dicky96 wrote:

If I put two Dimms in it hangs with code 40 (Preparing the descriptor tables next)

That's interesting, because it means that the next step of the POST fails, which is putting the CPU into virtual mode to test the memory. Did you test with a different CPU? (not a different one of the same type, but a different one that the board is compatible with)

A corrupt BIOS is very unlikely, as it does a self-test early and would not even come that far if the BIOS checksum would not match.

Did you rule out a bad power supply already?
When the CPU enters virtual mode there is probably a power surge, and if the voltage sags sufficiently, that may hang stuff.

Yes I ruled out a bad PSU, I have the motherboard bread boarded on my bench and using a known good PSU

Regards the CPU yes I tried a known good Pentium D (which the manual says is compatible). Also the Core 2 Quad I took from this motherboard boots OK in my known good Gigabyte GA-G31M-E2Sl (which previously had the Pentium D). I use the gigabyte as a general test board for LGA775 CPUs DDR2 Dimms and PCI-e graphics cards

I'll just check for bent pins on the socket now

@derrSammler
where did you find the order of the post codes sequence (which test comes next)

Sorry it's not a Pentium D that's in the Asus now, its a Pentium Dual Core E2180 and the manual says that is compatible

OK I had it under the microscope - there was a little bit of crap (looked like heatsink compound) on one pin in the 775 socket and also on the CPU

I cleaned that off. No bent pins. Put it back together and I still have the same codes

D0-98-F0 with only one of the DDR3 dimms in socket A1 or A2 (either dimm, doesn't matter)
D0-98-40 with two dimms in Ai/B1 or A2/B2

Actually occasionally I can get D0-98-40 with one dimm fitted, then try again and I get D0-98-F0

I also tried a couple other known good dimms. Same result

@retardware, I thought about what you said regards the power supply. I removed the dimms so the board beeps on power up (no ram) and measured the voltage on the MoFETS that power the CPU (seem to be 5 of them in pairs unless some are hidden under the heat sink. I get 1.302V there. I then fitted one dimm and booted so it hangs on code F0. I still get 1.302V. I suppose I could look for ripple or a spike on 1.302V rail on my scope...

y'all need HP Sure Start

derSammler wrote:

A corrupt BIOS is very unlikely, as it does a self-test early and would not even come that far if the BIOS checksum would not match.

Actually what would one normally observe if they had a corrupt BIOS?

Nothing - no POST code and no beeps. The BIOS stops immediately if the checksum is wrong. And should it be so badly damaged that it can not even check its checksum anymore, nothing will happen anyway, since no code execution happens or the CPU stalls.

Interesting
I have several faulty motherboards here that I picked up at the local flea market (along with numerous good items that I then sell to by more retro hardware)

I'm trying to learn component level motherboard repair so practice on these boards

So do you think this thread may be a faulty BIOS then?
ECS P6BAT-A+ Slot 1 + Skt370 Motherboard Problem

I posted the thread before i had the PCI analyser card but i have tried the analyser on it and it just sits at 00 with no activity. I replaced the corroded crystal I mention, X2

Have you checked for bad DIMM sockets? By which I don't mean broken sockets, but sockets with connection issues due to dirt, broken traces, etc.

No but I will have a look under magnification today.

Actually I was looking at those DDR2 and DDR3 slot testers yesterday, and also socket 775, 1155 etc testers and was thinking of getting some. But all the reviews/tutorials I could find for them on youtube etc were either in Russian or some Indian language. So I can't find any info on using them 🙁 It would seem some supposedly 'third world countries' are far ahead of the 'developed world' when it comes to fixing computers.

OK I had a look under the microscope and there diesn't seem to be a problem with the RAM slots. Also putting a single stick of memory into any slot seems to have pretty much the same effect so it's not like i have one slot behaving different form the others.

A single stick generally stops on F0 in slot A1 or A2 (but sometimes 40) and stops on 40 in slot B1 or B2.

Putting more than one dimm gives 40

I have some 1Gb Samsung PC3-8600U and some other brand (M logo with an ellipse around it) 2Gb PC3-12800U. Both types give the same result. With no Ram it stops at D3 and beeps like you would expect.

I'm wondering if maybe the North bridge has some problem seeing as that is how the CPU accesses memory? Maytbe if I bought those those DDR slot and CPU socket testers they would give some useful indication?

As an aside are there any other good online forums for PC (not Apple) motherboard repair? There are some clever guys here for sure and Badcaps has a section for laptop motherboard repair (though not for desktop mobo repair), but I didn't find any others and most repair videos on youtube are indian language for some reason. Yes I know about Eli Tech but he's pretty much GPU only.

Rich

dicky96 wrote:

I'm trying to learn component level motherboard repair so practice on these boards

So do you think this thread may be a faulty BIOS then?
ECS P6BAT-A+ Slot 1 + Skt370 Motherboard Problem

Hey guys - I had some success with the Slot 1 board. ECS P6BAT-A+ Slot 1 + Skt370 Motherboard Problem thanks to the posts from derSammler on this thread
😁

OK back to the Asus P5P43TD
I downloaded the latest version and flashed the BIOS . In this case it did not make any difference, the POST still hangs on F0

I've taken the heatsink off the North Bridge it is an AC82P43 SLB89. The heatsink compound was all hard and powdery like cement. Maybe that is a clue? I had a look on aliexpress and I can get these for less than 5 euros inc postage. I was thinking of ordering a couple and seeing if I could change it. https://es.aliexpress.com/item/32887250768.ht … rchweb201603_52

In the meantime I will order one of the DDR2/3 slot testers and also the CPU socket tester for LGA775 to see if they are any use.

I'm getting a better hot air station (Quick 861DW) next time I pop back to the UK in a few weeks so hopefully I would have that by the time the chips arrive, and maybe it would be up to the job.

I've never changed a BGA before so maybe it would be a useful experience to try anyway. Maybe changing it would even fix the motherboard? Can someone more experienced advise on how likely that is?

Rich