Yep. A lot of those 386 boards have the AMI BIOS. One of mine does, and I remember my dad's old 386DX/33 did as well. I like Award better -- usually has more options, IME -- but that color scheme is indeed quite ... uhm .. distinct. 😀

Congrats on the christening!

Just a suggestion: The first thing I do on a new motherboard acquisition is boot up my Memtest86 floppy (version 2 dot something, I think -- appropriate for old PCs) and let it run through a full RAM check. That verifies kind of the whole kernel system -- RAM, chipset, CPU, motherboard, etc. The next thing I do is partition and format the hard drive, and run Norton Disk Doctor (6.0) with the surface scan. You've got yourself an afternoon of staring at blue screens and progress bars, which IMO, is a great way to introduce yourself to a new build. 😁

I'm sure everyone has their own new-computer ritual. This is just mine. Thought I would throw it out there in case it's helpful.

Right now I'm in the "get the floppy drive working" ritual. =p Don't have the slightest bit of documentation on the controller card so I'm trying to dig that up, first.

edit: Well, I did find the documentation.

Kind of.

This is the best I can find...

And: https://store.cwc-group.com/isa1.html

Which only shows a portion of the jumper settings.

The drive correctly scans, it rotates the disk, the read head clamps down, I've cleaned the read head, and yet it simply refuses to read the boot disk that my friend made, which loads fine on my crappy USB floppy drive - and was tested on this exact same drive before shipping. 😒

Been trying all sorts of stuff. Set jumpers to a product photo I found, cleaned the read heads multiple times, tried booting the disk with my USB floppy drive and that worked fine so it is in fact bootable, tried making a new boot disk with the memtest86 floppy image, tried holding the damn thing at random angles to see if that helped anything. The multifunction card successfully reports parallel and serial just as it should, it operates the mechanical parts of the floppy drive, yet it just refuses to read disks.

Please send help.

This is actually quote a cool fun thread... I like it, in a coming of age kinda thing... and I especially would not knock or put down anyone trying their hand at building a 386.... as really, there are not enough of them!

I even like the case... it will be a "Sleeper" machine but in reverse. It gives me an idea to maybe use one of those see through cases... like Thermaltake, to be able to see all of the 386 as she runs!

Good luck with the build... I look forward to seeing more.

novasilisko wrote:

operates the mechanical parts of the floppy drive, yet it just refuses to read disks.

Please send help.

Can either be one of two things... the floppy disks are problematic...

or if they are proven to be good

You may need to experiment with another controller/floppy drive. I've performed surgery (cap replacement) on several OEM/proprietary floppy drives, that should read disks, but didn't.

It's just so very strange - this same disk + drive + cable were tested and working immediately before being sent to me, and it's bizarre that EVERYTHING on the drive would work before shipping, but then only reading fails afterward. It's so specific. Further, it's just as strange that everything on this controller card would work except for reading.

I think I need to take a break before I completely lose my head. I appreciate the words of support, though.

I'm sure you have already looked at this... but have you plugged in the floppy ribbon cable the right way around?

If one end is 180 degree out, this would stop a drive from doing anything, i'm sure you have checked this already though and is more like to be hardware.

386_junkie wrote:

I'm sure you have already looked at this... but have you plugged in the floppy ribbon cable the right way around?

If one end is 180 degree out, this would stop a drive from doing anything, i'm sure you have checked this already though and is more like to be hardware.

Yeah, the cable's in the right way around. Like I said, it properly spins and seeks, but it just won't read anything. Either that very specific part of the card is faulty, the drive got damaged in shipping, or something obscure is configured incorrectly... Unfortunately, I don't have access to another machine that actually has a floppy header, so I can't give it an independent test.

Here's my checklist for floppy problems, in no particular order. Complete list, even if it may not apply, or you know that already, or you've already checked it. Just in case.

- Floppy controller jumpers set to enable, and primary floppy if there's an option elsewise.
- Verify pin 1 on the controller to red or marked wire on cable.
- Long end of cable to controller. There's usually one or two middle connector(s), then a twist in the cable, then one or two end connector(s). The end connector goes to the drive you want to be A:. There may be no middle connector. In that case, the twist is in the middle of the cable and it doesn't matter as much which end goes where.
- Pin 1 on the drive is usually (but not always) the "inside" of the connector -- the one not aligned to the edge of the drive. Normally, if you're looking at the back of the drive, the pins are set to the right edge of the drive, and pin 1 is on the left. Power is usually on the leftmost edge (but could also be directly above the pins.) Check for a legend on the PCB or metal casing of the drive. If you get the orientation wrong, usually the drive LED will light up and stay that way. It will not be detected and won't work under DOS.
- Make sure pins are aligned and not off-by-one -- on both ends of the cable.
- Check for jumpers on the drive. If there is a way to choose "DS0" or "DS1", choose DS1. (PCs do not use DS0.) There may also be an option to pick A: or B:. One of those inserts a virtual cable twist in the drive. A twist in the cable plus a virtual twist on the drive equals no twist, and that makes the drive B: again. Try the other way if the first way doesn't work.
- Make sure the power connector is aligned right. It's often remarkably easy to get that offset by one, and possible to plug it in upside down. You'll usually know if you did this wrong by the presence of smoke. Don't ask how I know.
- Check the BIOS and make sure drive A: is set to 1.44MB 3.5".
- Make sure the BIOS option to swap floppy drive A:/B: is disabled (if applicable.)
- Make sure the BIOS floppy seek on boot is enabled (if applicable.)

If all of this is right, you should get a brief light on the floppy's access LED and a complimentary NRRRGGGHHH-uurggghh on boot. If you don't, something's broken.

Assuming you pass this far, your problem set has been whittled down to a malfunctioning drive or bad disk. You've ruled out the disk. So here's the drive checklist:

- If DOS does not think there's a disk in the drive, clean the sensors.
- If DOS can't read anything from the disk, the heads may be badly out of alignment. It is POSSIBLE to fix this, but very difficult.
- If the drive starts reading the disk (you hear a little TCHR, TCHR, TCHR of the drive heads advancing cylinders on the disk) then we're probably back to alignment again, but it's off by a little rather than off by a lot.
- If the drive doesn't really respond at all to I/O commands (no access LED, no head movement other than the initial clamp when you insert a disk), then it could be sensors again. Or it could be dead electronics.

SirNickity wrote:

- If the drive starts reading the disk (you hear a little TCHR, TCHR, TCHR of the drive heads advancing cylinders on the disk) then we're probably back to alignment again, but it's off by a little rather than off by a lot.

I don't quite reach this point. There's a little tick as the drive motor starts, a gentle whirr from the rotation of it for only a second or two, then it just gives up. It doesn't do this if I don't have a disk in the drive, so it evidently CAN tell whether or not a disk is in.

Here are two pictures of the heads.

https://i.imgur.com/iyWA4bQ.jpg

https://i.imgur.com/lziwHui.jpg

The top one does look a bit odd. I know it should be tilted back like that so it lays flat against the disk, but it does look slightly rotated laterally as well. I don't know how much is to be expected, though. It seems to level out when you put a disk in anyway, so...

Does the head move back and forth on the drive cleanly and easily.

The 1st most common problem I have with floppy drives is that the heads don't move properly due to the grease on the worm drive drying out due to lack of use. Giving the worm gear a bit of a clean and manually twisting the worm to break up hardened grease tends to get this working again.

The 2nd most common problem I have is that the capacitors die. Unfortunately, as the capacitors on these drives are tiny, you normally don't see too many physical signs they are dead, but often a drive I have written off for dead will suddenly spring back to life after replacing all the electrolytic capacitors on it.

canthearu wrote:

Does the head move back and forth on the drive cleanly and easily.

The 1st most common problem I have with floppy drives is that the heads don't move properly due to the grease on the worm drive drying out due to lack of use. Giving the worm gear a bit of a clean and manually twisting the worm to break up hardened grease tends to get this working again.

The 2nd most common problem I have is that the capacitors die. Unfortunately, as the capacitors on these drives are tiny, you normally don't see too many physical signs they are dead, but often a drive I have written off for dead will suddenly spring back to life after replacing all the electrolytic capacitors on it.

It seems to move just fine, and there's still some grease left although it definitely isn't as fresh as it could be. But like I said, this drive was documented as working just a few days before it was sent to me. The only major things that have changed since then are:

1. It was tested on a newer motherboard
2. It was tested on a motherboard floppy header
3. It was shipped

So... either something in the card is faulty (in such a way as to only affect reading data), something in the drive was damaged in shipping (in such a way as to only affect reading data), or something obscure is configured improperly.

novasilisko wrote:

Oh yeah. Now we're ready. […]
Show full quote

Oh yeah. Now we're ready.

OH YEAH. Celebatory pocky! I can officially start calling it Phoenix, now.

These are some very... enthusiastic BIOS colors (now with all the RAM installed - apparently I've got 4x 1MB and 4x 4MB)

I like that old 1993 Megatrends bios. remind me of my first 486dx-33mhz computer with 4mb ram, Oak SVGA graphics, Sony 2x CDROM with controller, 1440 modem, All ISA motherboard. Sound blaster 16, 14" SVGA monitor, and running DOS-6.22 / Win3.11.
Served me well for many years. Never broke, I just replaced it with a newer computer like a fool.

Chances are that during shipping, the read/write heads got knocked out of alignment, it's fixable (t'was easier on old Tandon full-height floppy drives) but very difficult on a 3½-inch drive.

That could explain why it does not want to read anything.

Regarding head alignment, on each head I see a white block with a black stripe. When the upper head presses down, the white blocks align fairly well, while the black stripes are offset by a few millimeters. I don't really know what it -should- look like.

Easy to see in the pic I linked earlier, just in the up position

It is quite dusty, but there isn't any on the heads.

Honestly, I keep feeling like it's a problem with the controller or some configuration thing rather than the drive mechanism/electronics...

Is it possible it's in exactly the same condition it was when it was tested before, but this controller/BIOS (I dunno where it'd be) just has poor quality error-correcting and can't work with the data it's getting?

I think I should still be able to get this to boot, though. I have a motherboard with an IDE header on it, so I should be able to just plug the module into that, format it correctly, put something bootable on it, and then swap it in. Assuming the IDE controller works 😒

I would just save time and replace parts for another i.e. controller/floppy drive... especially if they are generic.

The only time I would work to repair hardware is when there are no other alternatives, the part(s) are proprietary and no longer available (anywhere!), or the hardware is unique and could be of use in some way or another.

Floppy drives are plentiful and cheap as are ISA controllers. To save hassle just get a new one of each.

386_junkie wrote:

I would just save time and replace parts for another i.e. controller/floppy drive... especially if they are generic.

The only time I would work to repair hardware is when there are no other alternatives, the part(s) are proprietary and no longer available (anywhere!), or the hardware is unique and could be of use in some way or another.

Floppy drives are plentiful and cheap as are ISA controllers. To save hassle just get a new one of each.

Yeah. That's what I'll be doing.

Meanwhile, the damn disk module didn't show up today like it was supposed to, so I'm out of luck until monday at least...

So, I'm strategizing a bit, and I'm curious if my plot will work. I should be getting that disk-on-module monday or tuesday. But with no working floppy drive, I can't really install anything onto it in-situ... but one of my PCs does have an IDE header on the board. So I'm wondering - maybe I could unplug my other storage from that computer, install the DOM, install an OS to it there, then transfer it back to the 386 machine? I feel like this might be stretching the limits of backwards compatibility, though.

The biggest problem is how to get DOS onto your flash module. There's no reason, if you have an IDE interface, that it won't work. All you need to worry about is somehow getting a FAT16 filesystem on the module, and then transferring the DOS system files onto it. Booting from DOS would be the most direct route -- if you have the means on the new computer. Some kind of VM environment would be another. E.g., direct access to the flash module, and virtual floppy drives from disk images.

One thing I often do with Linux builds is a little more technical in nature, but it works. Create a disk image, and do your install to that in a sandbox / VM. Then write that disk image straight to the flash module. You will need to make sure the image has the same number of sectors as the module, and may need to mind the emulated geometry on both as well. (When I do this, it's typically LBA, so that part's easy. DOS would be another story.)

Worst case, maybe someone here (hello!) could create a disk image for you that you can write directly to your module.

Well, good news and bad news

After some super fiddly soldering to get a power connector attached to the teeny tiny connections on the drive, I got a FAT16 FreeDOS partition installed onto it (after hellish experience making a boot image. It boots properly in my newer machine (the only one with an IDE header). Putting it into the 386 is a mixed bag, though. It definitely detects it! The BIOS' auto-detect hard disk picks it up immediately and assigns geometry, but... as soon as I try to boot, it starts to load and then just stops. Stops right here, to be specific:

https://i.imgur.com/NTKnYpK.png

It's always something. I've let it set for a good long while, with no luck.

So, options:

- I made something that won't boot on a 386
- The auto-detected geometry was wrong
- Bad jumper settings for the controller
- The IDE controller has crapped out (which would be consistent with the floppy drive misbehaving...)

Here is the geometry the BIOS detects:
https://i.imgur.com/68E776v.png

And here's the datasheet for the drive series: http://www.mouser.com/ds/2/24/ap-fmxxxxex0xxr … ev10-293066.pdf

I've got the 2 GB standard version. It appears to match up...

Ignore all that crap, it friggen works!

🤣

https://www.youtube.com/watch?v=0VkrUG3OrPc