dominusprog wrote on Yesterday, 14:52:

Locutus wrote on Yesterday, 14:06:

Got myself a very nice XT JUKO board with 8MHz V20 CPU and C8087 FPU 😀

The attachment IMG_1680.jpg is no longer available

That ceramic 287 looks very nice.

Yup, but it is C8087 😉

This little Aptiva has had an interesting life!

I've been keeping my eye out for "original" IBM Aptivas since I first got interested in retro PCs. My first ever PC was a tower format Pentium 120 Aptiva in 1997, but I remember that even then I saw the illustrations of the desktop format version in the manuals, and thought it looked much sleeker and nicer (even though the tower had the great fun sliding drive bay cover). So when this P133 desktop 2134 model came up on eBay at a high price, I watched it for a while. The seller dropped the price a few times, and I finally bit when it came down again - still kind of expensive for what it is, but a "birthday present to myself"!

It was described as having been bought at a garage sale 15 years ago, and left in the loft ever since.

I checked it over and cleaned it up, and tested the PSU (this thread was useful for working out the custom soft-power on). To my surprise, absolutely everything worked - even the CD-ROM drive not only ejected, but read, and the PSU fan was so quiet that I could barely tell it was working! Even more of a surprise when I booted it from the original IBM FRU marked 1.2Gb hard drive ...

It booted into DOS, via a program called "Pause" that just said "Press any key to continue" - then straight into this bespoke industrial control app for Guinness in Runcorn (Cheshire, England) - the menus have all kinds of valve control options. Also amazingly, "ABM Ltd" still have a web page live with a press release from 2004 about their having renewed all of the "CIP" packaging plant at Guinness Runcorn. So I'd wager this little Aptiva had been sat there controlling Guinness bottling plant from 1997 until 2004? Which kind of fits with the eBay seller having bought it at a garage sale "15" years ago!

The inside of the machine was amazingly clean, considering - a fine layer of very dark dust on everything inside, but no big dust bunnies or seriously caked fans - either it had been in a very clean room, or had maybe been a spare machine that hadn't actually been used for that long?

Interesting choice for an industrial installation though - why an Aptiva, rather than an IBM PC series machine? I had been surprised it didn't have the usual mWave sound+modem card, but an empty slot. Perhaps it had had some kind of bespoke control card installed in place of the mWave?

I'll definitely back up the hard drive, and for now I've swapped in a 4Gb CF card as boot drive - I'll see if I can use the official IBM restore CD for a similar model Aptiva from Archive.org to get that full nostalgic experience!

Check out that white colour! Must have been hidden from the sun pretty well) If you can find time for that, could you please post a picture with the front panel open? It does open by touching that dimple on the left, is that correct?

justin1985 wrote on Yesterday, 17:27:

This little Aptiva has had an interesting life!

Btw, is your Dell sound bar still ok? I have one, but the speakers' cones degraded beyond redemption. They're like dried out and cracked gelatinous shells.

Archer57 wrote on Yesterday, 12:33:

Alesia wrote on Yesterday, 11:19:

That's too bad, I really feel like I got to this one in the nick of time, knowing how often things get overpriced in that shop and sit for months/years. I'm coming around to the idea of assembling some sort of socketed replacement for this PRAM battery situation but I'm unsure if a 3v coin cell would work in a spot that normally takes a 2.4v battery. I know all these assemblies protect the coin cells from charging but I don't know of one that tries to prevent the new battery from pumping too much voltage into the computer.

Hmm, would not a diode in series both protect it from charging and drop the voltage by exactly as much as needed?

Also Ni-Cd nominal voltage is 1.2v, but they are charged to ~1.4-1.45 IIRC, which would be 2.4 and 2.8-2.9 respectively for 2S battery. Jump from 2.8-2.9 to 3.0-3.2 for lithium cell should probably be fine...

The original Not-A-Varta replacements (the ones that are flat and use a 1220 battery) use a MAX40200AUK+T to prevent the charging voltage from reaching the battery, so that is exactly that is done as best as I can tell. I'm not versed in the function of electronic parts though, I just see a 3v lithium battery being used in place of a 2.4v Ni-Mh and get worried. https://github.com/wiretap-retro/Not-A-Varta <- the design I'm considering using due to space constraints.

Nexxen wrote on 2025-07-04, 11:36:

ChrisK wrote on 2025-07-04, 10:51:

The balls itself should be ok as far as can be seen. It's mainly the most upper layer of the package that was lifted.

Push it down with a finger and turn it on.
Worst case it doesn't POST, best, it works only when pushed down.

This has absolutely worked for me when troubleshooting a board. It's helpful to feel around for overheating chips to look for a short and to press down on components to test for cracked solder joints. I have diagnosed problems successfully with this technique before. I'm not certain that it will work for a cracked via, but you can still try.

Halofiber86 wrote on Yesterday, 17:38:

Check out that white colour! Must have been hidden from the sun pretty well) If you can find time for that, could you please post a picture with the front panel open? It does open by touching that dimple on the left, is that correct?

Correct! It's not as dramatic as the sliding cover on the towers, but it's great to see IBM went to the trouble of fitting a damping mechanism - press the dimple and it very gently rolls open.

The general Aptiva Restore Disk worked once I realised my CF card wasn't being detected because it was formatted with FAT32, so threw an "invalid drive specification" error. Great to see all the custom animated icons and everything again!

vstrakh wrote on Yesterday, 18:40:

Btw, is your Dell sound bar still ok? I have one, but the speakers' cones degraded beyond redemption. They're like dried out and cracked gelatinous shells.

I've not noticed any problems... It's the wrong one for this monitor (which is newer and the clips don't fit). The correct one is not only USB powered but only functions as a USB sound device. So this is one of the older ones with 3D printed replacement clips glued on the top. I'm powering it from one of the cheap USB to 12v barrel jack step-up adapters. Pretty ideal solution in my book ... So I hope they're not destined to disintegrate!

Goodies arrived yesterday.

- P3B-F works, slotket surprisingly supports 100FSB but it desperately needs a BIOS update
- FireGL V7200 works, currently in process of flashing to X1800
- AOpen i855GMEm-LFS works, I just need to get around setting up a XP HDD 🤣, and maybe a Pentium M 780. Has latest R1.08bn BIOS that I submitted to TRW. Also gotta restore the Apevia case.
- got a SL-75KAV double. Rev F1 as opposed to the F5 I previously got.

Alesia wrote on Yesterday, 19:07:

The original Not-A-Varta replacements (the ones that are flat and use a 1220 battery) use a MAX40200AUK+T to prevent the charging voltage from reaching the battery, so that is exactly that is done as best as I can tell. I'm not versed in the function of electronic parts though, I just see a 3v lithium battery being used in place of a 2.4v Ni-Mh and get worried. https://github.com/wiretap-retro/Not-A-Varta <- the design I'm considering using due to space constraints.

Well, they are using a fancy chip specifically to get as low voltage drop as possible, because it is intended to replace 3S battery (single cell voltage for this batteries is 1.2V, yours is 2 in series so 2.4V, one that is intended to be replaced by that design is 3 cells in series, 3.6v). Technically if you replace that fancy chip with simple regular diode, like 1N4148, it will cause ~0.6-0.7v drop, which is exactly what you need.

Also nominal voltage for this batteries is generally somewhere in the middle of actual operational voltage range. Like for li-ion we are all used to it is usually specified as 3.7v or 3.6v, but actual operational range is 2.5-4.2V. For Ni-Cd chemistry used in this batteries operational range is not defined as precisely, but it is something like 0.9V-1.45V. Obviously circuitry used has to be able to handle whole range, not just nominal voltage, usually with some margins. High end of this range is very close to 3V, so it should be fine. Though those 3V battery is not exactly 3V either, it will be around 3.2V when new and drop as it discharges.

I am pretty sure it should work fine because it is so close, but if you want to be absolutely sure you can look up clock chip datasheet (if that's even possible) to see what voltages it is rated for, or just replace those MAX40200AUK+T with a simple diode like 1N4148 and then voltage which reaches the clock chip will absolutely be within range of what original battery had. The diode will have a different footprint than those chip though, so that'll require either modifying the PCB or just finding a way to solder it in there as is...

Archer57 wrote on Today, 00:43:

Well, they are using a fancy chip specifically to get as low voltage drop as possible, because it is intended to replace 3S batt […]
Show full quote

Alesia wrote on Yesterday, 19:07:

The original Not-A-Varta replacements (the ones that are flat and use a 1220 battery) use a MAX40200AUK+T to prevent the charging voltage from reaching the battery, so that is exactly that is done as best as I can tell. I'm not versed in the function of electronic parts though, I just see a 3v lithium battery being used in place of a 2.4v Ni-Mh and get worried. https://github.com/wiretap-retro/Not-A-Varta <- the design I'm considering using due to space constraints.

Well, they are using a fancy chip specifically to get as low voltage drop as possible, because it is intended to replace 3S battery (single cell voltage for this batteries is 1.2V, yours is 2 in series so 2.4V, one that is intended to be replaced by that design is 3 cells in series, 3.6v). Technically if you replace that fancy chip with simple regular diode, like 1N4148, it will cause ~0.6-0.7v drop, which is exactly what you need.

Also nominal voltage for this batteries is generally somewhere in the middle of actual operational voltage range. Like for li-ion we are all used to it is usually specified as 3.7v or 3.6v, but actual operational range is 2.5-4.2V. For Ni-Cd chemistry used in this batteries operational range is not defined as precisely, but it is something like 0.9V-1.45V. Obviously circuitry used has to be able to handle whole range, not just nominal voltage, usually with some margins. High end of this range is very close to 3V, so it should be fine. Though those 3V battery is not exactly 3V either, it will be around 3.2V when new and drop as it discharges.

I am pretty sure it should work fine because it is so close, but if you want to be absolutely sure you can look up clock chip datasheet (if that's even possible) to see what voltages it is rated for, or just replace those MAX40200AUK+T with a simple diode like 1N4148 and then voltage which reaches the clock chip will absolutely be within range of what original battery had. The diode will have a different footprint than those chip though, so that'll require either modifying the PCB or just finding a way to solder it in there as is...

I took it up to design a custom solution for this using 2x 1.2v rechargeable ni-mh batteries in series. I'll probably post about it once I get details more settled (mostly dimensions determining what batteries/battery mounts can be used) so that people can double check the work. Keeping the chemistry and voltages all the same lets me keep the design simple, at least. I really appreciate the thoughts, very much helped me decide which direction I wanted to try and go with finding a replacement solution for this powerbook.