BitWrangler wrote on 2022-10-09, 04:45:

Due to minor differences in geometry translation and there not really being an official standard until midway through the 1990s, minor differences meant that not every HDD formatted on given controller, with given BIOS and chipset, could necessarily be read on a different machine with a different controller/BIOS/chipset. The problem seemed to go away totally after ~27GB limit was sorted out, existed in edge cases in the post 8GB days, and everything before that, 90% of the time it worked all the time, though things get worse the further back you go. Maximum reliability on anything older than about 1998ish is best achieved by formatting your IDE drives on the machine you're gonna use them on... even though XT-IDE code and hardware is later you've still got original BIOS and chipset quirks that might pee in your cheerios. Some of it might be due to lack of care in making certain timing operations independent of processor execution speed.

Very good points. After setting up a brand new card everything does seem to be working fine now, so I am definitely going to keep all this in mind for future setups.

I did some retro repairs today, to clear off some of the projects that have taken up my workbench. I fixed two Radio Shack / Randy CCR-82 datacorders, and some tiny stereo speakers that had become damaged in shipping.

Radio Shack / Tandy CCR-82 repair

https://youtu.be/c-ktY5YOVPA

I also finished fixing up an AIWA AD-R40 cassette deck I picked up for cheap. It was in excellent condition and appeared to be barely used. All it needed were new belts, as the old ones had melted. It works great now.

Oof, a whole evening of floppy drive hell. The EME278TB drive that's seemingly unique to the T21xx series of Toshiba laptops is a pretty unique fitting and would need new cables to fit a later drive.
I chose to format the hard drive on my Toshiba T2130CT and load up the 17x restore disks only to find that the floppy drive wasn't as well restored as I thought.

The drive rails got stuck, I thought I broke the heads, the drive rail stepper motor jammed up to the point where anything past 80% of a disk could not be reached. For a while I thought I broke the disk spinner motor speed sensor.

I designed up a new belt though, this should fit any of the EME278TB and EME279TC drives and gives the best fit I've yet achieved with it being measured from the original belt that snapped instead of turning into goo. 204mm circumference, 0.4mm thick and 1.8mm tall. Prints nicely with TPU with a 0.25mm nozzle in vase mode.

Replacing the belt was my last ditch thing thing after it just stopped reading disks. Lubricated the drive spindles very carefully then cleaned off the belt contact areas with IPA. desoldered the motor from the board and ran it from 5v to see how it behaved - with the new belt it's much stronger. For good measure I replaced the little SMD capacitor on the motor/sensor board so I don't have to do this again in future.
For anyone that ever works on an EME278TB drive, when testing it make sure to never touch that screw on the back. That's the track position 0 sensor and is glued in place after being properly adjusted in the factory. The track position sensor uses a zebra strip to contact the drive PCB - which means that the aluminium top cover must be fitted and clipped in at the back for the track position 0 sensor to make proper contact.
To remove the drive tray which is necessary when replacing the belt, remove the top PCB carefully removing each of the flex connector cables. Mark on the metal floppy drive 'cage' which side has the long and short spring. Wedge something into the arm for the heads so they stay separated. My drives work after letting the two heads touch but they got really stuck one time and I could've ripped the head wires if separating them didn't go well. Easiest just to wedge something in there like in my picture.

To remove the floppy cage, push in the lower metal that connects to the eject arm plastic part. Once pushed back far enough the round pegs of the floppy cage should be freed up and then the floppy cage can be carefully lifted up from the front, avoiding the heads. Same thing to put it back in place later just be careful of the heads and when the springs are put back in, everything will hold in place. IF you find the floppy drive won't accept disks after reassembly, check that the lower nylon arm down by the head is installed the right way around, it can fit in a couple ways but only one way works. My picture is a good indication of how it fits together.

So many of my disks are in marginal condition 🙁 The SuperDrive can read them fine but regular floppy drives just see bad sectors.

Repairing floppy drives is really delicate work.
I once ruined one after cleaning, but not directly through the cleaning itself. After reassembly it just wouldn't rotate the disc so I took it apart again.
Couldn't find the issue at first but noticed that the upper head was bend a bit, that little metal spring part the actual head is fixed on to say exactly.
I don't know how this could happen but I suppose it must have got caught at the protection slider of a disc while ejecting. But this wasn't the actual initial problem though it could also be one reason the drive doesn't read any discs any more.
The actual cause of the drive not working was that little cam gripping into the metal ring of a disc and rotating it. It was stuck in its hole and did not grip the disk although the rotating mechanism itself worked.
Ridiculously easy to fix.
The problem was I mindlessly had already separated the upper from the lower head destroying the alignment between both. Never found a way for realigning them again. Still feeling so dumb...

Adrian from Adrian's Digital Basement has done a lot of floppy drive teardown and repairs. Maybe he has some advice on alignment. I don't recall if he ever touched on this inhl his videos, or if he ever did this off camera.

I think I can remember this name. I have spent quite some time on youtube and other sites on the web searching for some advice regarding this but most only deal with track zero problems.
I also found some specs regarding theoretical alignment of upper and lower head but none of that "knowledge" lead to anything useful for me in the end. It's just very difficult to move the head a defined amount just by hand.
I think there would be some more sophisticated apparatus needed for exact mechanical calibration as well as some test discs with special patterns and some knowledge how to interprete the read data on a scope.
Without special tools I don't see any real chance bringing this back to life, unfortunately.

Anyways, thanks for the tip. Maybe I will contact Adrain directly some day but for now I think I already spent too much time on this. In the end it's just some standard 3.5" Mitsumi drive...

Edit:
The conclusion I wanted to tell is:
Never ever separate the two heads from each other, meaning never loosen or remove the two brass screws that can be seen in Thermalwrongs last photo just right above the cap he has put between the heads.
It is possible to remove both heads together as a unit from the drive but this unit should not be separated.

Learning about the socket A 5v amp limitation the hard way...

gmaverick2k wrote on 2022-10-13, 21:27:

Learning about the socket A 5v amp limitation the hard way...
For an athlon xp 2800+(?) and ati rage 128 pro. Swapped out multiple cx500 until cx600 played nice. 9600xt flakes out in this system, suspect 5v rail weak on these supplies...

ChrisK wrote on 2022-10-13, 11:19:

I think I can remember this name. I have spent quite some time on youtube and other sites on the web searching for some advice r […]
Show full quote

I think I can remember this name. I have spent quite some time on youtube and other sites on the web searching for some advice regarding this but most only deal with track zero problems.
I also found some specs regarding theoretical alignment of upper and lower head but none of that "knowledge" lead to anything useful for me in the end. It's just very difficult to move the head a defined amount just by hand.
I think there would be some more sophisticated apparatus needed for exact mechanical calibration as well as some test discs with special patterns and some knowledge how to interprete the read data on a scope.
Without special tools I don't see any real chance bringing this back to life, unfortunately.

Anyways, thanks for the tip. Maybe I will contact Adrain directly some day but for now I think I already spent too much time on this. In the end it's just some standard 3.5" Mitsumi drive...

Edit:
The conclusion I wanted to tell is:
Never ever separate the two heads from each other, meaning never loosen or remove the two brass screws that can be seen in Thermalwrongs last photo just right above the cap he has put between the heads.
It is possible to remove both heads together as a unit from the drive but this unit should not be separated.

Are you using IMD to do this? it's the best alignment tool I've yet used since you can see in real time how the alignment is per head and per track. Adrian's said he uses it but I don't recall a video on it. PCRetroTech did a good video on how to use the tool here: https://youtu.be/VZ9xhFkHZ5c?t=525 I'm sure there are others but this is the video I followed to get a 5.25" drive working by combining 2 broken ones - I did a top head transplant on a 5.25" Toshiba floppy drive and had to align the replacement head with the lower which I didn't alter. So the track 0 position was unaltered but the lower head wasn't aligned with the top. It took a while but using IMD I was able to get the heads closely enough aligned that the drive is now able to read proper 5.25" disks.

With the size and density difference, aligning a 3.5" floppy drive is tougher. It's probably possible and I recall watching someone on youtube align a 3.5" track 0 and I think the heads through trial and error on an Amiga without realtime alignment software / scope.
It can help to have an oscilloscope to get the alignment perfect and 'retrofriends' youtube channel has done some videos on it, the most in-depth I've found so far for floppy drive alignment with a scope: https://www.youtube.com/c/retrofriends/search?query=align

Im frazzled, the laptop drive (also an EME278TB) that I used to make the disks for the T2130CT that I loaded the disks onto, that drive has broken somehow and I seem to be making it worse, can't even read the RPM 🙁
I think I need a break from floppy drives for a while again, into the 'fix later' box you go little floppy drive.

Here's an STL for the EME279TC's belt. Seems to work pretty nicely.

edit: For the alignment disk, I just use a new known good disk usually. Or one that was just formatted a couple of times by my nicest floppy drive, the LS-120 superdrive.

edit 2: IMD is daunting since it does so much stuff. The part I use is specifically - run IMD, press Escape to get out of the help system.
Since I'm replacing belts a lot, the "T" for test RPM is one that's good to verify the 3.5" floppy is spinning at 300RPM reliably.
To get alignment sorted, go into "A" to align/test the drive. If it's reading well the 3.5" 1.44MB floppy correctly it'll look just like this. Press the number keys 0 to 7/8 to get from Track 0 to 79. As the drive moves across the disk it'll drop out briefly then go back to '36' if it's all good:

Switch between the heads to see if they're both reading well at that position - that should be helpful to you but isn't needed for track 0 alignment.
Press R to make the drive go back to Track 0 and then back to the track it was just on. That can be helpful if the heads ended up in a funny position past 79 and you go back down to like track 70. I find that even good drives can get bad reads if pushed past the 79 track point then back down. Reseeking essentially recalibrates the head position.
Today's floppy repair station - the spare drive was broken too and needed recalibrating but somehow I nailed it on the first try and that's now superglued in place. Hope it lasts.

Try a different cable if you've got one around though, I am recalling a track 0 bad situation happening decades back with some cable freakiness.

Kahenraz wrote on 2022-10-12, 23:44:

I also finished fixing up an AIWA AD-R40 cassette deck I picked up for cheap. It was in excellent condition and appeared to be barely used. All it needed were new belts, as the old ones had melted. It works great now.

20221012_193008.jpg

just jumping in to admire this - looks good 😀

Many moths ago I bought this external harddrive for Amiga500's. It is a Rocktec RocHard RH800C. One of them side-card thingy, that I bought untested and without any PSU. I have been modding it today, so it can get power internally from the side-slot. Just two wires. One for 5v and one for 12v. I found that the power input on the backside, have direct connection to the HDD power plug inside, so my guess is that the external PSU delivers power to the HDD and the Ram gets the power from the Zorro-slot. Does not look pretty with the brown tape, but it will never be seen, once it is reassembled. And by golly, it was dirty inside. Big dark brown and black cakes in the internal fan. I hope it is working so I can use it on one of my 500's.

Yup..... Ram works at least....

trying to setup athlon64 3200+ nforce3 250 build. disabled a lot of things to minimise resource use. gotuit up and running but having issues installing daemon 3.47 and 7zip. get an error message page fault error. never had an 9issue installing these two important programs before. any advice?

gmaverick2k wrote on 2022-10-14, 18:13:

trying to setup athlon64 3200+ nforce3 250 build. disabled a lot of things to minimise resource use. gotuit up and running but having issues installing daemon 3.47 and 7zip. get an error message page fault error. never had an 9issue installing these two important programs before. any advice?

problem solved! weird... never had any other system do this, a quick google and installed 7zip v3.13 and daemon v3.44
Athlon64 system is off to a good start.... hopefully vortex2 doesnt have pci detect issue... thats next

thats a nope for dfi lanparty athlon64 build.... pci sound cards gimping out. resulting in vmouse.vxd error. no mouse
tried both vortex2 and cmi8738....

I tried to adjust the pots of the DVD drive in my original xbox because it wouldn't read any discs anymore.

Almost had them tweaked perfectly but then I broke the adjustment screw off. That's it, the laser is trash now.

Turns out, that replacement lasers for Philips DVD drives are around 50-70 bucks. Oh well...I guess I'll buy a complete junk Xbox to get the DVD drive out of it, saves some money.

Thermalwrong wrote on 2022-10-13, 22:23:

Are you using IMD to do this? it's the best alignment tool I've yet used since you can see in real time how the alignment is per […]
Show full quote

Are you using IMD to do this? it's the best alignment tool I've yet used since you can see in real time how the alignment is per head and per track. Adrian's said he uses it but I don't recall a video on it. PCRetroTech did a good video on how to use the tool here: https://youtu.be/VZ9xhFkHZ5c?t=525 I'm sure there are others but this is the video I followed to get a 5.25" drive working by combining 2 broken ones - I did a top head transplant on a 5.25" Toshiba floppy drive and had to align the replacement head with the lower which I didn't alter. So the track 0 position was unaltered but the lower head wasn't aligned with the top. It took a while but using IMD I was able to get the heads closely enough aligned that the drive is now able to read proper 5.25" disks.

With the size and density difference, aligning a 3.5" floppy drive is tougher. It's probably possible and I recall watching someone on youtube align a 3.5" track 0 and I think the heads through trial and error on an Amiga without realtime alignment software / scope.
It can help to have an oscilloscope to get the alignment perfect and 'retrofriends' youtube channel has done some videos on it, the most in-depth I've found so far for floppy drive alignment with a scope: https://www.youtube.com/c/retrofriends/search?query=align

Im frazzled, the laptop drive (also an EME278TB) that I used to make the disks for the T2130CT that I loaded the disks onto, that drive has broken somehow and I seem to be making it worse, can't even read the RPM 🙁
I think I need a break from floppy drives for a while again, into the 'fix later' box you go little floppy drive.

...

Thanks very much for these infos! I'll have a look into this as soon as I can find some free time for this.
I can remember I've read about some software tools (can't remember which ones though) but didn't use any as they were meant for some more ancient floppy drives iirc.
If yours can work on a "standard" PC (a.k.a. socket 7 class) this should definitely come in handy. Getting quite curious now...

While feeling lousy from the Flu and COVID Booster shots I got at work this afternoon (I work at a hospital) - I FINALLY nailed something I've been trying to for over a year.

So a year ago I bought a Panasonic KXL-D20 Sound/SCSI card off E-bay for $15 with the intents of pulling the audio signals - for the very least - for MIDI on my NEC Versa M/75 - well tonight.....I found them.

The KXL-D20 has a 32 or 34 pin header that goes into the Panasonic "Interface II" (IIRC) breakout box - the 34 pins all face DOWN when in the Versas PCMCIA Type II slot (Label up). I decided just to look at resources in Device Manager as I was working on tuning my Versa Video card to run on everything else....well, when I looked I saw ESS1688 in Windows 95's device manager with a bang on it (resource conflict methinks), but when I looked at one of the non-conflicting resources I saw port 388 listed.

So I rigged up a pair of cheap headphones I had with an aligator clip cable I use for testing guitar pickups - clipped the negative to the I/O sheild surrounding the pins, and then stuck a tiny wire into the positive for probing. Pretty much what I've been trying to do with zero results.

I fired up Monkey Island 2: Lechuck's Revenge and wandered on the bridge to summon up Largo LaGrande, poked pin one - nothing, pin 2, nothing, pin 3, nope, pin 4, uh-uh, pin 5 "dat dun dun dunnnn---dat dun dun dun" (!!!) - found it. Pin 6 "dat dun dun dun dun dun dun duuuuuuuu..." yeppers, left and right channels.

So if anyone is interested in coming up with some kind of bodge wire or funky rigging for their old laptop using a KXL-D20, there's a start, especially since those cards don't cost a kidney or a spleen. Now the big question....flat speaker.....or go for the modem sound input pin....hehe. I'm going to start looking at options next.

Technically this makes my Versa M/75 THE ULTIMATE MS-DOS Gaming laptop in my eyes, here I've got a 486 DX4, with over 32MB of RAM (40MB), an 80GB HDD, triple boots Win95/WFW311/DOS, and has WSS AND SoundBlaster/Adlib capabilities all in one box, and WiFi. The only thing that'd make this thing more awesome would be the ability to connect using WPA-PSK in DOS (actually could have that if I put 98 SE on it using Norton for WiFi management).