PC@LIVE wrote on 2025-10-31, 18:19:

Looking at the images, I can tell you that the CPU is blocked, you can only use two multi, 2.5X and 3X, the 2.5X one definitely […]
Show full quote

sunkindly wrote on 2025-10-31, 17:51:

Photos below, but in doing some more testing...it seems like the jumpers don't do anything at all to change it from 2.5x. […]
Show full quote

PC@LIVE wrote on 2025-10-31, 09:51:

Hello, if I understand correctly you have an S.7, but the CPU goes to a maximum of 166 MHz, so first of all 😨 I would check what is written on the CPU, for example Pentium XXX MMX SXXXX (the X's indicate what you will find written, they are important), already from there it would be possible to better understand the problem.
Then if maybe 🤔 you can take some photos, and attach them here, possibly motherboard and CPU, or if you can jumper zone (multiX), it would be possible to compare your settings with those of the manual (if you have one), but in any case even in the absence of the manual, it is possible to do some tests, to see to set the jumpers on 3X, currently it should be on 2.5X, if instead it is on 3X and does not go to 200, speaking of Intel CPU, you could have a 166 MHz CPU with multiplier blocked to go up.
If so, the only way to increase the frequency is to raise the FSB, if that board allows it, you can switch to 75 MHz, you get 188 MHz that the BIOS could interpret as 180 or 200 MHz, maybe that's how they showed you the 200 MHz.

Photos below, but in doing some more testing...it seems like the jumpers don't do anything at all to change it from 2.5x.

The attachment r529multjp.png is no longer available

So JP16 and JP17 seem to set the fsb, but for JP19, JP20, and JP22 no matter what I set them to...it didn't change from 2.5x at 166MHz. Even setting it to something "normal" like 133MHz at 2x, nothing changed. The only thing that seemed to affect change was changing the FSB through JP16-17. But unfortunately, it doesn't support more than 66MHz.

The attachment IMG_7685.jpg is no longer available

The attachment IMG_7682.jpg is no longer available

The attachment IMG_7683.jpg is no longer available

The attachment IMG_7684.JPG is no longer available

The silkscreen doesn't mention anything about P55C or a 3x multiplier, but it has the jumpers to set dual voltage (which I did) and since it's POSTing I'm assuming that's actually working. So I'm not sure...

Looking at the images, I can tell you that the CPU is blocked, you can only use two multi, 2.5X and 3X, the 2.5X one definitely works, while the 3X one doesn't seem to.
But it could be something else, first, you can use a software to see the actual frequency, it could for example be a dated BIOS that makes you see only 166 MHz, when in reality it is 200.
One that you can easily find is CHKCPU, it works with DOS, or if you have Windows 95-98 you can try CPU-Z, both can show the actual frequency.
For the rest it's a motherboard that's fine, if you don't need to update it, that is, you want to use it as it is, and according to what it says, the jumpers are set up in the right way.
Perhaps the BIOS version is the factory version, and it is not accurate in reading the frequency (?).

I was hoping that might be the case but sadly CHKCPU and Speedsys both report 166ish...maybe I have some kind of early revision or is it the CPU?

BIOS is updated to the latest I found.

Either way I have a QDI Titanium 1E on the way so we'll what happens with that one 😮‍💨

sunkindly wrote on 2025-10-31, 18:27:

I was hoping that might be the case but sadly CHKCPU and Speedsys both report 166ish...maybe I have some kind of early revision […]
Show full quote

PC@LIVE wrote on 2025-10-31, 18:19:

Looking at the images, I can tell you that the CPU is blocked, you can only use two multi, 2.5X and 3X, the 2.5X one definitely […]
Show full quote

sunkindly wrote on 2025-10-31, 17:51:

Photos below, but in doing some more testing...it seems like the jumpers don't do anything at all to change it from 2.5x. […]
Show full quote

Photos below, but in doing some more testing...it seems like the jumpers don't do anything at all to change it from 2.5x.

The attachment r529multjp.png is no longer available

So JP16 and JP17 seem to set the fsb, but for JP19, JP20, and JP22 no matter what I set them to...it didn't change from 2.5x at 166MHz. Even setting it to something "normal" like 133MHz at 2x, nothing changed. The only thing that seemed to affect change was changing the FSB through JP16-17. But unfortunately, it doesn't support more than 66MHz.

The attachment IMG_7685.jpg is no longer available

The attachment IMG_7682.jpg is no longer available

The attachment IMG_7683.jpg is no longer available

The attachment IMG_7684.JPG is no longer available

The silkscreen doesn't mention anything about P55C or a 3x multiplier, but it has the jumpers to set dual voltage (which I did) and since it's POSTing I'm assuming that's actually working. So I'm not sure...

Looking at the images, I can tell you that the CPU is blocked, you can only use two multi, 2.5X and 3X, the 2.5X one definitely works, while the 3X one doesn't seem to.
But it could be something else, first, you can use a software to see the actual frequency, it could for example be a dated BIOS that makes you see only 166 MHz, when in reality it is 200.
One that you can easily find is CHKCPU, it works with DOS, or if you have Windows 95-98 you can try CPU-Z, both can show the actual frequency.
For the rest it's a motherboard that's fine, if you don't need to update it, that is, you want to use it as it is, and according to what it says, the jumpers are set up in the right way.
Perhaps the BIOS version is the factory version, and it is not accurate in reading the frequency (?).

I was hoping that might be the case but sadly CHKCPU and Speedsys both report 166ish...maybe I have some kind of early revision or is it the CPU?

BIOS is updated to the latest I found.

Either way I have a QDI Titanium 1E on the way so we'll what happens with that one 😮‍💨

Well you can always make some attempts, try selecting the other multi, if you haven't done it before, that is, the 1.5X/3.5X and 2.0X, I don't think the CPU doesn't work with the 3X multi, but by trying it in the QDI you should have confirmation, whether it works at 200 MHz (or not).
The only thing that could be there, is a hw failure, I noticed some SMDs near the jumper of the multi, maybe if it is broken, it could deceive the board, making you believe that the jumper is open, instead of closed.

So I got back to that Hercules 9600 Pro. Had a bit of time the last days.

I now swapped / reballed all memory ICs on memory channel A, both ranks - the error does not really disappear nor change, even when mixing the memory chips.

The card sometimes posts, sometimes doesn't. When it posts, we get a pretty picture:

R3MEMID isn't happy either:

1Failing bits in test [1] Fill
2errcount: 17616237
3MDA0: 5 7 11 12 14
4MDB0: 11
5MDA1: 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
6Failing bits in test [2] Data line toggle (GUI)
7errcount: 5772202
8MDA0: 5 7 11 12 14
9MDB0: 11
10MDA1: 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
11Failing bits in test [3] RW page (GUI)
12errcount: 8663668
13MDA0: 5 7 11 12 14 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
14MDB0: 11
15MDA1: 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
16Failing bits in test [4] RW channel (GUI)
17errcount: 5672858
18MDA0: 5 7 11 12 14
19MDB0: 11
20MDA1: 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63

So I checked the data lines to check the bits with the scope.. I just wanted to see if there's activity or not.

So here we have it.. some data lines are just plain "dead" - for everything from bits 32-63 I assume it has to do with a clock signal missing or strobe signal missing. Hard to check, since these traces go directly to the memory ICs. The upper picture is a good data line, the lower is a bad one just floating at VTT.

Hmm.. I put pressure on the GPU where the data lines are - no change. Yikes...

Then I cleaned up the GPU to check it and whoops.. someone crumbled the cookie corners? I checked an older picture I had from the card: Same.. And I didn't notice since I just repasted it quickly before testing. Ouch..

I've ordered some RV350s now.. let's see how it goes when they arrive.

If I get this card to run again, I've got stories to tell I guess..

Restored an AT&T Globalist 200 Laptop to near factory.

Mostly yellowed keyboard melted floppy belt and missing battery latch. Got lucky with a 30 EUR ebay well rusted 200S donor and upgraded RAM from 4MB to 8MB plus took its keycaps, and the 486 DX4 100MHZ in it also worked but out of concern for thermals I left it out of this build.

Only thing that remained is the battery which I gutted, and I am wondering if I should rebuild?

Today I worked on a couple of 486 motherboards, first on a 4FLUD-1.0, then on an ATC-1411B, both have problems due to corrosion, due to the loss of acid from the battery.
On the first one the damage is quite limited, but today investigating, I found three interrupted tracks, I'll see if tomorrow I can reconnect them, currently the card does not start, and the post card marks — —, probably it will be one of the broken tracks 😑, a track brings to quartz, so I imagine it can be back in operation, after reconnecting them, I will do a boot test, and I hope to see both the post codes and the BIOS screen.
The other one, on the other hand, has several oxidized points, tonight I scraped the green oxide from the copper, tomorrow I'll try, if I have enough time, to see if I find any interrupted, one is surely broken 😑, and in the back there are two or four, which I suspect are interrupted.

I played Fade to Black from 1995! The less famous sequel to Flashback.

Started it as a kid, unfinished business.

Not sure if poorly optimised early 3D engine, or hardware limitations but saw some pretty nasty slowdowns at times. First tried with the 6x86L clocked @ 190MHz, then switched to Pentium 100 and the performance got worse - so probably not an FPU issue. My GPU on that system is just the onboard Rage Pro 3D, wondering if that's the culprit? Univbe and custom DOS don't help much.

On paper, the system should make light work of the game, it was designed to run on P75 and DX2-66 systems after all!

Getting the 5155 close to being complete. Just waiting for the power adapter to arrive for the 3.5" floppy drive so it'll be fully set since I usually use 1.44MB diskettes to transfer files from my systems to machines like this since I don't have an enclosure to do 360K diskettes at the moment.

Anyhow, here is Planet X3 in 16 shades of amber:

Another thing I need to do is wire the internal speaker correctly by soldering a male header onto the wiring and go from there. That and fix the geometry on the display since it's a bit off on the top right corner. And once I get all of that done, I'll upgrade the BIOS to the GLaBIOS and add GLaTICK to the U19 side once I triple the BASIC ROM contents and add go from there.

I still haven't worked on the 5150 since I have the Macintosh 512K project that I'll be working on (full recap and battery holder cleaning/repairs).

Back to the Beige part 2, I pulled apart an LCD monitor to try and make it a retro style beige for my setup, still not sure how I feel about how it turned out, video below

https://youtu.be/RjFtzkFAsOI

Finished resurrecting the 486DX2-66, on Modula Tech 4FLUD-1.0 motherboards, I made three bridges to as many tracks, reassembled everything and added the minimum for a startup test.
It works!!!

"Fixed" My Acer 486... Again
There is some cold join on the motherboard. I probably have to reflow the whole thing or something.
Surprisingly it did like 4 months without issue till last time I played Fallout for half a day then it did the usual "flashing cursor problem" next time I booted it (Next day, when I wanted to continue said Fallout game.

Now doing the usual batteries of tests because there is more than one cold join.

Played around with GW-Basic and made a simple 'guess the number' game.
I think I'll start learning basic again, just for fun

Today I tried the ISA video card, on the PC 486 4FLUD-1.0 with Intel DX2-66, unfortunately it didn't work, the PC beeps and you can't see anything.
I'll have to get another one, I don't know what to check, to try to repair it.

Nothing to do with games or what most think of as a "computer", but good reminder
of how fun it can be to be a retired "old phart" work on stuff you like instead
of having the "make a living".
---
I've had a lifelong need to constantly "build stuff"... made a lot of useful
technical tools etc. over the years, but sometimes I got distracted by more
pressing things, and don't get back to a project. Having lost over a year of
memory hasn't helped, but sometimes gives pleasent surprises.

I recently "found" this bit of "General Test Equipment" that I had once started
building and had since forgotten about, containing:

1-8031 CPU with external 8/15/32k (selectable) ROM
2-32k Dallas "Timekeeper" NV-RAM (incl. clock and alarm)
3-Onboard pizo speaker/buzzer (on 1khz output from Timekeeper)
4-8-bit output latch            -8253 timer
5-ADC0809 8-channel AnalogToDigital convertor
6-20 character large LCD display
7-48-key panel keypad (salvaged from an ancient logic analyser)
8-RS232 serial port             -50-pin cable socket with connections:
9  8-digital outputs (latch)    1-digital input    8-In/Outs (8031 P1)
10  8031 Int0/Int1/T0/T1 inputs  8253 Clk0/1 Gate0/1 Out0/1/2
11  8-Analog (0-5v) inputs       0.5/1/2/4mhz clock outputs
12  Multiple +5/Gnd

A very useful kit when building/testing something designed to be interfaced
to a system. Gives you lots of programmable inputs/outputs that you can write
test programs to excercise. The Dalla 32k NV-RAM allows me to load in "test code"
and have it stay there even during power-OFF till I change it. (I of course had
to grind the top off the Dallas chip and tack on a good coin-cell battery)

Looks like I finished the main unit, but hadn't built any test interfaces.
And ... I've not found any schematics (I did have a bad habit of ripping stuff
out relying on my knowlege/memory to wire it)

So I got to spend some time figuring out exactly what I had done, where things
were connected/how they are wired. Fortunately I did find the sources to it's
current 8K ROM which answered some of the questions. It contains:
- My full 8051 debug monitor.
- My C-FLEA virtual machine (more powerful processor to easily run C code)
- Setup application - use Keys/LCD to set numerous operating parameters.

I just built a connectable proto-board to be able to easily test various
circuits ... and just used it to "figure out" some in-term DTMF receiver boards
that I got when I worked at Mitel (getting back into Amateur Radio and want to
be able to make devices controllable by a handheld with DTMF pad).

That is really neat ~

I'm similar in some ways but I use 90s and 2000s parts to make stuff with few goodies like a 68K controlling everything ~

Resoldered a few hdmi-to-vga/dvi adapters, minidp-to-multi outputs.
The problem was always the same, the different wires were soldered directly to the small motherboard to small pads and kept in place with some hot glue.
Anchoring was worn by mechanical stress and the tiny wires broke off. Boring but I saved some 50€ in new adapters.

Cheap means cheap, no doubt about it.

Disassembled Zenith Z-286 LP that does not POST for inspection…

sunkindly wrote on 2025-10-31, 17:51:

So JP16 and JP17 seem to set the fsb, but for JP19, JP20, and JP22 no matter what I set them to...it didn't change from 2.5x at 166MHz. Even setting it to something "normal" like 133MHz at 2x, nothing changed. The only thing that seemed to affect change was changing the FSB through JP16-17. But unfortunately, it doesn't support more than 66MHz.

If that clockgenerator (W49C65-03H) is similar enough to W49C65-01, then according to https://public.dhe.ibm.com/rs6000/PPC_support … ass2/ltspec.pdf I would suspect that both JP16 and JP17 closed would give 75 FSB?

I can't seem to see where the other jumpers are located from your image, though? [Edit:] Found them. Weird, though that JP21 is "skipped" 🤣

Tiido wrote on 2025-11-02, 22:21:

... I'm similar in some ways ...

Always to good to know there are other people of like minds/obsessions!

(argh - sorry, In thinking about "Stuff I've done", this all came out! don't know if there's a better place to put it)

I've been building stuff and keenly interested in electronics, at a very young age (elementary) I got to know a guy who worked in the phone system, and had collected a lot of parts - he gave me a bunch of carbon mics, receivers, rotary dialers and at least one step-by-step switch. I began making a rudimentary phone system, phones and wired the forest behind my house so my friends and I could call "tree to tree".

A year or two later I discovered radio, and began making simple transmitters and receivers so we could talk across the village.

At one point I made an analog computer (but didn't know what to call it) - A big mess of Potentiometers, Meters, a few transistors, relays, Lights and more I can't remember - no "stored program" capability, but you could do various types of math and get reasonably good answers.

In a high-school project where everyone else was assembling digital clocks, I convinced my teacher (who did TV repair on the side) to let me make an Oscilloscope - next week, when there were "clock kits" on everyone elses desk, there was a 3KP1 video tube, and handful of other tubes and a copy of an Eico scope schematic on mine - We had the metal shop make a case, I built it and it worked - Got my pic in the local paper for that one!

This all would have been late 60s - early70s.

When I went to university, I "met" my first computers, an IBM 360 mainframe and a few PDP-11s. I quickly got hooked!
and ... I met someone in the computing center who was building his own computer using a (then new) Intel 8080 processor. I had to build one too.

My first system was a homebuilt 8080, had a 63x16 TV display (should have been 64x16 but I messed up the design a bit), a huge Model-28 teletype for a printer, and a homemade wooden keyboard with a mess of diodes to encode it.

My first commercial machine was an original Altair 8800. After school I worked for a company using the Motorola 6809 - it quickly became my all time favorite processor. By this time I was building computer related "stuff" like crazy. I built my first "portable" computer running my own CUBIX (6809 OS)

Looking around my area, I see about a dozen homebuilt things with CPUs:
Eprom programmer, AmateurRadio/repeater controller, RS-232 interfaced "learning" IR-remote control, Video game input automater (fakes keyboard, mouse, joystick inputs), Serial-port expander: easily/transparently switch multiple things to the limited COM ports available on PC, Three 6809/CUBIX "desktops" and 4-5 wirewrap boards... There have been a lot more over the years that have gone by the wayside.

And I had a LOT of different computers (see "Daves Old Computers" - although I didn't actually use some of those). I have written emulators for some of the early ones I did use - anyone who would like to, can experience actually using:
- Altair 8800 (8080 - NorthStar DOS, my own DMF)
- NorthStar Horizon (Z80 - NorthStar DOS, CP/M)
- Vector 1+ (Z80 - NorthStar DOS, CP/M)
- D6809 Portable (6809 - CUBIX)
- MIL MOD8 (8008 - MIL MONITOR-8, Scelbi BASIC)

You can find out more about myself and my history on my site, just click on my name!

I have been on your site, it is full of cool stuff ~

I wish I was around at that time in that part of the world, so much fun was going on. I was interested in electrical stuff since I was really young, and it naturally evolved into electronics but there was the problem that *nobody* knew anything about it, and I had zero support from anyone (and opposition from mom who kept throwing away stuff I had collected to make things out of). It was quite a bit of an uphill battle. Internet and English knowledge changed everything though, along with moving out in my early 20s (which was in 2010 or thereabouts).

I dream of my own chip(s) but for now I settle with programmable logic devices and elaborate analog sections since I don't have access to integration beyond opamps and DACs and ADCs 🤣

H3nrik V! wrote on Yesterday, 09:43:

sunkindly wrote on 2025-10-31, 17:51:

So JP16 and JP17 seem to set the fsb, but for JP19, JP20, and JP22 no matter what I set them to...it didn't change from 2.5x at 166MHz. Even setting it to something "normal" like 133MHz at 2x, nothing changed. The only thing that seemed to affect change was changing the FSB through JP16-17. But unfortunately, it doesn't support more than 66MHz.

If that clockgenerator (W49C65-03H) is similar enough to W49C65-01, then according to https://public.dhe.ibm.com/rs6000/PPC_support … ass2/ltspec.pdf I would suspect that both JP16 and JP17 closed would give 75 FSB?

I can't seem to see where the other jumpers are located from your image, though? [Edit:] Found them. Weird, though that JP21 is "skipped" 🤣

Unfortunately closing JP16-17 gives 55 fsb.

Yeah, some things are kind of strange in terms of the jumpers. JP15 is also unpopulated.

The QDI board comes with a 200 MMX so I'll have two. I might put the second one into the M-Tech board just to see if anything changes but for now it looks like the M-Tech will be a spare Socket 7 board.