VOGONS


First post, by Meoseko

User metadata
Rank Newbie
Rank
Newbie

Dear VOGONS people, I come in the need of help.

I've been lurking these forums for the past couple months, mainly through google search results when trying to find solutions for various problems while building some old PCs.

Eventually, both of my machines were in a satisfactory state, at which point something in the floppy drive of my DOS machine fizzled, rendering the machine dead. Since then, I've been trying to figure out what parts of the machine were affected by that in an effort to get everything back to how it was. Here are the specs of the machine in question:

Motherboard: MB-8500TVX-A Ver 2.3 (Including a mod to socket the Dallas RTC for easy swapping as well as a battery holder)
CPU: Intel Pentium MMX SL27S
PSU: AT
RAM: Unsure since it came with the original MB but here is a picture.
GPU: Matrox 618-02 REV: A (MGA-MYST/21)

Of course the PC also has hard drives, a sound card and a 3.5" disk drive as well as a CD drive, but since the problem persists even if I remove all of these things, I would already be happy in getting the pc to turn on without having any of these hooked up.

Originally I only ordered a new motherboard, but when taking out the CPU of my old motherboard to put it into the new one, I noticed that it had several spots on it that looked like it had popped there. So, I ordered a replacement for that CPU as well, put that new CPU into the new motherboard along with the RAM and GPU from the old motherboard, configured the jumpers on the motherboard just like on the old one and hooked that up to my old PSU.

Now, here is where I'm at a loss:
Before that floppy drive mishap, the PC started without any problems.
Now, however, the PC only turned on after several attempts. If it wouldn't turn on, there were no beeps, no picture. The hard drives and CPU fan would spin up, the CD drive would be able to pop in and out, but that's it. After several more attempts, the PC would turn on with a picture, but the system would freeze before it counted up the RAM. (The RAM count showed something like "KB 0 KB".) After restarting once more, it would then boot into DOS just fine and run well through several games and restarts for well over 6 hours.
On the next day, same deal. Had to try a while before it would turn on. On the first turn on, it would freeze while I was in the Bios and after a restart, it would work fine for the rest of the day.
Today, it wouldn't turn on at all, no matter how often I tried.

My biggest hunch is that the CPU jumpers seem to not be configured correctly, but I didn't question it since the old one ran like this and this one also seems to sometimes run like this, and I just don't know any better. Here is what I've been able to gather about that:

Here's the manual: https://stason.org/TULARC/pc/motherboards/B/B … Pentiu-211.html
Here's the cpu: http://www.cpu-galerie.de/html/intelpentiump55-fv233.html

My jumper configurations are:
Flash Memory Voltage: +12V type Flash Memory used (Other option, +5V) (JP3: 2-3 Closed)
Cache RAM: 512 KB Cache RAM size (JP7: Closed)
CPU Voltage: Doesn't look like a valid setup in the manual, but that's how I got the original board and it ran that same model CPU like that. The two options here are single voltage or dual voltage, either selected by JP6 1- 2 Closed + JP11 1-4 Closed or JP6 2-3 Cosed + JP11 1-4 Open. My setup is JP6 2-3 Closed, JP11 1-4 Closed. Possible cause of error?
CPU I/O Voltage: 3.5V (other option: 3.45V) (JP12: Open)
CPU Voltage Select for dual Voltage CPU: 2.9V (other option: 2.6V) (JP14: Open)
CPU Clock Select: 100 Mhz (that's not what it showed when the PC was turned one as far as I know) (JP5: 1-2 Open, 3-4 Closed; JP8 Open; JP9 Open)

Now the CPU specs tell me the core voltage is 2.8V +- 5%, the I/O voltage is 3,3V +- 10% and it's also a dual voltage cpu. Wouldn't that mean these voltage jumpers are setup all wrong? Could this be the problem? If yes, why did this configuration work without problems with the old CPU/MB that were exactly the same model?

Last thing I tried was to put the RAM from slot 3 and 4 into 1 and 2 but that didn't do anything either. I hope it's just me setting up the jumpers in a configuration that supplies the CPU with so little power that it only turns on sometimes or something like that. Hope anybody has any ideas. Thanks so much in advance!

Reply 1 of 27, by dionb

User metadata
Rank l33t++
Rank
l33t++
Meoseko wrote on 2020-08-10, 20:12:
[...] […]
Show full quote

[...]

My biggest hunch is that the CPU jumpers seem to not be configured correctly, but I didn't question it since the old one ran like this and this one also seems to sometimes run like this, and I just don't know any better. Here is what I've been able to gather about that:

Here's the manual: https://stason.org/TULARC/pc/motherboards/B/B … Pentiu-211.html
Here's the cpu: http://www.cpu-galerie.de/html/intelpentiump55-fv233.html

My jumper configurations are:
Flash Memory Voltage: +12V type Flash Memory used (Other option, +5V) (JP3: 2-3 Closed)
Cache RAM: 512 KB Cache RAM size (JP7: Closed)
CPU Voltage: Doesn't look like a valid setup in the manual, but that's how I got the original board and it ran that same model CPU like that. The two options here are single voltage or dual voltage, either selected by JP6 1- 2 Closed + JP11 1-4 Closed or JP6 2-3 Cosed + JP11 1-4 Open. My setup is JP6 2-3 Closed, JP11 1-4 Closed. Possible cause of error?
CPU I/O Voltage: 3.5V (other option: 3.45V) (JP12: Open)
CPU Voltage Select for dual Voltage CPU: 2.9V (other option: 2.6V) (JP14: Open)
CPU Clock Select: 100 Mhz (that's not what it showed when the PC was turned one as far as I know) (JP5: 1-2 Open, 3-4 Closed; JP8 Open; JP9 Open)

Now the CPU specs tell me the core voltage is 2.8V +- 5%, the I/O voltage is 3,3V +- 10% and it's also a dual voltage cpu. Wouldn't that mean these voltage jumpers are setup all wrong? Could this be the problem? If yes, why did this configuration work without problems with the old CPU/MB that were exactly the same model?

Several things going on here:
- multiplier of P233MMX is 3.5x. Intel re-used the 1.5x setting to allow 233MHz to be used without requiring a third multiplier pin. So to get 233MHz you use the exact same setting as 100MHz, a P54(C) interprets it as 1.5x giving 100MHz, a P55C interprets it as 3.5x giving 233MHz. This is correct.
- voltage of 2.9V/3.5V is overvolted, but not excessively so. If your only dual voltage options are 2.6V and 2.9V, 2.9V is the better match for stability (within the +-5%). I'd recommend setting VIO to 3.4V, so JP14 closed, as that is already above nominal, if in-spec (as is 3.5V)
- JP11 1-4 is not described in TH99, so can't say what that should do. It should be open for dual voltage in any event. Possibly 1-4 is effectively open (linking two ground pins)

Last thing I tried was to put the RAM from slot 3 and 4 into 1 and 2 but that didn't do anything either. I hope it's just me setting up the jumpers in a configuration that supplies the CPU with so little power that it only turns on sometimes or something like that. Hope anybody has any ideas. Thanks so much in advance!

You're supplying the CPU with marginally too much (dual voltage) or far too much (single voltage) power, so too little power is definitely not the problem.

Reading your story about the FDD shorting out, I'd strongly suspect the PSU. Have you tried a different PSU yet in this system?

Reply 2 of 27, by Doornkaat

User metadata
Rank l33t
Rank
l33t

That's a lot to take in. 😉
First, something on your FDD fried, drive motors work but no (or irregular) system activity. - Most chips on that system use the +5V rail of the PSU while drive motors use +12V. The PSU wasn't changed after the FDD incident and is very old. From those clues I would start troubleshooting by checking voltages and ripple on the PSU.

As for jumpers (I'll admit I didn't fully understand what your current settings are😉) you should configure them according to the manual first before doing any other troubleshooting. Maybe they were misconfigured. The P1MMX need split voltage, ~2.8V core / ~3.3V io. 2.9V core / 3.45V io are fine too. I have had multiple systems where Pentium MMXs had been run completely out of spec but were working ok. In my opinion those CPUs are rather tolerant to wrong voltages but that may change over time as the hardware ages.

Hope this helps! 🙂👍

Reply 3 of 27, by Horun

User metadata
Rank l33t++
Rank
l33t++

Agree with both of you ! Set it for Intel 100Mhz. Set voltages to 3.4v and 2.9v. Set cpu type as P55C on Jp6 and Jp11. Try a different PSU.
Try a lessor cpu like a Pentium 100 (you will have to set jumpers as P54C, single 3.4v voltage) just to make sure the board boots at all, possible the BIOS does not like the 233MMX.
IF you can get it to boot again: take a POST screen picture with the BIOS string at the bottom.

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. https://archive.org/details/@horun

Reply 4 of 27, by dionb

User metadata
Rank l33t++
Rank
l33t++
Doornkaat wrote on 2020-08-10, 22:03:

[...]
In my opinion those CPUs are rather tolerant to wrong voltages but that may change over time as the hardware ages.

P55C is possibly the most over-engineered and tolerant CPU ever made.

At work some ancient teletext-related system died after almost 20 years solid service. Turned out to be a fried P166MMX. Why? Because the heatsink had fallen off, almost certainly at time of installation. This P166MMX had survived almost 20 years of running without a heatsink. The black OPGA had been baked reddish brown after that time. I keep it as a totem on my desk.

They also almost invariably worked on single-voltage motherboards, being fed 3.3V-3.5VCore where 2.8V +-5% is specced. Inversely you could generally undervolt them to 2.5V or sometimes even 2.2V to save power and so heat production, allowing very compact systems to be passively cooled. Or just keep it at stock voltage and overclock massively. I've had them running at 3.5x100MHz with 2.8VCore.

If any CPU could handle slightly off voltage settings, it's a Pentium MMX. So no, that's about the last place I'd look for the cause of the problems. Start with the PSU.

Reply 5 of 27, by Meoseko

User metadata
Rank Newbie
Rank
Newbie

Thank you all for the insightful responses. I didn't expect anyone to take the time to read all that, so I'm very thankful to all of you. Here is what I've tried in response. I powered the system on between each of those steps.

Tried to turn it on to see if the problem was still persisting. Same as before. Fans would spin, HDDs would spin, motherboard would do nothing.
Put the RAM back into 3/4 because that's how I had the system working before and after the incident.
Removed all IDE cables and disconnected the PSU from all hard drives and disk/CD drives, leaving only the connector to the motherboard.
Disconnected PSU from motherboard, connected a different AT PSU to the motherboard.
Opened JP11 (To select dual voltage CPU)
Closed JP12 (I/O voltage from 3.5V to 3.45V)
Connected the original PSU to the motherboard again.
Replaced the CPU with an SL27J.
Set CPU clock jumpers to be CPU speed 200Mhz, Bus Clock & Multiplier 66MHz x 3.
Switched to alternative PSU again.
Switched CPU to SL27S again.
Put CPU clock jumpers back to original position.
Switched to original PSU again.

I also found the exact manual that I've been using in printed form: https://www.epanorama.net/sff/Computer/Mother … Motherboard.pdf

At this point I want to assume that either both of these CPUs/PSUs are faulty or neither of those are the problem. I have 2 ATX PSUs as well but don't know how I would be able to hook these up to this MB for testing. I also still have the old MB in which the old CPU pretty much exploded. Don't feel like it's safe to grab that one for testing though, since the CPU socket in that one looks like this.

Reply 6 of 27, by Doornkaat

User metadata
Rank l33t
Rank
l33t

EDIT: I'm still not certain I understand your current jumper settings. Can you please post a picture of the board with all jumpers visible?

Further troubleshooting ideas:
Do you have a multimeter to check voltages on the PSU(s)?
->If so, reduce the setup to PSU, motherboard and CPU (no RAM) and test the voltages delivered by the PSU.
->If not, go and buy one. 😉
Do you have a POST diagnostics card?
->If so connect it and see what it reads.
->If not order one and for now substitute via PC speaker:
Connect the PC speaker to the motherboard and use the PSU, motherboard and CPU combo again to see (or hear 😉 ) wether no RAM is reported over the speaker.
->If it is reported continue to add RAM and see wether it still reports no RAM or now reports no VGA adaptor.
-->If it reports no VGA, add a VGA card, see if it POSTs, then add a floppy drive, try to boot from floppy and so on.
-->If it keeps reporting no RAM try different RAM sticks.
--->If it keeps on reporting no RAM try cleaning the contacts and reflashing the BIOS.
---->If it still keeps on reporting no RAM chances are your motherboard is damaged.
->If no RAM isn't reported, leave it running a bit. Some motherboards will report unstable voltages after about a minute of not POSTing. This may indicate a PSU or motherboard failure. If your PSU tests fine the motherboard is likely damaged.
->If the speaker doesn't beep at all try another known good CPU.
-->If it still doesn't beep try reflashing the BIOS.
--->If it still doesn't beep chances are your motherboard is dead.

Reply 7 of 27, by Meoseko

User metadata
Rank Newbie
Rank
Newbie

Thank you very much! I'm going out to the get a multimeter from the store now. Say, if I'm checking the voltages with the PSU connected to the motherboard with a CPU in it, where would I take the voltages from? I only found tutorial videos where people put the multimeter directly to the PSU with a paperclip and stuff.

Reply 9 of 27, by Doornkaat

User metadata
Rank l33t
Rank
l33t
Meoseko wrote on 2020-08-11, 12:55:

Here are the jumpers: https://i.imgur.com/QoeM74f.jpg
And without markings in case I obscured something important: https://i.imgur.com/5IKzVA3.jpg

The CPU is currently configured for 2.5x multiplier / 66MHz FSB. Some P1MMX CPUs will not operate when configured for certain multipliers. Please set JP9 open (keeping JP8 closed) for P1MMX 200MHz (3x multi) or JP8 open & JP9 open for 233MHz (1.5x/3.5x multi).
Also the current CPU voltage setting is 2.6V core. (JP12 closed / JP14 open) Please change JP12 to open (keeping JP14 open) for 2.9V core/3.5V io.
Overall your settings for P1 MMX CPUs should be:
JP4 OPEN
JP5 1-2 OPEN & 3-4 CLOSED (66MHz FSB)
JP6 2-3
JP7 OPEN
JP8 CLOSED for 200MHz (3x multi) /// OPEN for 233MHz (1.5/3.5x multi)
JP9 OPEN for 200MHz (3x multi) /// OPEN for 233MHz (1.5/3.5x multi)
JP11 1-2 OPEN & 3-4 OPEN
JP12 OPEN
JP14 OPEN

If this boots we can try reducing i/o voltage by changing JP14 to closed if you wish so.

If this doesn't help I also noticed JP3 is set for neither 12V nor 5V. I don't think this is required outside of flashing the EEPROM but it's worth a try setting it to the 12V setting (the seller used this setting in his photos) and seeing wether this helps.

As for measuring AT PSU voltages: +5V and +12V can easily be measured at the drive connectors. black is ground, red is +5V, yellow is +12V.
For measuring -5V and -12V you can either probe the backside of the board or stick a thin piece of solid wire into the AT connector from the cable side so it makes contact with the metal pieces in the connector. You can use the drive connectors as well as the keyboard connector and other places for ground reference.

Of course you're doing all this at your own risk. 😉

Reply 10 of 27, by Meoseko

User metadata
Rank Newbie
Rank
Newbie

Changed the jumpers! These should be exactly as you described now: https://imgur.com/a/Seh1L9y
Also, JP3 did have a jumper on it already (2-3) but it's black so it looked like it didn't. I tried to get it at an angle this time so it's more visible.

Unfortunately, nothing has changed. No boot (no beeps, no picture), even after waiting for several minutes on either PSU.

I got the multimeter now but I'll probably have to watch a bunch of Youtube tutorials before knowing how exactly to work that thing.

What I find weird is that this is exactly how the system behaved while I was still trying to get it to work after the floppy drive mishap. I was switching PSUs, CPUs and jumpers in an effort to somehow get it to run which it eventually did as I had described, only to then not work again without me changing anything inbetween. I'm kind of just waiting for it to suddenly work again only for it to go back to not working shortly after, regardless of the changes we are making.

Reply 12 of 27, by Doornkaat

User metadata
Rank l33t
Rank
l33t

Oh, right - I simply missed that jumper.
Have you tested with the 200MHz CPU as well (configured for 200MHz of course)?
Maybe one is bad?

Otherwise I'm afraid there are failing parts in this setup, and without a set of known good parts (or better equipment) troubleshooting is a pain in the ass.

You bought from a German auction. Are you located in Germany? If so, what general area if I may ask?

Reply 13 of 27, by Meoseko

User metadata
Rank Newbie
Rank
Newbie
kingcake wrote on 2020-08-11, 15:39:

You might want to invest in one of those $7 post diag cards from ebay

Would this be something? https://www.amazon.de/4-stellige-Analyzer-Com … 97167450&sr=8-1

All I could find on Ebay would ship from China.

Reply 14 of 27, by Meoseko

User metadata
Rank Newbie
Rank
Newbie
Doornkaat wrote on 2020-08-11, 15:40:
Oh, right - I simply missed that jumper. Have you tested with the 200MHz CPU as well (configured for 200MHz of course)? Maybe on […]
Show full quote

Oh, right - I simply missed that jumper.
Have you tested with the 200MHz CPU as well (configured for 200MHz of course)?
Maybe one is bad?

Otherwise I'm afraid there are failing parts in this setup, and without a set of known good parts (or better equipment) troubleshooting is a pain in the ass.

You bought from a German auction. Are you located in Germany? If so, what general area if I may ask?

I'm from Berlin! Will get back after checking the 200MHz.

Reply 17 of 27, by Meoseko

User metadata
Rank Newbie
Rank
Newbie
kingcake wrote on 2020-08-11, 19:24:

Copy! Just getting that POST diagnostic card then.