VOGONS


First post, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

Hello, some hours ago I put back in place this mb (https://4.bp.blogspot.com/-HPRP48X6cos/V2O8bq … Corp%2BP55C.jpg), but despite paying attention (clearly not enough), I connected the mb with one pin left on the right side, then I switched on the psu, it went to protection immediately, then I switched it off after some moments, but now when I power on the system I just get a blank screen with no beeps, on the debug card I just get 0000 code, with or withtout cpu and ram, I noticed that the cpu gets warm anyway.

I tried another cpu, another bank of sdram and another video card, same result...

Did I kill the system for good?

Reply 1 of 14, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie
Nemo1985 wrote on 2020-04-29, 23:21:

Did I kill the system for good?

I'm not sure what you mean by "I connected the mb with one pin left on the right side" - which one do you consider right and left? P8 or P9? My guess would be though you've plugged P8 one pin off, somehow, and that put +12V on "key" position. This is usually a missing pin on XT systems to prevent such mistakes but on AT it's another +5V. So the mobo got a surge of 12V on 5V line before the PSU noticed the outputs being wrong and shut down.

So yeah, I'd say good chances are it's toast.

EDIT: Oh and BTW if your memory was SIMMs or 5V DIMM stick then that could be toast too, as well as CPU and any ISA cards. Best to test everything in a known working mobo.

Reply 2 of 14, by shock__

User metadata
Rank Oldbie
Rank
Oldbie

http://www.minuszerodegrees.net/5150_5160/MDC … 8_p9_pinout.jpg
You can guess best what you did there.
If you're lucky you just killed the +5V/-5V rails on your PSU ... did you check the voltages?

Yup, you can be pretty sure you killed either the PSU or mainboard for good. The Debug card being stuck at 0000 also indicates the CPU won't initialize (which can be due to a number of reasons).

Last edited by shock__ on 2020-04-29, 23:48. Edited 1 time in total.

Current Project: new GUS PnP compatible soundcard

[Z?]

Reply 3 of 14, by shock__

User metadata
Rank Oldbie
Rank
Oldbie
Deunan wrote on 2020-04-29, 23:37:

EDIT: Oh and BTW if your memory was SIMMs or 5V DIMM stick then that could be toast too, as well as CPU and any ISA cards. Best to test everything in a known working mobo.

And potentially kill the working config along? Let's resort to finding out what happened first, before killing more good hardware 😉

Current Project: new GUS PnP compatible soundcard

[Z?]

Reply 4 of 14, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie

Eh, if nothing is smoking now when power is on, there can't be any serious shorts. It should be safe enough to try and swap the parts one by one. ISA cards and RAMs can be tested in any cheap mobo that at least boots to BIOS.

Reply 5 of 14, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

Psu is working fine, memory is working fine, video card is working fine, tested on another mb. Since I can't properly explain what I did I upload a picture, this was the mistake: https://i.imgur.com/OfyDWOs.jpg
As I said the psu immediately cut off the power as soon it detected something was wrong.
I tried to program and change the bios eeprom, I tried every single pci slot and an ISA SVGA, always 0000

Thanks for the help

Reply 7 of 14, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie

Hm. That would put +12V on -12V pin, short -12V to ground, and short -5V to 5V.
Now, -5V has pitiful current capability so that shouldn't do any damage, unless to the PSU. Make sure your PSU still delivers clean -12V and -5V. Start with -12V as -5V is often just a 7905 regulator on that line.
I wonder if your mobo needs -12V for something important and putting +12V there killed it. By the way, the test card, what does the RESET LED on it do? Does it react to pressing the reset switch? Did you test that PSU with another mobo, maybe you've busted the PWR_GOOD driver transistor by shorting it to 5V.

Reply 8 of 14, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

I can't test the cpu on another board but I tested another cpu (same as the one used on that mb) but same results 0000 and the cpu gets warm almost immediately.
I switched debug card, the second says: ----
Tester result:
+5V: 4,91
-5V: -4,67
+12V: 12,01
-12V: -11,67
Debug card says:
-12V On
+12V On
+5V On
+3.3V Off
CLK On
IRDY, FRAME, RESET Off

Reply 10 of 14, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie
Deunan wrote on 2020-04-30, 08:30:
Nemo1985 wrote on 2020-04-30, 00:19:

IRDY, FRAME, RESET Off

RESET should be ON briefly after you apply the power, then go off. Same when you press the reset switch. Make sure it works like that.

Yes it's like that, half second when I turn the system on, but what shouldn't stay on is the clk led I suppose

Reply 11 of 14, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie
Nemo1985 wrote on 2020-04-30, 09:06:

Yes it's like that, half second when I turn the system on, but what shouldn't stay on is the clk led I suppose

It's probably blinking but at multi-MHz frequency so you have no way of telling. That LED is not very useful, if you have a scope check the clock presence with that.

Reply 13 of 14, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

I did some other testing:
With the ram and without the cpu the debug card shows CLK and RESET ON, while with all the other trying the CLK is always on (or blinking?) and reset off after the first half second, after half minute it start to blink on and off, with all the other setting the CLK is always on while other leds are off.
Since the mb has the RTC socketed (VIA vt82885n), I tried to take it off, I sometimes some burning smell around the RTC area, (where also the mosfets are), I could buy some new rtc or would it be a waste of money?
I suppose something has been damaged to the cpu line but it still get warm if I keep the system on for like a minute...