kotel wrote on 2024-08-10, 18:30:
Thank you for the reply. Yes I do understand that the coil has not enough windings, but its not like I will game on it due to the broken AGP slot. I think the cause for the dead AGP slot will be missing connections from and to NB. What I don't know is the cause for the CMOS not keeping settings.
It doesn't really matter if you will game on this board or not or even if you just plan to use it "lightly".
For starters, socket 462 CPUs don't have the idle power saving features/states that the equivalent Intel CPUs had at the time. As such, socket 462 CPUs pretty much run full-tilt both in idle and under 100% load. The difference is usually about 10-20% difference in the power consumption, if even that much. For example, an Athlon XP 2500+ system of mine uses about 100 Watts in "idle" mode and 117 Watts under 100% CPU load. The only way I can make a dent in the power consumption of that system is when I swap the CPU with something less powerful, like my Duron Applebred 1400 (T-bred "B" core), which has a max TDP of 54W (IIRC) compared to the almost 80 Watts of the 2500+.
But anyways, all I'm trying to say is you need to get the right inductors / coils from the start for the board to run right.
That being said, the inductors you chose for L1 and L3 (the toroidal ones near the CPU) appear to be more or less close enough to what was there originally, at least compared to the pictures on the board from TRW.
HOW TO PICK EQUIVALENT INDUCTORS:
1) The core type of the inductor must match.
I'm not exactly sure what that light-green core's parameters are. But if my memory serves me right, the Xbox 360 motherboards used those and the all-black toroidal inductors almost interchangeably between various designs, so they must be closely related. The all-black cores should be Micrometals -45 type core. It's a popular core type that was on many motherboards of that era (Pentium 4 / Athlon XP / Athlon 64), especially on ASUS and AsRock motherboards.
2) The inductor windings must be relatively close to the originals in terms of (equivalent) thickness. In regards to that, the replacements you used appear to have only 2 wires, whereas the originals seen in the picture on TRW appear to have 3 wires. Otherwise, the # of turns appears to be the same (I'm counting 4 turns for your replacement coils and also 4 turns from the picture on TRW.) And given the core appears to be the same type, that should be making the inductance the same. Only the wire thickness is smaller on your replacements. This means the inductors will heat up more at higher currents compared to the originals. To make sure you don't over-heat them, use a CPU with a lower-rated TDP (lower power dissipation).
As for L2, I'm guessing that's the one in the upper-right corner of the motherboard near the ATX connector? If so, that one isn't too critical to match, especially when it's a "rod" type rather than a toroidal one, like the two near the CPU socket. All that L2 does in this case (assuming the above assumption is true) is filter out EMI/RFI from the CPU VRM to get back to the PSU (and the rest of the components in the computer.)
majestyk wrote on 2024-08-10, 16:15:
It´s hard to guess the exact electrical parameters of the inductors without any(!) pictures provided.
Agreed.
For these kind of questions, pictures should always be provided.
Not to sound too harsh here, but I really don't understand how people expect anyone to know the exact answer to their very very specific technician-type of question without providing any pictures. Electronics repair is largely a "must see with my eyes" type of work. Not providing pictures is just not serious. It's the equivalent of calling your doctor and telling him/her you have a cough and then expecting them to know exactly what medicine or treatment to give you based on that . Sorry, but it's not gonna work. Either the doctor needs to see you or you need to describe your symptoms very very precisely for them to get even a remote idea of what the problem might be. Going back to the question in hand in this thread, you could have done this by at least describing all of the inductors' parameters (size, color, # of turns, # of wire strands, and etc.) on all of the motherboards you had available for parts. Of course, at that point, you might as well snap a few pictures instead. Or as the saying goes "a picture is worth a 1000 words". For some of these technical questions, it can be worth even more.
Again, forgive me if I sound harsh and long-winded here, but it just really irks me when I see HW troubleshooting threads without pictures, especially when it comes down to component-level troubleshooting.
Horun wrote on 2024-08-10, 20:12:
Does it keep time ? The SB VT8237 has the RTC and cmos ram. Maybe the CLRCMOS jumper or traces is shorting somewhere...
Probably unrelated but One of my VIA based boards had issues with savings settings and it was the EEPROM where the ESCD stuff is stored, it had been written too so many times that portion of the eeprom went bad. Replaced it and all good...
Good call.
I think I've read about this issue / remedy once on badcaps too, so it's probably not that of an uncommon issue.
