First post, by kuurakarahka
Greetings! Seeing as this forum popped up in many a search about older hardware, I decided to join. I'm not well versed in motherboard repair, and this current project seems like one benefiting from asking others for a change, to speed up both my learning and the mobo's possible resurrection.
This A7V133 I bought new in 2000-2001 with a 1 GHz Thunderbird. Pretty soon I found out the lackluster OC of my chip, and set about making a voltmod as per instructions online. All worked well, lots of volts, lots of heat, lots of noise, some seconds off piFast.
Anyway, I began working on this to make myself a year 2000 PC: Thunderbird with PAL8045, TNT2, the case is late 90s (remarkably I found a case with the very same front panel as the one I got in 1997-1998 with Pentium 2 300 MHz, still in daily use with a "maxed out" LGA775). I have a MSI KT7 Turbo and Soltek 75KAV, too, but the former fell victim to someone impatient (me) and the old PSU took out a little chip SC 1205CS and the parallel port terminator chip. CPU and memory survived. Latter works, but the leaky, bulging mess of caps I do not trust. So, candidate number three it is.
I've primarily tested the mobo with a Duron 800 MHz, jumpers set at 8x100 MHz and 1,85 V. Memory has been random sticks from 64 MB PC100 to 128 and 256 MB PC133. A modern PSU that pushes out 29 A at 5V.
I compiled a sheets doc to collect the troubleshooting so far, which is copied here sans the mosfet measurements, link to the sheet:https://docs.google.com/spreadsheets/d/1xIV_w … dit?usp=sharing
- Fault(s)
no POST, numlock yes, no beep, power led yes, fans spin
VID doesn't work: measures correctly when jumper set at 1,85 (from VRM board pin 1,85V to screw hole, to PSU GND 1,89V). Otherwise: tried setting 1,5, 1,8, measured 1,198V and 1,496V)
DIMM voltage 2,7-3V, measured at slot pin 6
- Misc
voltage reg chip 5322 0035S pin 2 from right bottom (VSB) measures 7,6k to GND
Mosfet 1 is 10N03L, 2,3 and 4 are 32N03L.
The larger coil by ATX connector whines when measuring mosfet 1 voltage from drain to ground
SMD C198 by mosfet 1 look damaged, seems to work, however
- What's done
Has trace damage above the DIMM slot closest to CPU. Traces measured to be intact, they lead to NB.
Has trace damage between CPU and VRM board, right next to the socket (cooler bungle?). Traces repaired: cleaned, scraped open, 1st one only solder, 2nd small copper wire soldered. Repair sealed with nail varnish. Measured working, leading to NB, no shorts.
Had a voltmod back in the day: reg chip pin 2 to GND via a potentiometer (50k, IIRC). Removed, also back in the day.
Mosfets 1 and 3 have been taken off and resoldered at some point.
The black cap by ATX connector has been measured good.
After the trace repair, it stopped beeping low-high-low like a siren
What's next in my opinion? Well, the voltage mosfet 1 outputs is strange, ranging from 2,7 to 3,1, and would seem likely it prevented booting. I'll check the caps out of circuit by the power connector. The other mosfets seem fine, other than #4. D-G and G-S voltages were both 9V, a pattern unlike the others.
Also, how VID behaves is baffling. It's like setting all four jumpers to 1-2 enables some static voltage, 1,85 V, and all the others introduce a modifier, which doesn't work, instead lowering Vcore by ca. 0,3 V. I recall the voltmod spoofed the regulating chip by adding resistance between voltage feedback VSB and GND, so it saw a lower voltage the lower the R was, thereby increasing Vcore. A page describing it:https://www.angelfire.com/electronic/Dulanic/voltmod.htm. Perhaps at 1,85 V setting Vcore comes by some other method, and lower settings all introduce resistance to VSB-GND?
I must mention this project has already provided valuable experience, especially the trace repair, my first, by the socket. It took hours of crouching over the mobo in the early hours of night turned dawn, but finally the traces accepted the lone strand of copper in union by solder. What a harrowing location, some 5 mm from the socket, almost under the center tab. A more open location would have allowed for taping a soldered wire taut over the trace and kept it in place for joining. But no, dangling the almost invisible wire I was.
That's all for now, let me know if you can help and/or need more information. Thank you for reading this far.
