Grem Five wrote on 2021-10-12, 02:20:

My question is why does an AGP slot have a +12 line that isnt even used on the cards I have looked at (just a copper pad that goes no where) and what is that surface mount component circled in orange and where does that trace go to besides the AGP slot? I look for that hole at the end of the trace on the other side of the board and it goes no where from there.... is there a trace inside the board that is tapped off of it somewhere? And if it is tapped off of it somewhere inside the layered board where does it go and could something on the other end of it had caused my short?

Designers are allowed to draw 1A through the 12V pin, so that's another 12W if they need the power (they can regulate it down to whatever voltage they actually need). I found a picture of an MX200 that looks like it uses the 12V for something (or at least has a track going from A1, if I've got my side A and B right): http://www.vgamuseum.info/images/palcal/nvidi … 0_bottom_hq.jpg

The component you've circled looks like a capacitor, likely should be between 12V and GND. Check the resistance across it to make sure it wasn't the cause of the short. Also check the resistance on the graphics card between A1 and GND (should be able to use the back plate), and A2 and GND. I'd be surprised if the A2 pin managed to contact the A1 pin, it looks like the spacing is ok, probably >0.5mm from the A2 pin to the edge of the A1 copper.

-

snufkin wrote on 2021-10-12, 11:25:

Designers are allowed to draw 1A through the 12V pin, so that's another 12W if they need the power (they can regulate it down to whatever voltage they actually need). I found a picture of an MX200 that looks like it uses the 12V for something (or at least has a track going from A1, if I've got my side A and B right): http://www.vgamuseum.info/images/palcal/nvidi … 0_bottom_hq.jpg

In the 3rd pic I posted the A1 pad is the one farthest to the right and you can see on that card it goes nowhere, the pic you linked it does indeed have a trace linked to it. The card that was in the machine looks like an OEM later version of it.

snufkin wrote on 2021-10-12, 11:25:

The component you've circled looks like a capacitor, likely should be between 12V and GND. Check the resistance across it to make sure it wasn't the cause of the short. Also check the resistance on the graphics card between A1 and GND (should be able to use the back plate), and A2 and GND. I'd be surprised if the A2 pin managed to contact the A1 pin, it looks like the spacing is ok, probably >0.5mm from the A2 pin to the edge of the A1 copper.

I thought I checked that side of the smd for connection to ground but I guess not. The left side (towards the rear I/O) of it is ground and the right side is +12v (well would be if the trace didnt burn) Resistance across it is 3.76 M ohms

On the card A1 to GND is open and from A2 to GND is 0.7 ohms which is what you would expect for the card to work in 1.5 volt. I can only guess if the card was in crooked or something and both A1 & A2 pins were touching A1 pad. I dont know I have taken out and put in many AGP cards it other motherboards and never seen this happen..... if IF that is what happened.

I took a look at more of my cards and I see my Geforce 3 based cards have a trace running from A1 to a smd that is connected to the ground plane of the card and they measure about 3.0 M ohms across them so I guess they are doing the same thing the mother board is with that +12 volt line, other cards had a trace coming off A1 and ending up at an empty pad with an empty ground plane pad straight across from it.

Thank you so much for your help understanding whats going on.... I never looked into the AGP slot before and had no idea it had a +12 volt line into it. Unless I find a card that needs that +12 volt I guess there is no hurry to fix that trace or put in a bodge wire.

Some (if not most) AGP cards power the fan from that line. I wouldn't leave it like that.

snufkin wrote on 2021-10-12, 11:25:

I'd be surprised if the A2 pin managed to contact the A1 pin, it looks like the spacing is ok, probably >0.5mm from the A2 pin to the edge of the A1 copper.

Well I tried to test this by probing ground and then the +12 volt trace that burnt and pulling out and reseating the card and trying to move it around in the slot as much as possible and not once did that trace go to ground so I would have to agree with you which leads me totally baffled why that trace burned in the 1st place.

That makes me more nervous for fixing the trace incase it happens again inside a buttoned up case where I cant smell it. At least if I could track down what caused it I would feel much better but unless something dawns on me why it might get set aside again.

Could you post a picture of the front and back of your motherboard without anything fitted (try to avoid any reflections from direct lighting if you can)? It might be worth checking the +12V connections for any other slots, that trace burning might have disconnected them all. There might be something else on that trace that caused the short. It might just be the lighting in your first photo, but it does look like there's a corrosion problem with the vias on the back of the motherboard, they all look a bit green/blue copper oxide tinged. There might some bubbling on the surface just to the right of the AGP mounting post. The motherboard probably has a couple of internal routeing/copper layers, so the bubbling could be a sign of internal heating.

Best Pics I was able to do

I can't see anything else that jumps out as being odd. Maybe that capacitor you circled in the first post failed short and took out the trace with it. Seems unlikely. I'm still a bit suspicious of the blueish tinge to all the vias on the back. At this point I'd probably solder a wire to replace the burnt trace (maybe connect a 1A fuse as well?), then check the resistance between 12V and all the other voltages to make sure there's nothing short. Then try turning it on with no cpu or RAM and measure all the voltages on the AGP and PCI slots.

its a pity some agp cards burn away but have incompatable voltage rails.

Well I was just trying to figure out if this board was worth being in my collection. I dont really have any nostalgia for P4 machines as after P2 & P3 I moved on to Athlons and never had a P4.

Its a 1.6 GHz processor that I thought might fit in nice but I was hoping it was a socket 423 but it was a socket 478 instead. I also actually like the case and thought it would be nice to keep a gateway system in its original case but I do have 2 other P3 MATX boards I could put in there. Plus now that I'm living in old Gateway country I thought it would be nice to have one system from them which is sad as all I can find around here is old Dell E-waste.... heck I had enough of that when I lived in Austin,TX and you couldnt swing a dead cat without hitting a mountain of old Dell computers.

6 months ago I was given a bunch of old Dell P4s that I harvested a few parts from and trying to decided if there is a single one worth keeping, they range from 1.8 GHZ to 3.+ GHz