the bridge again before testing it. Also I have no idea what the deal is with the resistor on the underside of the board.

I flipped the PCB side and aligned the top and bottom photos as best as I could for position and perspective.

Please, ask away if more info is needed. There's really no reason for me to use this PSU instead of replacing it with an ATX beyond me wanting to. Same as the rest of the clone really. I just want it to work again.

 

edit: I have absolutely no idea what I'm doing with the photos. Sorry!

 

Offline
Last seen: 9 hours 13 min ago
Joined: Feb 27 2021 - 18:59
Posts: 73
Peasy

This is a single-layer circuit board (the copper traces are all a single layer of wiring) and tracing out the schematic is quite easy. You should try it for yourself. Just remember the cardinal rules for schematics:  power on top, ground on the bottom, input on the left and output on the right. Paper and pencil is all you need to do this.

Some notes: the two bolts in the middle are connected together through the TO-3 transistor's case. The beige 6-pin chip is an optoisolator—an electrically separate LED and photodiode in the same case. It provides feedback from the regulated rail (+5V ?) on the secondary side, back to the primary side chopper transistor. The green and black wires wound on a ferrite doughnut are a common-mode choke: a radio-frequency interference (RFI) suppression filter that cleans up the mains (120-240VAC) power before it is used. The three 3-pin chips with heat sinks are linear regulators, each with reverse protection diodes attached.

Offline
Last seen: 1 week 6 days ago
Joined: Oct 8 2021 - 02:14
Posts: 3
robespierre wrote:This is a
robespierre wrote:

This is a single-layer circuit board (the copper traces are all a single layer of wiring) and tracing out the schematic is quite easy. You should try it for yourself. Just remember the cardinal rules for schematics:  power on top, ground on the bottom, input on the left and output on the right. Paper and pencil is all you need to do this.

Been meaning to do that properly for this. Recently I invested most of my time into collating my notes and drawings from the II+ clone keyboard into a schematic on the computer. This should be way easier. But at the same time, while I know what the components are on the PSU, it's an area of electronics I'm not knowledgeable in. For example, I couldn't say with certainty why the -12v rail has two low value resistors in series connected to ground. My gut tells me it's for stability under low load but I don't actually know. I also still have absolutely no idea whether the severed trace near one of the transistor's bolts was meant to be. 

 

3 pin chips? Do you mean the TO-220 near the bottom right in the pictures? I think it was labelled SCR1. Assuming it is an SCR, does anyone have any hints on a possible part type? It seems to be unlabelled.

 

Really in the end, I can identify the parts etc but the way they do what they do is still a bit mysterious to me. I'm not going to go licking a power diode or anything. It's just how these things function I'm not great at understanding. For example where theres some resistors and a diode in parallel. I can see that'd let through more current in one direction than the other but I have no idea why this is being done. Clearly I need to educate myself more on switch mode power supplies. But I'd also love some help understanding why I'm seeing the issues that I am on this PSU. To me it feels like the ground plane for the negative rails is floating or seeing leakage from the positive side. But it's a guess and nothing more.

 

 

Offline
Last seen: 1 week 6 days ago
Joined: Oct 8 2021 - 02:14
Posts: 3
I "fixed" it!

I'm not sure which action got all the rails working properly, but besides a lot of probing of components I did two other things.

 

The mains power connector and switch wires were terrible so I decided to replace them. At this time I noticed the switch was connected to neutral so I corrected that too. I chose live to be connected to the AC side of the circuit which has the cut jumper wire parallel to a big ceramic resistor. Presumably the 120 / 240v selection.

 

After doing that and noticing I'd failed to thread the ground through the socket hole, I decided to reflow everything while I was fixing my routing error. Tested it after and everything was good. Texted it again today. Still good. I installed it back in the clone. Yet to test that. Can't do much though because it's keyboard is fried and also made of distilled unicorn horn oil. So it's going to get a microcontroller chop job. Not a choice I take lightly. I have spent countless hours with a multimeter and scope trying to work out what all the sanded chips are and whether they work. In the end I have three unidentified, two of which seem to be cooked. Two positively identified and one tentative. The ROM is full of garbage so I think it is dead too.

Then I need to fix the floppy controller and drives. The mainboard is already done. I'm bringing this thing back kicking and screaming from the electronic afterlife. 

Log in or register to post comments