VOGONS


First post, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Dono that it’s old but this is the best category I could find for this.

Anyway I have a msi ms-7758 ver 1.3 (w/ intel i5 2500 processor)

It has been buggy for some time and I finally tracked it down to it being intermittently picky about my pci sound blaster live 24 so I pulled that card out and it’s been pretty stable. But that leaves like only two cards out of my collection that it likes, it’s super picky about pci cards and I can’t figure out why.

No issues It seems with a geforce 6200 pci and a usb2.0 card all are pci (regular old school pci)

But when I install a pci to pci express card, a 610gt, a usb 3.0 card, etc it fails to post. (Again, all old school pci)

So today I got a diagnostic card to read codes. And wouldn’t you know it, the manual has over 100 codes in it but it goes straight from 93 to 95 and I have a code 94 which isn’t even in the book. Typical. Lol.

The only pci related setting I see in my bios is a latency timer. Which I’ve tried messing with to no effect.

I’m not sure who makes the bios. It’s got msi branding all over the mouse enabled bios though.

Any ideas?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 2 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
LewisRaz wrote on 2021-01-18, 15:59:

Looks like an AMI bios and that gives this: "94 Memory size set"

Any idea what it means and why specific pci cards would have anything to do with memory

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 3 of 15, by mr.cat

User metadata
Rank Member
Rank
Member

Well, one thing that could be done is to disassemble the BIOS to see where that code is set.
(I don't know very much about this subject, but I think BIOS just pokes the code to port 80h.)
That's an UEFI BIOS right? Could it be possible that the bridge card needs some update or UEFI firmware?
What's the make and model of that card?

There's also a Voodoo trick you could try aka "did you try turning it off and back on again":
1. Try booting, see it fail.
2. Turn off the power.
3. Wait a while (10 secs or so) and try booting again.

That probably doesn't help, but doesn't cost anything to try...

One more suggestion, take a look at your caps near the PCI ports if you didn't already.
This mobo is not from the "bad caps" era I think, but you never know.

Reply 4 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
mr.cat wrote on 2021-01-18, 23:42:
Well, one thing that could be done is to disassemble the BIOS to see where that code is set. (I don't know very much about this […]
Show full quote

Well, one thing that could be done is to disassemble the BIOS to see where that code is set.
(I don't know very much about this subject, but I think BIOS just pokes the code to port 80h.)
That's an UEFI BIOS right? Could it be possible that the bridge card needs some update or UEFI firmware?
What's the make and model of that card?

There's also a Voodoo trick you could try aka "did you try turning it off and back on again":
1. Try booting, see it fail.
2. Turn off the power.
3. Wait a while (10 secs or so) and try booting again.

That probably doesn't help, but doesn't cost anything to try...

One more suggestion, take a look at your caps near the PCI ports if you didn't already.
This mobo is not from the "bad caps" era I think, but you never know.

It’s not that old of a motherboard 2013 or 2013 I think..

I can dump the bios if someone who knows how to do that comes along but I wouldn’t have the first clue. Also the bios I believe is available online still.

I will look around at the caps but they are all solid caps. So unlikely to find anything.

Well, it’s not just that card it’s more than half the cards I own so I’m thinking it is not the cards ya know? Or rather, it is but that can be ignored haha cause they ALL should work and if it was just one that may be a avenue to look at.

Odly, the off then back on trick does sometimes work with the sound blaster. But not at all with the others.

The 6200 works fine, the usb2.0 works fine and the sound blaster works 50% of the time the rest of them..... omg, I may have just figured something out, the usb3.0 (which has a bridge chip), the 610 (which has one also), the pci express (which is a bridge chip) all have bridge chips..

Why wouldn’t bridge chips work in this motherboard?

This doesn’t explain the soundblaster though that unit is from well into the pci-agp era.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 5 of 15, by mr.cat

User metadata
Rank Member
Rank
Member

Yes, I did omit one Voodoo step "do the little chicken-dance, run around the house and let out a little shriek three times", I didn't think it would make a very big difference though 😉
Regarding USB, you could also try booting without any USB devices and see if that helps. These are somehow UEFI related issues, but I don't know the actual root cause.
Caps are not to blame in this case.

Reply 6 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
mr.cat wrote on 2021-01-19, 04:04:

Yes, I did omit one Voodoo step "do the little chicken-dance, run around the house and let out a little shriek three times", I didn't think it would make a very big difference though 😉
Regarding USB, you could also try booting without any USB devices and see if that helps. These are somehow UEFI related issues, but I don't know the actual root cause.
Caps are not to blame in this case.

I’m not entirely sure what uefi is tbh. But I remember a few settings in the bios about it. Do you think those may help or are those settings for after post, like only for booting the OS

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 7 of 15, by mr.cat

User metadata
Rank Member
Rank
Member
Sphere478 wrote on 2021-01-19, 04:10:
mr.cat wrote on 2021-01-19, 04:04:

Yes, I did omit one Voodoo step "do the little chicken-dance, run around the house and let out a little shriek three times", I didn't think it would make a very big difference though 😉
Regarding USB, you could also try booting without any USB devices and see if that helps. These are somehow UEFI related issues, but I don't know the actual root cause.
Caps are not to blame in this case.

I’m not entirely sure what uefi is tbh. But I remember a few settings in the bios about it. Do you think those may help or are those settings for after post, like only for booting the OS

Perhaps you have heard about the end of BIOS as we know it? UEFI is "simply" the modern replacement for BIOS and Intel is going to pull the plug on traditional BIOS at some point.
The rationale being, we can't use 70s tech anymore because it's not safe (which is a fair point).
It does seem like hugely complicated beast, but I guess that's what we got now.

Well, this is all guesswork, don't take it too seriously. It's just the problems you're having sound awful lot like some of the old devices not getting along with UEFI.
About BIOS settings, there's probably nothing of much interest there. Maybe PCI and/or USB related settings would be the ones to consider (I see you tried with PCI latency already).
Hmm, GT 610 is at least supposed to be UEFI-capable. But that doesn't necessarily mean that it actually has UEFI firmware installed. Maybe that's something to look into.
EDIT: This little program can be used to parse roms to see if they contain EFI: https://github.com/awilliam/rom-parser

Sphere478 wrote on 2021-01-19, 03:41:

The 6200 works fine, the usb2.0 works fine and the sound blaster works 50% of the time the rest of them..... omg, I may have just figured something out, the usb3.0 (which has a bridge chip), the 610 (which has one also), the pci express (which is a bridge chip) all have bridge chips..

Why wouldn’t bridge chips work in this motherboard?

That's kind of why I asked about make and model. There are folks here who actually know about them there chips, u know...

I have a neat little theory which could explain the symptoms. It may be probably totally bonkers but I am presenting it here anyway, humor me 😁
(And please shoot it down if you can and show where I'm going wrong...)
I'm thinking these devices require a piece of init code that is simply not present or differs in some crucial way from what it should be. Maybe the UEFI is trying to init them with some generic code that doesn't work.
It could also be just that there is some timing issue caused by the bridging that messes up the init.

Last edited by mr.cat on 2021-01-19, 15:24. Edited 1 time in total.

Reply 8 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
mr.cat wrote on 2021-01-19, 04:29:
Perhaps you have heard about the end of BIOS as we know it? UEFI is "simply" the modern replacement for BIOS and Intel is going […]
Show full quote
Sphere478 wrote on 2021-01-19, 04:10:
mr.cat wrote on 2021-01-19, 04:04:

Yes, I did omit one Voodoo step "do the little chicken-dance, run around the house and let out a little shriek three times", I didn't think it would make a very big difference though 😉
Regarding USB, you could also try booting without any USB devices and see if that helps. These are somehow UEFI related issues, but I don't know the actual root cause.
Caps are not to blame in this case.

I’m not entirely sure what uefi is tbh. But I remember a few settings in the bios about it. Do you think those may help or are those settings for after post, like only for booting the OS

Perhaps you have heard about the end of BIOS as we know it? UEFI is "simply" the modern replacement for BIOS and Intel is going to pull the plug on traditional BIOS at some point.
The rationale being, we can't use 70s tech anymore because it's not safe (which is a fair point).
It does seem like hugely complicated beast, but I guess that's what we got now.

Well, this is all guesswork, don't take it too seriously. It's just the problems you're having sound awful lot like some of the old devices not getting along with UEFI.
About BIOS settings, there's probably nothing of much interest there. Maybe PCI and/or USB related settings would be the ones to consider (I see you tried with PCI latency already).
Hmm, GT 610 is at least supposed to be UEFI-capable. But that doesn't necessarily mean that it actually has UEFI firmware installed. Maybe that's something to look into.

Sphere478 wrote on 2021-01-19, 03:41:

The 6200 works fine, the usb2.0 works fine and the sound blaster works 50% of the time the rest of them..... omg, I may have just figured something out, the usb3.0 (which has a bridge chip), the 610 (which has one also), the pci express (which is a bridge chip) all have bridge chips..

Why wouldn’t bridge chips work in this motherboard?

That's kind of why I asked about make and model. There are folks here who actually know about them there chips, u know...

I have a neat little theory which could explain the symptoms. It may be probably totally bonkers but I am presenting it here anyway, humor me 😁
(And please shoot it down if you can and show where I'm going wrong...)
I'm thinking these devices require a piece of init code that is simply not present or differs in some crucial way from what it should be. Maybe the UEFI is trying to init them with some generic code that doesn't work.

Here is a pic of them

Attachments

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 9 of 15, by mr.cat

User metadata
Rank Member
Rank
Member

OK...looks like pretty standard stuff to me. I think I even might have one similar to the green one myself, somewhere.
EDIT: Yep, found it. But looks like I can't replicate your setup at all, so it doesn't really help.

Last edited by mr.cat on 2021-01-19, 08:23. Edited 1 time in total.

Reply 10 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
mr.cat wrote on 2021-01-19, 07:22:

OK...looks like pretty standard stuff to me. I think I even might have one similar to the green one myself, somewhere.

yeah. Idono, as long as they work in the socket 7 board I’ll be happy. Lol that i5 is my only pci board at the moment. It’s just so weird that they don’t work in the i5 board. Never had a problem like that before

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 11 of 15, by mr.cat

User metadata
Rank Member
Rank
Member
Sphere478 wrote on 2021-01-18, 22:31:
LewisRaz wrote on 2021-01-18, 15:59:

Looks like an AMI bios and that gives this: "94 Memory size set"

Any idea what it means and why specific pci cards would have anything to do with memory

I may have found an explanation to that code. With UEFI, status codes are still written to port 0x80, but in a different way (see page 13):
http://www.uefi.org/sites/default/files/resou … In_UEFI_101.pdf

Basically code 0x94 should be interpreted as 0x0205 and that means "PCI bus enumeration". Makes a bit better sense I suppose...

Also, I found another board that has a similar problem:
https://rog.asus.com/forum/showthread.php?748 … ode-94-Problems

Reply 12 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
mr.cat wrote on 2021-01-20, 02:28:
I may have found an explanation to that code. With UEFI, status codes are still written to port 0x80, but in a different way (se […]
Show full quote
Sphere478 wrote on 2021-01-18, 22:31:
LewisRaz wrote on 2021-01-18, 15:59:

Looks like an AMI bios and that gives this: "94 Memory size set"

Any idea what it means and why specific pci cards would have anything to do with memory

I may have found an explanation to that code. With UEFI, status codes are still written to port 0x80, but in a different way (see page 13):
http://www.uefi.org/sites/default/files/resou … In_UEFI_101.pdf

Basically code 0x94 should be interpreted as 0x0205 and that means "PCI bus enumeration". Makes a bit better sense I suppose...

Also, I found another board that has a similar problem:
https://rog.asus.com/forum/showthread.php?748 … ode-94-Problems

Good sleuthing! 🤔 eh, I wonder if it wouldn’t just be easier to replace the board. Sounds like whatever this is is beyond my ability to fix?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 13 of 15, by mr.cat

User metadata
Rank Member
Rank
Member

Yeah, it doesn't seem like an easy fix. Is your BIOS (UEFI) up to date?

Sphere478 wrote on 2021-01-18, 09:57:

... mouse enabled bios ...

That's one way to determine that you're using UEFI. Mouse & graphical mode I mean.
Most of the traditional BIOSes use text mode (with some exceptions).

Reply 14 of 15, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
mr.cat wrote on 2021-01-20, 10:39:
Yeah, it doesn't seem like an easy fix. Is your BIOS (UEFI) up to date? […]
Show full quote

Yeah, it doesn't seem like an easy fix. Is your BIOS (UEFI) up to date?

Sphere478 wrote on 2021-01-18, 09:57:

... mouse enabled bios ...

That's one way to determine that you're using UEFI. Mouse & graphical mode I mean.
Most of the traditional BIOSes use text mode (with some exceptions).

I shit you not, I once came across I think a 486 that had a mouse enabled bios that looked like windows 3.1 ish.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 15 of 15, by mr.cat

User metadata
Rank Member
Rank
Member
Sphere478 wrote on 2021-01-20, 15:40:
mr.cat wrote on 2021-01-20, 10:39:
Yeah, it doesn't seem like an easy fix. Is your BIOS (UEFI) up to date? […]
Show full quote

Yeah, it doesn't seem like an easy fix. Is your BIOS (UEFI) up to date?

Sphere478 wrote on 2021-01-18, 09:57:

... mouse enabled bios ...

That's one way to determine that you're using UEFI. Mouse & graphical mode I mean.
Most of the traditional BIOSes use text mode (with some exceptions).

I shit you not, I once came across I think a 486 that had a mouse enabled bios that looked like windows 3.1 ish.

Yes I beliiieve that...
Some of those old ones even had some neat EGA trick that gave you a mouse pointer in a otherwise text-based BIOS. Very slick for the time.