I'm turning a dead EQ into a 10x stereo input mixer, with differential inputs to successfully deal with the inevitable ground loops that form between all the computers that this thing will connect to. Tomorrow I'll build a power supply section and the actual mix amp, neither of which I got PCB made for but perhaps should have 🤣.

I have just completed my MT32-pi project, which included assembling the MIDI Hat and a 3D-printed case. Thanks to the interwebs!

Prepping a slightly-unusual-than-usual dual-boot machine to be sold.

P4 631, P5P800, a 160GB HDD halved in two 80GB partitions, FX5700LE and 2GB worth of DDR.

(I'm screwed, aren't I?)

The mixer has all inputs connected now. It took some effort to get the holes drilled and in the end they ended up a little offset still...

Next up is making the summing amplifier and output buffers. This thing will feed monitor on the desk, amp on the desk and output to the big amp in other side of the room and one output will be spare. I'll make outputs all differential too.

Bought an Acer 2560x1440 "gaming" screen to try the 3x upscale option of the OSSC and it does only work in DOS text mode - perfect picture.
"320x200" does not work though.
The problem seems to be related to the OSSC as the LED is turning red at that mode... one has to get the Pro version because there is a certain limit to the scanning capacity which just doesn't cut it.

Cleaned the heads of a 5.25" FDD (TEAC FD-55GFR), which were contaminated with magnetic film from some old rotting diskettes. It caked up so hard that even a cleaning diskette couldn't deal with the dirt, so I had to completely disassemble heads and scrape that stuff off.

Interesting observation - I tried to align heads using IMD alignment test and the drive passed it with a formatted HD diskette, but in reality, the FDD was still unable to read formatted DD or HD diskettes, so I had to align heads the old-fashioned "trial and error" way.

Did I miss setting stomething up in IMD regarding 5.25" drives, because with 3.5" drives, IMD has always been accurate in alignment tests.

SSTV2 wrote on 2025-02-25, 15:56:

Interesting observation - I tried to align heads using IMD alignment test and the drive passed it with a formatted HD diskette, but in reality, the FDD was still unable to read formatted DD or HD diskettes, so I had to align heads the old-fashioned "trial and error" way.

This is an inconsistent statement "aligned with IMD .. drive passed" and "was unable to read DD or HD disks"

IMD Test/Align does not write the disk - it reads the disk and beeping indicates when it could read the sectors
on a track matching the selected track...

So how could it "pass" alignment if it couldn't read the disks?

To align with IMD, you have to have a formatted diskette of known good alignment (I usually use some commercial
package distribution disk) - ** That matches the drive (DD or HD)

You step to a track, move the heads to the edges of where it can read the matching sectors, note positions
and clamp it down in the center of that range... rinse and repeat on various tracks over the range of stepping
till you find the best compromise...

If it can't read the disk, would never indicates that it could read sectors... so how could you figure out where to clamp
it down.

Another possibility - double-step setting should be OFF, use 40 track (360k) "master" for 40track (DD) drives,
and 80track "master"for 80-track (HD) drives - if you have any of these wrong, things won't work out.

IMD doesn't know of care about 3.5" .vs. 5.25" drives - it just knows stepping and data rates... the main difference
between 5.25" drives is that there are both 40 and 80 track drives where all 3.5" drives are 80 track!

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

The mixer is now pretty much finished. It would be nice to have some knobs on the sliders and buttons on the switches but overall it does what it was going to. There's tiny amount of hiss that is head from headphones that isn't reacting to slider positions but I can improve it by paralleling opamps on the summing amplifier. I also had to reorient the power transformer since there was a strong hum in the sound that didn't react to moving or shielding other things in the device. Overall this thing does what I made it for ~

Now it will have to find its way on the computer desk, among the KVM, VGA matrix and HDMI matrix 🤣

I had a mixer that I sold cheap on eBay because one of the sliders was crushed and there was no way to find a replacement part.

DaveDDS wrote on 2025-02-26, 02:14:

This is an inconsistent statement "aligned with IMD .. drive passed" and "was unable to read DD or HD disks" […]
Show full quote

SSTV2 wrote on 2025-02-25, 15:56:

Interesting observation - I tried to align heads using IMD alignment test and the drive passed it with a formatted HD diskette, but in reality, the FDD was still unable to read formatted DD or HD diskettes, so I had to align heads the old-fashioned "trial and error" way.

This is an inconsistent statement "aligned with IMD .. drive passed" and "was unable to read DD or HD disks"

IMD Test/Align does not write the disk - it reads the disk and beeping indicates when it could read the sectors
on a track matching the selected track...

So how could it "pass" alignment if it couldn't read the disks?

To align with IMD, you have to have a formatted diskette of known good alignment (I usually use some commercial
package distribution disk) - ** That matches the drive (DD or HD)

You step to a track, move the heads to the edges of where it can read the matching sectors, note positions
and clamp it down in the center of that range... rinse and repeat on various tracks over the range of stepping
till you find the best compromise...

If it can't read the disk, would never indicates that it could read sectors... so how could you figure out where to clamp
it down.

Another possibility - double-step setting should be OFF, use 40 track (360k) "master" for 40track (DD) drives,
and 80track "master"for 80-track (HD) drives - if you have any of these wrong, things won't work out.

IMD doesn't know of care about 3.5" .vs. 5.25" drives - it just knows stepping and data rates... the main difference
between 5.25" drives is that there are both 40 and 80 track drives where all 3.5" drives are 80 track!

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Thank you for the lengthy explanation, I think I figured out why I was getting perfect readings from the alignment test despite the upper head being misaligned - I was most likely checking the alignment of the lower head instead, because when I moved the upper head around, the number of sectors being read did not change at all, uh oh...

SSTV2 wrote on 2025-02-26, 15:06:

... I think I figured out why I was getting perfect readings from the alignment test despite the upper head being misaligned - I was most likely checking the alignment of the lower head instead, because when I moved the upper head around, the number of sectors being read did not change at all, uh oh...

Good point, and I did forget to mention that as you are moving around tracks to test, you should also
be changing 'H'eads - not something I (and most others) normally think about as the heads on most
drives are so well fixed relative to one another that you don't have to concern yourself too much
about aligning both...

But, since you removed the upper head, you would have to align them independently which can
be a bit tricky .. I'd align the bottom first, then hopefully the drive will let you loosen the upper
one enough to move it a little while you test .. otherwise you would have to do power down, adjust
and try to note the positions of the upper while finding the outer limits...

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

I realigned the heads on a floppy drive... ONCE, and it was a nightmare. I didn't know how to hook up a scope to the drive and was blindly moving the top head ever so slightly in different directions and validating it with a Compaq diagnostics tool. This took hours and I almost gave up multiple times. To make things worse, I had to account for the slight horizontal movement that occurred when I tightened the screw to lock it in place; I had it aligned once until I locked it down and then it was misaligned again. What a mess.

You have to be super careful because it's not just about aligning the drive to itself. You have to validate that it's able to read and write disks written from another known-good (aligned) drive. Otherwise you might end up writing misaligned disks that can only be read in that one drive.

Here is some information on my adventure. I did eventually succeed, but it was not an easy thing to do.

Re: If you have a Compaq LTE series laptop with a broken floppy drive, check the belt!

Here is some additional information on how precise the alignment process needs to be for a 1.44MB drive and why it's so hard to get it right.

https://retrocomputing.stackexchange.com/ques … ads-of-a-floppy

If you are reading floppies written with the same drive, it doesn't matter at all. […]
Show full quote

If you are reading floppies written with the same drive, it doesn't matter at all.

If you are reading floppies written with a different drive, then if those drives are differently aligned, the read head can start to pick up the flux from a neighbouring track. This materializes as "random noise" in the flux of the track you want to read, so as long as those signals are stronger, you'll still be able to read it. When the head is so far out of alignment that those signals get significantly distorted, you'll get read errors. It's even more fun if the head of the other drive wasn't properly aligned to start with.

So yes, there is a degree of slop, and actual results depend on a lot on which floppies you want to read, and where and how they were written.

I don't know how exactly you aligned your drive, but normally you do this with an oscilloscope and a floppy that is known to be written by a properly aligned drive, and then you can see directly the effects and the "slop" you get when moving the head, and how the signal distorts.

TLDR; never "transplant" the heads of a floppy drive if you can help it. Be prepared for potentially a lot of work getting them realigned otherwise if you're doing it blind. Extra brownie points if you can figure out how to hook up an oscilloscope to do it properly. I would like to learn how to do this someday.

I somehow got quite far with the 2GB RAM 98SE/XP 775-based AGP machine.

Granted, I used HIMEMX for 98SE, and RAM seems to be limited around 1150MB for 98SE. No crashes or weird juju so far.

Sidenote: whoever's idea was to bake custom 82.69 NVIDIA drivers into the REMOVED I am using... that wasn't a bright choice. FX5700LE AGP running @ PCI speeds clearly ain't good.

PcBytes wrote on 2025-02-26, 19:30:

I somehow got quite far with the 2GB RAM 98SE/XP 775-based AGP machine.

Windows 98SE should work great and it's easy to setup on your: "Main PC: i5 3470, GB B75M-D3H, 16GB RAM, 2x1TB." since you have PCI slots, USB should work with NUSB and so on...

OMORES wrote on 2025-02-26, 19:49:

PcBytes wrote on 2025-02-26, 19:30:

I somehow got quite far with the 2GB RAM 98SE/XP 775-based AGP machine.

Windows 98SE should work great and it's easy to setup on your: "Main PC: i5 3470, GB B75M-D3H, 16GB RAM, 2x1TB." since you have PCI slots, USB should work with NUSB and so on...

Not that one. It's a build I have put up for sale, made from a P4 631, ASUS P5P800, 2GB DDR400, 160GB SATA HDD (split in half to acommodate XP as well, over a NTFS partition so as to not have it accessible within 9x normally), FX5700LE.

Kahenraz wrote on 2025-02-26, 18:35:

I realigned the heads on a floppy drive... ONCE, and it was a nightmare. I didn't know how to hook up a scope to the drive and w […]
Show full quote

I realigned the heads on a floppy drive... ONCE, and it was a nightmare. I didn't know how to hook up a scope to the drive and was blindly moving the top head ever so slightly in different directions and validating it with a Compaq diagnostics tool. This took hours and I almost gave up multiple times. To make things worse, I had to account for the slight horizontal movement that occurred when I tightened the screw to lock it in place; I had it aligned once until I locked it down and then it was misaligned again. What a mess.

You have to be super careful because it's not just about aligning the drive to itself. You have to validate that it's able to read and write disks written from another known-good (aligned) drive. Otherwise you might end up writing misaligned disks that can only be read in that one drive.

Here is some information on my adventure. I did eventually succeed, but it was not an easy thing to do.

Re: If you have a Compaq LTE series laptop with a broken floppy drive, check the belt!

Here is some additional information on how precise the alignment process needs to be for a 1.44MB drive and why it's so hard to get it right.

https://retrocomputing.stackexchange.com/ques … ads-of-a-floppy

If you are reading floppies written with the same drive, it doesn't matter at all. […]
Show full quote

If you are reading floppies written with the same drive, it doesn't matter at all.

If you are reading floppies written with a different drive, then if those drives are differently aligned, the read head can start to pick up the flux from a neighbouring track. This materializes as "random noise" in the flux of the track you want to read, so as long as those signals are stronger, you'll still be able to read it. When the head is so far out of alignment that those signals get significantly distorted, you'll get read errors. It's even more fun if the head of the other drive wasn't properly aligned to start with.

So yes, there is a degree of slop, and actual results depend on a lot on which floppies you want to read, and where and how they were written.

I don't know how exactly you aligned your drive, but normally you do this with an oscilloscope and a floppy that is known to be written by a properly aligned drive, and then you can see directly the effects and the "slop" you get when moving the head, and how the signal distorts.

TLDR; never "transplant" the heads of a floppy drive if you can help it. Be prepared for potentially a lot of work getting them realigned otherwise if you're doing it blind. Extra brownie points if you can figure out how to hook up an oscilloscope to do it properly. I would like to learn how to do this someday.

I can really copy this.
I once cleaned the guts of a 3.5" drive WITHOUT removing the heads. After cleaning I put it all together and everything was good until the first inserted disk couldn't get ejected.
It turned out that the spring pushing this little knob upwards that's hooking into the disk to turn it was jammed and couldn't let the disk go.
But I only recognized that after I "freed" the disk in some way (naturally you can't see this knob when a disk is inserted into the drive).
With that "freeing" action the disk must have hooked into the upper head in some way and twisted the metal sheet where the actual head is mounted on a little too much.
It ended in the removal of the upper head to "correct" this deformation but I could never realign it correctly again.
Like Kahenraz wrote it is absolutely delicate to get the head positioned right again, especially in relation to the lower head, and after the final fixation it could still be misaligned (again).
I still wonder how they did in the factory.
I found only some little information about the functional work of a floppy in the net, with things like track width, radial offset between upper and lower heads and stuff and didn't see why the upper head wouldn't work again and spend hours with trial and error (and IMD) but eventually I gave up and the drive went into the parts bin.

Lesson learned: NEVER remove heads of a floppy.

I also wonder how the aligned the heads, especially with this laptop drive I was working on, because the head pivots slightly when you tighten the screw down. I ended up tightening lightly and then adding glue on top to prevent it from coming out from any vibrations. I also tweaked the alignment so much that the head started to tilt. I fixed this by slipping a folded piece of paper in the mechanism to raise one side slightly. It was such a mess. Anyways, I got it aligned (somehow) and it seems to be working fine so far.

The Emperor of S478 in Via.
Summary with Ti4200:

Replacement for 6600GT:

Tiido wrote on 2025-02-24, 12:59:

I'm turning a dead EQ into a 10x stereo input mixer, with differential inputs to successfully deal with the inevitable ground loops that form between all the computers that this thing will connect to.

Ha, that's a pretty neat project!
I probably could use something like that myself.
Instead, I've been doing it the manual way for longer than I can remember now: each PC has a 3.5<->3.5 mm cable plugged in its Line Out jack, and all of the ends from these cables are on top of my desk, to be plugged/unplugged manually either directly into a pair of headphones, or through my headphone amp (and into another pair of headphones.) I used to have "bookshelf" speakers at my old place too... but this has been on the "wait list" for my new place/setup.
In regards to ground loops, I've never had issues with that back when I lived in the US. But here in Europe, I have seen it a few times at other people's houses and even have a monitor that buzzes rather loudly (probably) due to a ground loop.
I wonder if just using a somewhat low value resistor between the ground of the output device and the ground of the receiver device can correct for this. If I recall, this is what some PSU manufacturers did to minimize ground loop noise for audio.

Tiido wrote on 2025-02-25, 08:30:

It took some effort to get the holes drilled and in the end they ended up a little offset still...

Did you use a smaller (well, the smallest you can get) drill bit to drill their centers first? I typically use a 2 mm one to get the centers drilled out, then move up to either 4 or 5 mm, then up to 8-10 mm... and widen once more for larger sizes.

Tiido wrote on 2025-02-26, 11:13:

I also had to reorient the power transformer since there was a strong hum in the sound that didn't react to moving or shielding other things in the device.

At least you have plenty of space for that inside. 😀
I wasn't so lucky with a subwoofer I was rebuilding for my dad. Its original transformer shorted on the primary, so I replaced it with another... but I just couldn't get the hum out, unless the transformer was placed outside the sub and further away from it. I guess I could have just re-boxed the transformer... but what I decided to do instead is use two 20V laptop power adapters in series, which not only gets rid of the hum, but also uses a lot less power in "standby" mode.

ChrisK wrote on 2025-02-27, 10:14:

Lesson learned: NEVER remove heads of a floppy.

Wish I read or asked about this a month ago. 🙁
I had a really dusty and dirty PC roll through. So every single component had to be taken apart for a wash, including the floppy (which had sucked in a lot of dust.) When I started taking that apart and got the the head part, it did cross my mind that perhaps this might mess up the alignment. But I figured, if the heads were held together with only 1 screw that was easily made to be removable, then surely this shouldn't be an issue. Guess I was quite WRONG on that. 🤣
We will see, though - I haven't tested the floppy drive yet. Maybe I got super lucky and it's all working fine. TBC... :p:

The drilling was an ordeal, because the side where I could get accurate measurements I couldn't drill on, and trying to get measures on the other side ended up being offset. But as far as centers go, I used small nails to punch detents into the metal, that will then keep the drill centered. I hammered the nails actually almost through the metal to get stuff on the other side where I actually could drill proper.

I was lucky that reorienting actually worked there, the other option would have been to make it external or perhaps find an alternative power source. I have some SMPS saved from things like set-top-boxes, and I'm sure I can get them modded enough to get a negative voltage out of them too (the opamps get +/-15V here, as they should).