VOGONS


First post, by Victorkorp99

User metadata
Rank Newbie
Rank
Newbie

Hi y'all, I recently bought a Gigabyte GA-5AX Rev 5.2 wich came in a small lot and after doing a few tests it ran unstable and after a while I found an almost knocked cap near the ram slots, so I decided to take advantage of the situation and do a full recap using polymers due to how long they last and their low ESR

The issues came after being able to use the machine when I noticed some coil whine (I guess it was), when testing a K6-2 550, which dissapeared after downgrading to a P1 MMX 233 and other issue I noticed was the linear regulator near the AGP port (LX8383A) that gets really hot at normal use and a lot more when playing, I can't measure the temperatures rn but when is idle I can maintain my finger over the plastic part of the component but not after playing something demanding (I used UT99). So far I tried to change the jumpers for VCC3 to free some stress of the regulator (iirc, it routes the PSU 3.3v to the AGP slot), and changed the first GPU I was using (GF4 MX440) for a GF2 MX400

Neither of those changes made the overheating issue lesser and I'm regretting not having measured the regulator before doing the recap so I could knew if it was an issue before. I will add photos of the motherboard as is now, a photo of the problematic regulator and some of the capacitors

Also I want to made clear the changes I made regarding the so called "poly mod" :
Originally the board had 25 330uf 25v, some of those had 10mm of diameter, others 8mm, I changed accordingly to what I had available in my local distributor and almost all were replaced with the Chemi-Cons rated at 270uf 16v 8mm and I put a pair of Lelon OCRZ 1000uf 16v (the ones at the right of the CPU), replacing 4/5 capacitors (1 was originally unpopulated, so I assume the board was prepared for use it) , with values of 330uf 25v 10mm in a parallel configuration, with those two

I pretty much respected the around 20% tolerance, regarding the capacitance value, I don't think those are the culprits but I'm not entirely sure, so I posted for asking about some advice about what to do, the board works just fine but I'm concerned on the mediun-long term before running into a problem with this rare board, anything is welcomed

Attachments

  • 20230125_151632.jpg
    Filename
    20230125_151632.jpg
    File size
    1007.34 KiB
    Views
    317 views
    File comment
    LX8383a
    File license
    Public domain
  • 20230126_004328.jpg
    Filename
    20230126_004328.jpg
    File size
    679.75 KiB
    Views
    317 views
    File comment
    Caps
    File license
    Public domain
  • 20230126_004356.jpg
    Filename
    20230126_004356.jpg
    File size
    2 MiB
    Views
    317 views
    File comment
    Motherboard already modified
    File license
    Public domain
Last edited by Victorkorp99 on 2023-01-26, 06:13. Edited 1 time in total.

Argentina, Córdoba

System: - Pentium III-S 1.4 (Tualatin) + 512mb RAM PC133 + Lite-On TR100 (i815ep) + Ati Radeon 9550 256mb + SB Live! + Generic Usb 2.0 Card + 3Com Etherlink III+ 64gb SD Card (IDE to SD adapter)

Reply 1 of 2, by Victorkorp99

User metadata
Rank Newbie
Rank
Newbie

Small update:

- I changed one polymer capacitor of 1000uf 16v for two 270uf 16v polymers I had to lower the overall capacitance to reach tolerance values. No change

- Measured [Vout] of the regulator and throws 3.5v and 5v for [Vin]

Argentina, Córdoba

System: - Pentium III-S 1.4 (Tualatin) + 512mb RAM PC133 + Lite-On TR100 (i815ep) + Ati Radeon 9550 256mb + SB Live! + Generic Usb 2.0 Card + 3Com Etherlink III+ 64gb SD Card (IDE to SD adapter)

Reply 2 of 2, by Victorkorp99

User metadata
Rank Newbie
Rank
Newbie

Update 2:

Nobody wouldn't believe me but I solved the issue with a extremely easy move

What happened is that is was looking at the board and I saw the jumpers for VCC3 (jp13) that adjusts the output of the regulator, the jumper was shorting pins 2-3 and it was supposed to throw 3.3v but I was getting 3.5v , so I changed the jumper to position 1-3 wich gives you 3.58 and that's what I got but when I changed the jumper to 2-3, the multimeter showed the 3.3v and the regulator didn't overheated anymore!

Now, the friggin component Is just slightly warm at the touch, and i can finally build the machine

I guess it was an issue related to bad contacts, but I'm so glad, it was that simple in the end, thanks to everyone who reach up until this part to found the end!

:'D

Argentina, Córdoba

System: - Pentium III-S 1.4 (Tualatin) + 512mb RAM PC133 + Lite-On TR100 (i815ep) + Ati Radeon 9550 256mb + SB Live! + Generic Usb 2.0 Card + 3Com Etherlink III+ 64gb SD Card (IDE to SD adapter)