Arbuthnot wrote on 2020-07-26, 16:07:

Today I made a quick and ugly BIOS battery holder for my 486.

Better than my battery holder for my 386/48yDLC machine. Mine is a slightly yellowed DMG-01 bottom chassis that's seen better days, but it works.

All day yesterday, after doing a bunch of work for my future guitar company (building some bodies), I started working on the Versa M75. I did a long writeup today of the initial thing in my new Build Log's thread (I plan to post past and present and even my future builds/modifications/restorations) into this thread. So the Versa, I'm figuring out the audio issues, I think I've got a line on 2 digitizers for it (I Found 2 on E-bay), and am now in process of plotting making a new hinge cover out of aluminum, or piecing the original together using JB Weld Steel Stik and J.B. Weld original.....after I take apart the hinge and fix the friction bits. Spent a lot of last night on the M75 and 40EC on the patio downloading files.

I also opened up the CommPac Words + - can't find a charger that works with it yet......looks like it uses a NiCad or NiMH Battery.....I plan to fix that up and get that going. Interesting mess of EEPROMs and even an 80186 CPU. This must have been when System 2000/Words+ were just starting out, the whole thing looks like it was built using either cast-off, heavily modified, or parts from Fry's/Radio Shack circa 1994. Also, all the dates inside the laptop ahd Words Plus put it around December 1994 build date.

Lubricant for hinges is heavy molybdenum grease is used when reassembling hinges.

Cheers,

Tested a bunch of PSU units to see if they still work...

6 ATX units working
1 ATX unit broken (suspect it already)
2 AT units working
3 AT units broken (2 suspected, one surprise).
1 Tulip PC Compact 2 unit working
1 weird form factor AT compatible unit working
1 PS/2 model 30 8086 PSU working

Grab an AT motherboard (not the most valuable one), install CPU/ram/graphics, hook up kB/mouse.

Hook up PSU no 1 and if the system starts, run a few loops of 3D mark 99 while measuring voltages. Then PSU no 2. And so on.
Then switch to ATX motherboard and go on.
Takes quite a while but at least I know what I have now!

aha2940 wrote on 2020-07-26, 16:32:

Yes, much time went into the locking part of the hard drive, plus the cable replacement required for the SATA drive (original 40-thread PATA cable would not work, had to use an 80-thread one), plus the eeprom.bin thing. We learned a bit, though. We had to softmod the xbox because it's one of the later ones which are harder to have the BIOS modded.

Which revision specifically? Also, if you can, you’re going to want to disassemble it again in due time to replace the thermal paste, since it is most likely a gummy wad of material that’s bringing your temps up.

ragefury32 wrote on 2020-07-21, 03:30:

Ah, crap. Had to perform some emergency triage on my IBM Thinkpad T21 - looks like leaving it on AC for the weekend to charge the batteries might've given it the infamous blink of death. Let's see if it comes back up to life within a day or 2 - if not, it might be time for a new motherboard.

Swapped the motherboard between a perfectly working Thinkpad T21 (800MHz) with one suffering from blink-of-death (850MHz). Well, now I have 2 machines with blink-of-death. Freaking T2x machines...

BetaC wrote on 2020-07-26, 21:46:

aha2940 wrote on 2020-07-26, 16:32:

Yes, much time went into the locking part of the hard drive, plus the cable replacement required for the SATA drive (original 40-thread PATA cable would not work, had to use an 80-thread one), plus the eeprom.bin thing. We learned a bit, though. We had to softmod the xbox because it's one of the later ones which are harder to have the BIOS modded.

Which revision specifically? Also, if you can, you’re going to want to disassemble it again in due time to replace the thermal paste, since it is most likely a gummy wad of material that’s bringing your temps up.

It's a 1.6b for sure, the one that does not have the issue with the clock capacitor. Thanks for the tip on the thermal paste, will do when I get some more time (next weekend, maybe).

Well, the retro activity did not stop at all today - in between movies I worked more on the M/75, including I dove back into the dumpster and retrieved the Digitizer Panel one last time. I was going through the photos and realized there were more numbers and things on it than I initially thought. Not sure what numbers are what - tempted to e-mail 3M (the maker) to find out.

However, I did find this is a VERY simple device. Five wires. I made a diagram which I have posted. I also put it on my website.

In other oddness, I picked up the M/75 to dink around in FreeDOS for a bit and found out that it's actually attempting to charge it's batteries, and it seems after a 2nd charge cycle, it's trying to run off those batteries a little more each time.......sheesh...Might be time to go into setup and set it for conservative power management so I can maybe recondition them. It's a smart battery so no real risk of blowing TBH. It'll be nuts if it actually starts running off battery power, 🤣. Too bad COVID-19, or i'd be gung-ho to show up at a Starbucks in flannel, torn jeans, wearing a beanie as "90's man"....hehehehe, sippin coffee, surfing on free WiFi with my phone's mobile hotspot as hidden proxy.....hey, I got the car and the music to go with it.

I narrowed down the exact MR-BIOS rom image I needed for my 486 board (GA-486VM, UM480 chipset, probably rev B) and got to troubleshooting my second copy of the IODE-3295UMC card -- it works, but neither the DOS driver nor a utility a friend wrote for setting drive speeds on its UM8672 IDE controller works. I noticed a little bit of rework on the card (specifically, near the UM8672 chip), and wonder if that might not be the problem. I'm not ordering another one -- I'll just have to see if I can either fix the problem, or, failing that, I'll just live with PIO0 I guess... changing slots doesn't do anything, removing the XT-IDE ROM'd NIC does nothing (I thought maybe that BIOS was hijacking the card), booting in turbo mode does nothing (my friend thought maybe it had to do with needing to slow down its addressing of the 8672 chip).

Checking out a Digilog serial analyzer (from the 80's) we're tossing at work to see if I want to turn it into a cyberdeck. Has a B&W CRT (white phosphor) integrated that I could easily replace with a 5" or 6" color CRT.

Gallery: https://imgur.com/a/gJADnq3

I received some bad news and have been distracted all day. I ended up making a basic Windows XP PC, using IDE drives and everything, to keep me occupied but my only spare case was this tiny 1999 Tiny PC case...

Removed the Varta battery from my Amiga 2000. The area looks pretty bad.. very corroded. However, it is really only the solder mask that's screwed up from the battery acid. If you scrape it all away with a fiberglass scratch pen, it is like brand new copper underneath. All traces have full continuity and no breaks - in fact almost no damage at all. I have to remove the CPU socket next, then clean up the solder mask under there too. I'm going to repaint with fresh UV curable solder mask paint to protect it, once I finish removing the acid impregnated solder mask areas.

Wowzers.

My JVC Kaboom! box (original 1999 model) finally came in the mail today!

I just love this thing, and lemme tell you, I definitely enjoyed some digital audio in a very 1999 kinda way today. I fired up the old Win98 machine, grabbed a bunch of music (dang, the 9x network stack is slow!), and burned it to CD. 80 glorious minutes of Jamiroquai, Backstreet Boys, Spice Girls, New Radicals, Hanson, Smash Mouth -- heck, I even learned what a scrub is!

Best of all, the Kaboom's Super Woofers sound exactly like I remember: not strictly accurate, but they sure do give a little bump up around 60Hz, which gives a goose to the bass right where I guiltily like it to be goosed up. Way cool!!

bjwil1991 wrote on 2020-07-28, 04:45:

Wowzers.

My thoughts exactly. Pretty scary, but necessary. The battery acid penetrates the solder mask and sits on the copper layer, slowly eating away it it. The only cure is to remove it and rework it to prevent future degradation.

I've got one of these Dell batteries:

Overheats like crazy when charged to 100%, i'm currently checking which cells are overheating... (there are 3 in parallel, maybe I can cut one row off, or reposition the thermistor)
i asked professional company to replace the cells but they refused, I guess i will have to do it by myself then, the battery controller shows only 25% wear so it's probably quite tolerant and maybe replacing the cells will not lock the battery (first i will connect new ones in parallel so voltage won't be cut off), I guess that's one of the reasons these batteries were exploding... I will probably just order 3 high current cells and leave empty space in the battery, I don't need large capacity

You'll want to replace all the batteries at once or make a new resized capacity battery. If you mix new cells in with old ones, you'll have uneven loads on the cells, different drain rates, and different temperatures. Don't worry about charging cell voltage cutoff - if you buy new batteries, they'll be at storage voltage upon arrival and you can charge them once installed.

wiretap wrote on 2020-07-28, 16:46:

If you mix new cells in with old ones, you'll have uneven loads on the cells, different drain rates, and different temperatures.

Yeah, i want them to be gone completely because this series has confirmed safety issues, there was recall of 4 million battery packs. I just read that if battery controller gets down to 0V it may disable itself by flagging failure in EEPROM or something like that. So i thought of soldering similarly charged cells in parallel to the old ones (on a wire, if i leave empty space there will be place for wires), then cutting the old ones off. I will maybe buy 6 cells, because the charging current is 3.5A and that may be over spec for single cells.

adalbert wrote on 2020-07-28, 16:54:

wiretap wrote on 2020-07-28, 16:46:

If you mix new cells in with old ones, you'll have uneven loads on the cells, different drain rates, and different temperatures.

Yeah, i want them to be gone completely because this series has confirmed safety issues, there was recall of 4 million battery packs. I just read that if battery controller gets down to 0V it may disable itself by flagging failure in EEPROM or something like that. So i thought of soldering similarly charged cells in parallel to the old ones (on a wire, if i leave empty space there will be place for wires), then cutting the old ones off. I will maybe buy 6 cells, because the charging current is 3.5A and that may be over spec for single cells.

I haven't heard of a charge controller never being able to work again once it senses 0V (or probably lower than 3.5V per cell). If that's really the case, I would temporarily solder a power supply (like a wall wart) onto the charge controller where it senses voltage so you can work on the battery pack separately and just drop it into place.

Bought an AT desktop case from eBay, which arrived today -

Had to do a bit of cleanup inside to remove some leftover battery leakage & other corrosion spots, some of the cleaning I've done can be seen in the 2nd pic. But otherwise it was in very good condition externally.

And the purpose of this new addition to my collection - to finally transfer my XT/8088 build into something more suitable and closer to it's era.

From this -

To this -

Still a work in progress though, only just finished putting everything together and haven't connected anything up yet - will do that tomorrow. The PSU that came with the case is going to be tested first before I put it anywhere near my motherboard. If it tests out ok, it can stay in there otherwise I'll reuse the PSU from the old case. I'm also going to replace the temporary card blanking plate with a proper black plastic one & print out a better looking fake IBM case badge, the one in the pic was just an example to make sure I had my measurements right.

Later this year I'll probably fill the blank space above the MFM HDD with a black 5.25 to 3.5 bracket & a black Gotek. I didn't like the idea of a vertically mounted Gotek, so the 3.5" floppy drive will stay as a placeholder/B: drive.

This case has a MHz display too, so I'll try and figure out how to set it to 4... or 5? 🤣

The previous tower case I'm going to use to rebuild a Super Socket 7 K6-2 PC that has been sitting in parts awaiting a case for a few years. Might even convert the tower case back to ATX, as it's currently modified to use an AT-style power switch (with ATX PSU with AT adapter) and the SS7 board can use either AT or ATX so seems senseless to stick with AT.