Anyone know how deep the traces are on the PCB around the NB for a A7N8X?

B

Barnaby

Supreme [H]ardness
Joined
Aug 7, 2001
Messages
5,280
Ok, so I wanted some more performance out of my A7N8X Deluxe v1.04, so I decided to upgrade the NB cooler to a swiftech 159..

so, in the process of removing the stock HS, I managed to indent part of the PCB around the NB (the PCB on the northbridge itself, just around the heatspreader) (frickin' TIM was like concrete on that mug).. looking closely at it, I can see traces EVERYWHERE on the PCB, and I'm sure theres no way that the ding wouldn't be on top of one.

It's not a gash, and it's not really big, but it's definately noticeable.. so my question is, how deep are the traces on the pcb? are they really as close to the surface as they look? and if that's the case do you think my NB is going to have problems?

I only had a enough time to boot up, and see what happened.. It seemed to be ok, but I'm definately going to stress test it by running everything I can think of at it for awhile..

Would you trust it, after running numerous stress tests? do you think they'd expose some kind of issue caused by potentially hurting a trace?
 
Surface (microstrip) traces are usually under about 1 mil of soldermask, and are about 2 mils thick (about 1.5 ounces after plating). Ya can generally LOOK to see if you cut a trace - it's possible to do, but harder than it looks.

A mil's 1/1000 inch.
An ounce is something like 1.37 mils; it's the thickness of a sheet of copper foil that weighs one ounce per square foot.
 
I don't think I cut it, I couldn't see anything cut, but it was definately dented a tad.. not sure if it just dented the surface, or if it hit a trace... I'm probably just being paranoid, as the machine did boot like normal.. so, If a trace was cut, I'd think booting up would show some kind of error, since every system get's used during boot up..
 
well, upon my initial testing, I'm not sure... there are still a few variables left to test, but upon some quick stress testing, my system proved to be unstable..

I tried to loop 3dmark for awhile, as I ran a bunch of tests using prime95 last night just fine, but it crashed to the desktop on me.. granted 3dmark is kinda buggy (using 3dmark2001se for testing, dx8 card) so, I didn't get to upset at first..

then I tried a quick OC on my FSB, and I couldn't get past 190 without it failing windows.. (i've had it in windows @ 195 before)

to be fair that was with all the tightest timings, I normally use @ 166, as I was tired, and didn't have alot of time to do it proper..

I tightened the MX159 on the NB alittle more, just to make sure it was on there nice and tight, and I left it running seti for the day.. I'll test oc'ing, and other bench's over the weekend, and possibly try reseating the NB cooler.. but if all that fails, guess I'll be posting about how my dfi nfII infinity kicks ass over the a7n8x.. ;)
 
Originally posted by KalTorak
An ounce is something like 1.37 mils; it's the thickness of a sheet of copper foil that weighs one ounce per square foot.
Click to expand...

That really depends on the absolute density of the material based on chemical composition but your numbers are close enough :D
 
You must log in or register to reply here.
Back
Top