VOGONS


First post, by tony359

User metadata
Rank Member
Rank
Member

Hi all,

I am diagnosing an Acer V58 - https://theretroweb.com/motherboards/s/acer-v58#docs - the manual is like 95% accurate as I see something is not there.

I took a look at this board in one of my videos, first box: https://www.youtube.com/watch?v=pEeNqD2FOHQ

The board appears to be dead. The CPU is getting proper voltage, proper clock and the reset line works. But there is no code being executed.
I see NO activity whatsoever on the BIOS address/data lines.

I tried Deksor's 80h test BIOS - it works on a different MoBo but not on the V58.

I am following some leads. One being that all the BIOS data lines are like 2.5V as it there was something conflicting on the bus. I noticed the relevant 245 buffer IC is set to "B to A" that is, ISA to BIOS. It's weird as I'd expect that at that early stage it should be "BIOS to BUS" - but I am still looking into that.

I would like to understand better what happens when a system is powered up. I suppose the CPU will reach for the BIOS so I should see some address lines coming to life and then some data being exchanged on the BUS? Is that coming from the CPU itself? In other words, I am going to probe the address/data lines at the back of the CPU, am I supposed to see something there when RESET is released?

Oh and of course the CPU and the RAM do work on a different motherboard.

Thanks for now!

My Youtube channel: https://www.youtube.com/@tony359

Reply 1 of 10, by tony359

User metadata
Rank Member
Rank
Member

I found this document for the ALI chipset. So it seems the BIOS is not on the ISA bus which is weird as it goes onto the ISA bus via a LS245.

Anyways, the BIOS communicates with the CPU via a bus which goes to the Southbridge which goes to the PCI bus which goes to the Northbridge which goes to the CPU. Complicated.

Can anybody think of a way to remove some variables here?

Attachments

  • ALI.JPG
    Filename
    ALI.JPG
    File size
    35.41 KiB
    Views
    506 views
    File license
    Public domain

My Youtube channel: https://www.youtube.com/@tony359

Reply 2 of 10, by Roman555

User metadata
Rank Oldbie
Rank
Oldbie
tony359 wrote on 2024-02-26, 17:21:
Hi all, ... I would like to understand better what happens when a system is powered up. ... Thanks for now! […]
Show full quote

Hi all,
...
I would like to understand better what happens when a system is powered up.
...
Thanks for now!

Hi!
I think you know about a thing called "Power-On Sequence diagram". Though I really doubt that there're ones for vintage motherboards.
IMO, a datasheet of chipset is a good starting point to fully understand how a system works. Of course it's a hard way. I hope the schematics will help a little bit

Attachments

  • Filename
    asus_motherboard_elementary_troubleshooting_guide.zip
    File size
    168.6 KiB
    Downloads
    17 downloads
    File comment
    old troubleshooting guide
    File license
    Fair use/fair dealing exception
  • Filename
    GA-5AA.zip
    File size
    544.37 KiB
    Downloads
    16 downloads
    File comment
    schematic manual
    File license
    Fair use/fair dealing exception

[ MS6168/PII-350/YMF754/98SE ]
[ 775i65G/E5500/9800Pro/Vortex2/ME ]

Reply 4 of 10, by tony359

User metadata
Rank Member
Rank
Member

I've now probed things on the CPU end (at the back of the CPU) and I do not see any activity whatsoever. Nothing.

There is clock, voltage and reset works. But the CPU does not start. I checked the HOLD, HLDA (Hold Acknowledge), STPCLCK pins, nothing points to something requesting the CPU to stop.

Are there other pins which might be preventing the CPU from starting? I do feel it's the motherboard somehow (a jumper?) to tell the CPU not to start. My guts feelings 😀

My Youtube channel: https://www.youtube.com/@tony359

Reply 5 of 10, by Roman555

User metadata
Rank Oldbie
Rank
Oldbie
tony359 wrote on 2024-02-27, 12:17:

There is clock, voltage and reset works. But the CPU does not start.

Do you mean CPURST signal is 0 (inactive) ?

[ MS6168/PII-350/YMF754/98SE ]
[ 775i65G/E5500/9800Pro/Vortex2/ME ]

Reply 7 of 10, by Roman555

User metadata
Rank Oldbie
Rank
Oldbie
tony359 wrote on 2024-02-27, 19:11:

Can I check which one is CPURST? I have a RESET on the pinout, AK20 - that goes high for a moment and then it goes low as expected.

Then it's OK. Also I would inspect for scratches places near board's mounting holes and socket7 CPU-FAN hooks.
Once I repaired a mainboard that behaved like this (turned on but didn't post) because of just one broken wire near a CPU socket.

[ MS6168/PII-350/YMF754/98SE ]
[ 775i65G/E5500/9800Pro/Vortex2/ME ]

Reply 8 of 10, by tony359

User metadata
Rank Member
Rank
Member

Unfortunately I've inspected the board physically multiple times, it's in very good nic and I cannot find any damage. Hence why I started probing pins! 😀
My guts feeling say that it's the motherboard itself stopping the CPU. When I got it, the RESET line would not disable and turned out there is a "Resume Power Supply" feature which can be disabled. I suppose the board supports some sort of pre-ATX power supplies - it also has a Power On pin header for a switch.

Once the "Resume PSU" is set to "Traditional", the reset line works. But is still doesn't do anything. I have a feeling there is some other jumper which might be causing this. The manual is only 90% accurate, some jumpers are mentioned incorrectly. There is also a "HW Monitor" feature which is barely mentioned in the manual.

My Youtube channel: https://www.youtube.com/@tony359

Reply 9 of 10, by tony359

User metadata
Rank Member
Rank
Member

Just for the records, I finally got that board working! 😀

Final video is here: https://www.youtube.com/watch?v=Q8inNVAGj4A

I'm told the V58 was used on Fujitsu MV205, MVII265, MVII267 Micro Tower sized system. Is there a manual somewhere? Searching for those systems I stumbled into this

dosvboard-768x576.jpg
Filename
dosvboard-768x576.jpg
File size
101.82 KiB
Views
171 views
File license
Public domain

And the "health monitor" connector at the bottom right is NOT populated!!

My Youtube channel: https://www.youtube.com/@tony359

Reply 10 of 10, by rasz_pl

User metadata
Rank l33t
Rank
l33t

Capacitor starts shorted until it charges up - thats how most early Reset circuits work. Much simpler than weird circuits with relays.

Open Source AT&T Globalyst/NCR/FIC 486-GAC-2 proprietary Cache Module reproduction