VOGONS


First post, by Charleston

User metadata
Rank Newbie
Rank
Newbie

Got this board on ebay, identified it, thought I was clean and good to go until I tried to setup my jumpers and pop in the dx4-100. I get no boot, no beeps, nothing. I spent hours working with the thing and as a last ditch effort set all the jumpers to as they were when I received the board (using an image from the listing) and popped the 50mhz sx2 back into place. Many of the jumpers were set under "Cyrix CPU" and not dx2 or other as I imagined they would be but apparently this worked.

(Note that when I replace the sx2 with the dx4-100 I get a boot but it locks up after several seconds)

https://www.elhvb.com/webhq/models/486vlb3/v4p895v1.htm

Fast forward and I've gotten to the point where I can change almost every jumper to the proper locations after much trial and error and get a boot from the dx4-100 (using the 50mhz clock which causes the system to freeze and give a 7 beep post after a minute) and the only jumpers I can not change whatsoever without going back to a black screen of nothingness are:

- JP1 (switches to 3V, tested the regulator and I get an output of 3.6 through my dmm so I'm at a loss here)
- JP43 (Switches the clock. Every manual says something different but it's just 2x, 2.5x, and 3x for dx4s)
- JP 38 and 39 (These two set the clock speed. Typically both are closed which gives 50Mhz, but with even one of them open to give either 25mhz or 33mhz I get absolutely nothing)

Changing any of these in any way at all leads me to nothing. As far as I'm aware I've done every possible config.

I'd really appreciate any help at all. You would think this board would be easy when it has everything labeled all over it.

Edit: I don't actually know if the 50mhz is a sx2 or a dx.

Reply 2 of 43, by Horun

User metadata
Rank l33t
Rank
l33t

What is the actual markings on your DX4-100 cpu ? All of my DX4-100's run fine as 33Mhz x3 (and do not overclock the memory, ISA or VLB slots on any board).

Hate posting a reply and have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. 🤣 Second computer a 286 12Mhz with real IDE drive ! After that came 386, 486, Pentium, P.Pro and everything after....

Reply 3 of 43, by jakethompson1

User metadata
Rank Oldbie
Rank
Oldbie
Charleston wrote on 2021-07-21, 18:27:

Gonna bump since I posted this late last night and it hasn't gotten any views.

maybe take picture of your setup so everyone can see the video card, ram, jumpers, etc?

Reply 4 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie

I apologize, should have definitely been more specific. Note that the manual I posted is a bit off of what's actually marked, and there seems to be another jumper manual online that's got different markings.

One picture is with and another without flash to make it easier to see details.

This board is the biggest pain. I feel like I've tried every combination. Video card is a Trident 9200cxr

As I said, this PC does post properly with the 50mhz in with weird jumper settings. The 30 pin was pulled out of my 386 and the video card is known to be working as well, unless there's a jumper on it that can affect the system although the system won't even make a post beep regardless.

Edit: keep in mind system won't boot with any config of jp43.

Attachments

Reply 5 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie

Really hate doing this but I'm gonna bump again. I am at a complete loss. I know this board works, is the voltage regulator dead? I get 3.6V testing it when set to the 3v jumper. I'm just so confused but really want this working. I think I've tried almost every configuration of jumper settings at this point maybe.

Reply 6 of 43, by mkarcher

User metadata
Rank Oldbie
Rank
Oldbie
Charleston wrote on 2021-07-21, 06:20:

- JP1 (switches to 3V, tested the regulator and I get an output of 3.6 through my dmm so I'm at a loss here)

Did you measure the 3.6V without a CPU inserted? We had several users here on VOGONs with bad transistors in the 3V regulator that didn't manage to deliver enough current to run a CPU, even if the voltage was fine without load.

Reply 7 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie

No, I didn't. I'll definitely do that when I get back to it later. What should I be getting without a cpu? The same 3.6? And if not the transistors are bad inside the package?

Do you know a good replacement I can buy quickly? I can't read the model number in regulator and won't be home until after the weekend.

Reply 8 of 43, by mkarcher

User metadata
Rank Oldbie
Rank
Oldbie
Charleston wrote on 2021-07-24, 19:33:

No, I didn't. I'll definitely do that when I get back to it later. What should I be getting without a cpu? The same 3.6?

I'm sorry, it seems I wasn't clear enough. If you get 3.6V with a CPU inserted, I don't know why the board won't boot with the 486DX4. I guessed that maybe you measured the 3.6V without a CPU, which wouldn't be a sure sign that a CPU (if inserted later) gets enough power. You don't need to test without a CPU if you already tested with a CPU. The latter is the more conclusive test.

Reply 9 of 43, by jakethompson1

User metadata
Rank Oldbie
Rank
Oldbie

I don't have much advice either after seeing the pictures, other than to note that apparently this was known as a high quality board but with difficult jumpers.

https://www.redhill.net.au/b/b-95.html#895

Reply 10 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie
jakethompson1 wrote on 2021-07-24, 19:43:

I don't have much advice either after seeing the pictures, other than to note that apparently this was known as a high quality board but with difficult jumpers.

https://www.redhill.net.au/b/b-95.html#895

I appreciate the help regardless. You'd think with the jumpers stamped it'd be easy. I'm leaning towards the issue being the voltage regulator now since I can get it working with 5v cpus and not any 3v cpus.

Reply 11 of 43, by Doornkaat

User metadata
Rank Oldbie
Rank
Oldbie

Afaik 3.6V is within tolerance for regular DX4 CPUs and it is very possible they chose the higher voltage intentionally to help with stable operation of any 3.xV CPU and maybe overclocking. 3.6V should not cause damage to any 486 with heatsink and fan so I wouldn't worry about it.

The clock generator IC could be damaged if you get no activity with two of three settings.
At this point it sounds like an oscilloscope might help clear things up.

Reply 12 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie
Doornkaat wrote on 2021-07-24, 19:49:

Afaik 3.6V is within tolerance for regular DX4 CPUs and it is very possible they chose the higher voltage intentionally to help with stable operation of any 3.xV CPU and maybe overclocking. 3.6V should not cause damage to any 486 with heatsink and fan so I wouldn't worry about it.

The clock generator IC could be damaged if you get no activity with two of three settings.
At this point it sounds like an oscilloscope might help clear things up.

What should I be looking for? 50mhz doesn't post if the voltage is ever changed. I don't remember if there are any 3v 50mhz cpus.

Sounds like I'm setting up the big fat oscope again. Where should I be testing? Not super experienced with older mainboards.

Reply 13 of 43, by Doornkaat

User metadata
Rank Oldbie
Rank
Oldbie

I'm not 100% sure but I think you can set a DX4 to run at 2x multi so with some boards you should be able to run it at 50MHz FSB / 100MHz internal @3.xV.

I think the clock generator is the IC between CPU socket and TAG RAM chip. I can't read it though. Can you find the datasheet and confirm wether that's the clock generator?

Reply 15 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie
Doornkaat wrote on 2021-07-24, 19:58:

I'm not 100% sure but I think you can set a DX4 to run at 2x multi so with some boards you should be able to run it at 50MHz FSB / 100MHz internal @3.xV.

I think the clock generator is the IC between CPU socket and TAG RAM chip. I can't read it though. Can you find the datasheet and confirm wether that's the clock generator?

I'll take a look.

Reply 17 of 43, by mkarcher

User metadata
Rank Oldbie
Rank
Oldbie
Charleston wrote on 2021-07-24, 19:52:

What should I be looking for? 50mhz doesn't post if the voltage is ever changed. I don't remember if there are any 3v 50mhz cpus.

Sounds like I'm setting up the big fat oscope again. Where should I be testing? Not super experienced with older mainboards.

You said you can not get the board to post if you either change the voltage jumper or the multiplier jumper. Your DX4 is expected to work at 50x2 (100MHz), but is likely to fail at 50x2.5 (125MHz) and 50x3 (150MHz). It's not impossible that overclocking the processor to 125 or 150 will work for some time at 5V, but fail at 3V. JP43 should be a 3-pin jumper, with GND and +5V on the two outer pins and CLKMUL in the middle. If you don't have it plugged, you get x3. If you connect GND with the center pin, you get x2. If you connect +5V with the center pin, you get x2.5.

If the only jumper setting that works is x2, it makes some sense, because that means your processor is not overclockable. If the only working setting is different (e.g. x3), it's likely that your BIOS is incompatible with a DX4 (but happens to work at 150MHz for some strange reason).

Reply 18 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie

According to this that I found, the IMI chip is the generator.

Will look for the datasheet in a bit.

Attachments

Reply 19 of 43, by Charleston

User metadata
Rank Newbie
Rank
Newbie
mkarcher wrote on 2021-07-24, 20:27:
Charleston wrote on 2021-07-24, 19:52:

What should I be looking for? 50mhz doesn't post if the voltage is ever changed. I don't remember if there are any 3v 50mhz cpus.

Sounds like I'm setting up the big fat oscope again. Where should I be testing? Not super experienced with older mainboards.

You said you can not get the board to post if you either change the voltage jumper or the multiplier jumper. Your DX4 is expected to work at 50x2 (100MHz), but is likely to fail at 50x2.5 (125MHz) and 50x3 (150MHz). It's not impossible that overclocking the processor to 125 or 150 will work for some time at 5V, but fail at 3V. JP43 should be a 3-pin jumper, with GND and +5V on the two outer pins and CLKMUL in the middle. If you don't have it plugged, you get x3. If you connect GND with the center pin, you get x2. If you connect +5V with the center pin, you get x2.5.

If the only jumper setting that works is x2, it makes some sense, because that means your processor is not overclockable. If the only working setting is different (e.g. x3), it's likely that your BIOS is incompatible with a DX4 (but happens to work at 150MHz for some strange reason).

Issue with this is that a dx4 is 33.3x3 not 50x2 to reach 100mhz. The dx4 only works when set specifically to 50mhz as it was setup for the sx or dx2 50. 3x clock will boot for moments before crashing in bios.

With the original settings with the 50mhz cpu in place, it was set to 3x which would be 75hz. I think the jumpers are wrong. I'm just highly confused.