Seller Note “The card physical looks perfect. No blown caps or damage I can see. Replaced the thermal paste on the cooler. Tried a different system and PSU. Whatever I try, Windows always errors with the card, giving error 43 in device manager.”
Summary
- Resistances seemed OK.
- Fails MATS see below, consistent with code 43, had hoped it could be a memory issue.
- Seems repaired after replacing the faulty memory chip 🙂
mats version 400.184. Testing GP106 with 20 MB of memory starting with 0 MB.
Read Error Count: 0
Write Error Count: 2631040
Unknown Error Count: 0
=== MEMORY ERRORS BY SUBPARTITION ===
SUBPART READ ERRORS WRITE ERRORS UNKNOWN ERRS
------- ----------- ------------ ------------
FBIOA0 0 0 0
FBIOA1 0 0 0
FBIOB0 0 0 0
FBIOB1 0 0 0
FBIOC0 0 2631040 0
FBIOC1 0 0 0
Failing Bits:
C024 C025 C026 C027 C028 C029 C030 C031
=== MEMORY ERRORS BY BIT ===
P : Partition (FBIO)
READ 0 READ 1 READ ?
P BIT READ ERRORS WRITE ERRORS UNKNOWN ERRS EXP. 1 EXP. 0 EXP. ?
- --- ----------- ------------ ------------ ------ ------ ------
C 024 0 1099934 0 437035 662899 0
C 025 0 1423214 0 331613 1091601 0
C 026 0 1315553 0 873601 441952 0
C 027 0 1530895 0 221224 1309671 0
C 028 0 1423257 0 1091564 331693 0
C 029 0 1423218 0 331637 1091581 0
C 030 0 1207877 0 655531 552346 0
C 031 0 1530982 0 221154 1309828 0
=== MEMORY ERRORS BY ADDRESS ===
ADDRESS : Failing memory address, or buffer offset if starting with 'X+'
T : Type of memory error: W = write, R = read
P : Partition (FBIO)
S : Subpartition
B : Bank
E : Beat
ADDRESS EXPECTED ACTUAL REREAD1 REREAD2 FAILBITS TPSBE ROW COL BIT(s)
------- -------- ------ ------- ------- -------- ----- --- --- ------
000135ecbc 00000000 39000000 39000000 39000000 39000000 WC097 0067 025 C024,C027,C028,C029
000135ecb8 00000000 ae000000 ae000000 ae000000 ae000000 WC096 0067 025 C025,C026,C027,C029,C031
000135ecb4 00000000 c9000000 c9000000 c9000000 c9000000 WC095 0067 025 C024,C027,C030,C031
000135ecb0 00000000 8f000000 8f000000 8f000000 8f000000 WC094 0067 025 C024,C025,C026,C027,C031
000135ecac 00000000 e7000000 e7000000 e7000000 e7000000 WC093 0067 025 C024,C025,C026,C029,C030,C031
000135eca8 00000000 dd000000 dd000000 dd000000 dd000000 WC092 0067 025 C024,C026,C027,C028,C030,C031
000135eca4 00000000 6a000000 6a000000 6a000000 6a000000 WC091 0067 025 C025,C027,C029,C030
000135eca0 00000000 f3000000 f3000000 f3000000 f3000000 WC090 0067 025 C024,C025,C028,C029,C030,C031
000135ec9c 00000000 39000000 39000000 39000000 39000000 WC097 0067 024 C024,C027,C028,C029
000135ec98 00000000 ae000000 ae000000 ae000000 ae000000 WC096 0067 024 C025,C026,C027,C029,C031
000135ec94 00000000 c9000000 c9000000 c9000000 c9000000 WC095 0067 024 C024,C027,C030,C031
000135ec90 00000000 8f000000 8f000000 8f000000 8f000000 WC094 0067 024 C024,C025,C026,C027,C031
000135ec8c 00000000 e7000000 e7000000 e7000000 e7000000 WC093 0067 024 C024,C025,C026,C029,C030,C031
000135ec88 00000000 dd000000 dd000000 dd000000 dd000000 WC092 0067 024 C024,C026,C027,C028,C030,C031
........
000041a734 00000000 c9000000 c9000000 c9000000 c9000000 WC095 0015 009 C024,C027,C030,C031
000041a730 00000000 8f000000 8f000000 8f000000 8f000000 WC094 0015 009 C024,C025,C026,C027,C031
If you are getting failure for first MB of FB then try option -no_scan_out
Error Code = 00000001
####### #### ######## ###
####### ###### ######## ###
## ## ## ## ###
## ## ## ## ###
####### ######## ## ###
####### ######## ## ###
## ## ## ## ###
## ## ## ######## ########
## ## ## ######## ######## VRAM is Samsung K4G41325FE-HC25, which is good, as I have a couple of spares (same as the MSI GTX 1050 2Gb that also had a chip fail in the exact same position). Hopefully, the replacement can be as successful.
Replacement
- The VRAM chip is surrounded by 0402 capacitors, unfortunately I managed to dislodge 2 when removing the chip!
- After measuring them, they both seem to be 1uF, which I have in stock. Soldered the new ones in.
- After the initial replacement I also didn’t use enough heat, as MATS showed failures on all bits of Chip C0. I reflowed the chip, this time making sure the ‘side nudge test’ was more committed and now it seems to work!
Testing Notes
- The card seems to be passing all benchmarks
- Kombuster HD
- Furmark HD
- Time Spy
- Soak tests
- Heaven – hours testing, was hotter about 62 deg C, but still very good.
- Subnautica going well ~ 52 deg C, Hotspot ~ 60 deg C
- All ports work, but the HDMI port seems quite loose – this should be reported when selling, I wont bother to fix it, as it still works fine.
- Needs a bit of a clean.