VOGONS


First post, by d00mo

User metadata
Rank Newbie
Rank
Newbie

My current retro build has been working fine, but suddenly it won’t start anymore…
I have stripped it to minimal setup with just PSU, CPU, memory and graphic card.
When turning on, the PSU starts and the CPU fan starts spinning, but nothing more happens after that. No POST signal or picture shown on monitor.

Any ideas what could be wrong?
This has worked flawlessly up until a few days ago…

IMG_2884.jpeg
Filename
IMG_2884.jpeg
File size
1.04 MiB
Views
889 views
File license
Fair use/fair dealing exception

Reply 1 of 17, by oh2ftu

User metadata
Rank Member
Rank
Member

Check PSU voltages (without it connected to the motherboard).

Remove CPU, check voltages. Test the different components in a different mobo (CPU, RAM, VGA). Check CMOS-battery voltage.

Check for an FSB (can't recall the pin)

I had an Epox VP3 that crapped it's 3,3V rail when the PWM-controlled died. Pushed 5V *everywhere* where it should not be. That board became spares ...
Also killed the CPU, no surprise here.

Reply 4 of 17, by d00mo

User metadata
Rank Newbie
Rank
Newbie

Thanks for all replys and suggestions!
I think the standalone components are working and that is something with the power on the motherboard. When I connect the PSU to an outlet, the PSU immediately starts up and the CPU fan starts as well, before I use the power switch button connected to the motherboard. But no beep, and nothing more happens, only spinning fans from PSU and CPU fan.
Tried out with another PSU, and got the same result.
Any thoughts why that happen?

Reply 6 of 17, by rasz_pl

User metadata
Rank l33t
Rank
l33t

sadly no diagram/boardview available for this board
PSU is sadly controlled directly by PIIX4 (southbridge), there is a slight chance board generates its own 3.3V and its bad, another option is Power button signal being buffered with a transistor that died, all in all you need to measure some voltages and trace atx power on pin

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

Reply 7 of 17, by d00mo

User metadata
Rank Newbie
Rank
Newbie
rasz_pl wrote on 2024-04-27, 14:37:

sadly no diagram/boardview available for this board
PSU is sadly controlled directly by PIIX4 (southbridge), there is a slight chance board generates its own 3.3V and its bad, another option is Power button signal being buffered with a transistor that died, all in all you need to measure some voltages and trace atx power on pin

Thanks for detailed explanation!
These things are a bit more complex than what I am used to handle, but sounds like very likely causes.
So if it is any of those two (bad 3,3v generation or bad transistor), is there any simple way of fixing /replacing them?

Reply 8 of 17, by lti

User metadata
Rank Member
Rank
Member

I see a broken cap next to the CPU socket.

Using these pictures as reference:
http://hw-museum.cz/mb/101/asus-tx97-xe

Q8 might be a 3.3V regulator (with Q9 used to connect 3.3V to the CPU core supply if an old single-rail CPU is installed), but it's hard to tell with only some pictures. However, that's a small FET without much cooling, so I think it's more likely that 3.3V is coming directly from the power supply. The CPU voltage should be on the inductor next to the RAM slots (L6?). You might see something if you measure the voltage on all three pins of Q8 and Q9 along with the inductors (L15 should be connected to 5V). Note that Q1 is actually a diode, and the two pins close to the board edge should connect to ground.

Also, what CPU do you have installed, and what are the Vcore jumpers set to? That's a reference point for whether the voltage readings above are correct.

Coincidentally, I have a different model of motherboard with the same chipset and the same symptoms. I haven't tried to fix it yet because I don't have a spare power supply.

Reply 9 of 17, by d00mo

User metadata
Rank Newbie
Rank
Newbie
lti wrote on 2024-04-29, 04:17:
I see a broken cap next to the CPU socket. […]
Show full quote

I see a broken cap next to the CPU socket.

Using these pictures as reference:
http://hw-museum.cz/mb/101/asus-tx97-xe

Q8 might be a 3.3V regulator (with Q9 used to connect 3.3V to the CPU core supply if an old single-rail CPU is installed), but it's hard to tell with only some pictures. However, that's a small FET without much cooling, so I think it's more likely that 3.3V is coming directly from the power supply. The CPU voltage should be on the inductor next to the RAM slots (L6?). You might see something if you measure the voltage on all three pins of Q8 and Q9 along with the inductors (L15 should be connected to 5V). Note that Q1 is actually a diode, and the two pins close to the board edge should connect to ground.

Also, what CPU do you have installed, and what are the Vcore jumpers set to? That's a reference point for whether the voltage readings above are correct.

Coincidentally, I have a different model of motherboard with the same chipset and the same symptoms. I haven't tried to fix it yet because I don't have a spare power supply.

Ok! Do you mean this one?
I thought this was just bent, but perhaps that indicates that it is broken?

IMG_2887.jpeg
Filename
IMG_2887.jpeg
File size
106.23 KiB
Views
625 views
File license
Fair use/fair dealing exception

Reply 11 of 17, by d00mo

User metadata
Rank Newbie
Rank
Newbie
lti wrote on 2024-05-01, 04:09:

At that angle, it looks like one lead is ripped out. That might not be the only problem with this board, though.

I see… I don’t even have a multimeter so not sure if I wanna go down the road and troubleshoot each pin/connection 😀
Replacing a capacitor I could try out, sounds like quite doable.. but as you say, might not end there.
Really appreciate all your input and thoughts, though!
I might reconsider and revisit this as a project 😀

Reply 12 of 17, by d00mo

User metadata
Rank Newbie
Rank
Newbie
lti wrote on 2024-04-29, 04:17:
I see a broken cap next to the CPU socket. […]
Show full quote

I see a broken cap next to the CPU socket.

Using these pictures as reference:
http://hw-museum.cz/mb/101/asus-tx97-xe

Q8 might be a 3.3V regulator (with Q9 used to connect 3.3V to the CPU core supply if an old single-rail CPU is installed), but it's hard to tell with only some pictures. However, that's a small FET without much cooling, so I think it's more likely that 3.3V is coming directly from the power supply. The CPU voltage should be on the inductor next to the RAM slots (L6?). You might see something if you measure the voltage on all three pins of Q8 and Q9 along with the inductors (L15 should be connected to 5V). Note that Q1 is actually a diode, and the two pins close to the board edge should connect to ground.

Also, what CPU do you have installed, and what are the Vcore jumpers set to? That's a reference point for whether the voltage readings above are correct.

Coincidentally, I have a different model of motherboard with the same chipset and the same symptoms. I haven't tried to fix it yet because I don't have a spare power supply.

Ok, I got hold of a multimeter and have done some voltage measuring!

Inductor L16 (it's not L6): 1,00V
Inductor L15 : 5,03V (as you suggested)

Q8
Pin1: 3,63V
Pin2: 5,03V
Pin3: 5,90V

Q9
Pin1: 0,99V
Pin2: 3,63V
Pin3: 0,03V

Can anything be concluded from these values?

Btw, the capacitor gives a reading so I guess it's not broken.

Reply 14 of 17, by lti

User metadata
Rank Member
Rank
Member

The controller is an HIP6008, which seems to have been used on lots of other motherboards. However, 1.0V on L16 is way too low (assuming that this is actually Vcore - there's nothing else here that looks like it's capable of supplying that amount of current), so there's definitely a VRM problem. It sounds like the VRM is a common failure on these boards, but without a schematic, it will take some more time to figure out why Vcore is too low (unless someone else already figured it out). Some detailed close-up pictures of the corner of the board around the HIP6008 might help. Component markings aren't readable in the hw-museum.cz pictures, ans you appear to have an earlier revision without the VID3 jumper. I'm thinking that there are three options - Q2 is bad (unlikely), the output is overloaded (measure resistance from L16 to ground to verify), or something failed in the feedback/compensation components (connected to pins 7 and 8 of the HIP6008).

Based on those voltage measurements, Q8 really is a 3.3V regulator (actually slightly high at 3.6V, but that was common) instead of using the 3.3V rail from the power supply. It's just a FET and not the LT1084 marked on the silkscreen (my initial thought was that the board was designed for either, but the pinout is wrong for an LT1084 - input and output are swapped).

By the way, on these packages, pin 1 is the pin marked "gate" in the datasheet:
https://www.mouser.com/datasheet/2/308/rfd3055sm-1196049.pdf

Reply 15 of 17, by d00mo

User metadata
Rank Newbie
Rank
Newbie
lti wrote on 2024-05-14, 04:05:
The controller is an HIP6008, which seems to have been used on lots of other motherboards. However, 1.0V on L16 is way too low ( […]
Show full quote

The controller is an HIP6008, which seems to have been used on lots of other motherboards. However, 1.0V on L16 is way too low (assuming that this is actually Vcore - there's nothing else here that looks like it's capable of supplying that amount of current), so there's definitely a VRM problem. It sounds like the VRM is a common failure on these boards, but without a schematic, it will take some more time to figure out why Vcore is too low (unless someone else already figured it out). Some detailed close-up pictures of the corner of the board around the HIP6008 might help. Component markings aren't readable in the hw-museum.cz pictures, ans you appear to have an earlier revision without the VID3 jumper. I'm thinking that there are three options - Q2 is bad (unlikely), the output is overloaded (measure resistance from L16 to ground to verify), or something failed in the feedback/compensation components (connected to pins 7 and 8 of the HIP6008).

Based on those voltage measurements, Q8 really is a 3.3V regulator (actually slightly high at 3.6V, but that was common) instead of using the 3.3V rail from the power supply. It's just a FET and not the LT1084 marked on the silkscreen (my initial thought was that the board was designed for either, but the pinout is wrong for an LT1084 - input and output are swapped).

By the way, on these packages, pin 1 is the pin marked "gate" in the datasheet:
https://www.mouser.com/datasheet/2/308/rfd3055sm-1196049.pdf

Here are two closeups around HIP6008, with some slight different focus.
I don’t know if they are helpful?

I will try get a reading of resistance L16 to ground as well.

IMG_2934.jpeg
Filename
IMG_2934.jpeg
File size
1.22 MiB
Views
332 views
File license
Fair use/fair dealing exception
IMG_2937.jpeg
Filename
IMG_2937.jpeg
File size
893.22 KiB
Views
332 views
File license
Fair use/fair dealing exception

Reply 17 of 17, by lti

User metadata
Rank Member
Rank
Member

That resistance is correct.

Unfortunately, there are enough traces running under the chip that the circuit isn't obvious here (and in my earlier posts, I was just taking a guess based on other motherboards and cheap voltage regulator circuits I've seen).