• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Solaris17s:CPU pins broke? Fix it.

Status
Not open for further replies.
btw i lied. it was a 5200BE not a 5000. Guess when it sits on a shelf for over a year you kinda neglect to look at the IHS

1581008960.jpg
 
yup. also if you read it near the pin diagram it says the dangers of crossing pins etc. Be carefull and you should be fine.

Besides the issues of crossing the pins, the main issue with doing this on a 478 system is the extremely tiny pins. I seem to remember 478 pins and processor being extremely small, even compared to AM2/3. Maybe thats just me though, as it has been a while since I've messed with anything 478.
 
Besides the issues of crossing the pins, the main issue with doing this on a 478 system is the extremely tiny pins. I seem to remember 478 pins and processor being extremely small, even compared to AM2/3. Maybe thats just me though, as it has been a while since I've messed with anything 478.

478 is tiny, the pins are packed close together indeed.
 
just to follow up with mr. helios

while i havent ran it through prime again its been running WCG for about 4 days now (woops wrong system make that 2 days) non-stop 24/7 at 100% load with no issues while peaking at 56c. (it passed a few hours of the last prime. A little over 3 hours before i started WCG back up needed to format to an unused key)
 
Last edited:
68088296.jpg


bumped 300mhz everything is fine. missing pins seem to have no adverse effect. i wont take it any higher though. my PSU's running my AMD machines blow at these speeds the rails are dipping bad.
 
Last edited:
Great thread - thanks
Just a thought re. the alternatives of reattaching a pin to the CPU:
If you bond the pin strongly to the solder pad on the CPU silicon, and it gets another knock, could the solder pad be ripped off, leaving the CPU non-revivable? The conductive epoxy might be better because it would be weaker.
I don't know how conductive such materials are though, and your fix might not handle power supply.
Has anyone tested, eg Arctic Silver epoxy TIM, or Devcon metal filler or similar?

If you do want to try soldering, paste solder might be the go. I've never seen the stuff, but I believe it's used to attach surface-mount components to PC boards. It's a mixture of powdered solder plus solder flux. You'd put a dab on the CPU pad, hold the pin in place upon it, and hit it with a heat gun. Sounds easier than using an iron.
 
CPU BROKEN PIN RE-SOLDERING FOR EXPERTS ONLY IN SOLDERING

I just did this, de-solder pins off a non working CPU & re-soldered a single pin to broken CPU.

There two ways to carry out this fix.

FIRST METHOD.

I used a small tip soldering iron 0.4 to 0.8 & de-soldered a single pin from a dead CPU (480C) Touching the side of the pin ie the PCB & the side of the pin together. Do the same to the broken pin on the CPU.

Now hold the pin in place with precision tweezers & solder to broken CPU using the same method not forgetting a little flux. This method is not as clean as using a full infrared workstation.


SECOND METHOD (INFRARED WORKSTATION)

Same as above, de-solder CPU pins, but use tweezers to pick them off the dead CPU. Pick the cleanest looking pins from a bunch (workstation).

Do not use infrared to de-solder broken CPU pin, use soldering iron method.

Use flux on CPU pad & use tweezers to place it in the centre of the pad. Isolate as much as possible CPU core & most of the pins that are not being soldered.

Turn-on pre-heater & infrared & re-solder pin.

NOTE: I did not add any extra solder to the pad or any pins & FX-57 is now working.

SUMMARY

If you have many broken pins on your CPU, it makes no difference "ALL CPU'S" with broken pin(s) can be repaired. I can go as far as removing all of the pins & make it look like LGA & put them all back. Infrared workstation gives it a professional finish & you can't tell the difference with the naked eye.
 
Status
Not open for further replies.
Back
Top