VOGONS


First post, by konc

User metadata
Rank l33t
Rank
l33t

OK it seems that floppy drives don’t like me lately. I’ve got a very weird problem that I’ll try to describe along with the steps I’ve taken, mostly looking for ideas since this is quite unusual.

The problem: It’s a 386DX/40 dos 6.22-only. When trying to access the floppy drive in any way, eg. dir a: - copy a file - anything the drive light goes on, there is no mechanical movement at all and it fails with the classic “not ready reading drive a: abort, retry, fail” after a while. The message appears delayed when compared to not having a disk in the drive.

What I’ve tried:
-Brand new 1.44 floppy drives, confirmed working on another pc
-Changing the drive with a 5.25 one, also confirmed working
-Trying to access drive b: with both drives connected
-BIOS: reset settings, default settings, triple-checking drives types are correct, lower all timing to the lowest possible
-I/O card: works on another pc but I also tried other ISA slots and another I/O card and floppy cable just in case with the exact same behavior.
-Changed literally all the components: VGA, RAM SIMMs, disabled cache, another PSU, disconnected HDD. Booted clean from another HDD and different dos version, tried to boot without HDD from floppy.

Weirdly I have absolutely no other indication of problem with the system. It’s heavily used running games, demos, benchmarks in loop, memtest86+, not a single glitch. Just the floppy. So I guess only the motherboard remains to blame, right? OK I agree. But what on earth could be the problem on a m/b without onboard I/O and no apparent other problem? I hope someone has some idea on where to start looking, as I have no clue what else to check/replace.

What can cause such a behavior on a motherboard without onboard I/O?

Reply 1 of 17, by Horun

User metadata
Rank l33t++
Rank
l33t++

Something interfering with IRQ6 or DMA2 could cause floppy issues. So you changed out the 1.44m floppy drive with other working floppy drives, changed cables and controller and you still get "drive not ready errors". Have you tried booting clean ? F5 during DOS boot will boot without any drivers installed.

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Reply 3 of 17, by konc

User metadata
Rank l33t
Rank
l33t
Horun wrote on 2020-02-23, 20:44:

Something interfering with IRQ6 or DMA2 could cause floppy issues. So you changed out the 1.44m floppy drive with other working floppy drives, changed cables and controller and you still get "drive not ready errors". Have you tried booting clean ? F5 during DOS boot will boot without any drivers installed.

That's an interesting idea I haven't explore, I'll check what's going on with IRQ6 and DMA2. Thanks, that's the kind of things I'm hoping with this anyway, it's not realistic to expect a "hey replace chip x". I did boot clean: holding shit down, with another hdd just formatted and sys'ed, without an hdd from a floppy (failed of course), different dos versions 5-7.1

maxtherabbit wrote on 2020-02-23, 20:49:

someone else recently had this issue here and it turned out the 14.whatever MHz crystal oscillator on the motherboard was bad

😃Yeah that was me, that's why I wrote "floppy drives don’t like me lately". But that one was intermittent behavior, it was mostly working and then stopping eg. in the middle of a file copy or format. This one now is consistent, it just doesn't work. For what it's worth I did replace the 14.318 oscillator (I had ordered a couple because of the previous problem so I had spares and well... although improbable I just did it since this happened to me recently) with no change in behavior. Forgot to mention it as it didn't change anything.

Last edited by konc on 2020-02-23, 21:44. Edited 1 time in total.

Reply 5 of 17, by Baoran

User metadata
Rank l33t
Rank
l33t

I once had similar issue and it ended up being one of the pins of the chipset was bent and was touching the pin next to it. I would check out the motherboard with a magnifier to see if anything is visibly wrong with it.

Reply 6 of 17, by douglar

User metadata
Rank Oldbie
Rank
Oldbie

I have a 486 motherboard, G486PLB-3, with 1991 BIOS, that doesn't work with floppy drives, so I'm watching this thread closely. The IDE HDD works but I always get an "FDD controller error" at the end of the POST. When the board tries to use the drive, the light comes on, but there are no mechanical sounds.

  • using a known good cable, floppy drive, diskette & power supply combination that works with other 486 motherboards.
  • same symptoms if I use the proprietary local bus controller that came with the motherboard or if I use a basic known good ISA multi/IO card with a Trident 8900.
  • tried different RAM.
  • Tried changing the shadow ram settings.

I still get the "FDD controller error" at the end of the POST.

This system had some battery damage. I cleaned up the surface traces and cleaned out the battery mount vias.

The board has another odd symptom. If I use a DX2-66 or DX4-100 overdrive, I have to hit the reset switch once or twice before it will POST. Doesn't need that when using a DX33 or SX33. Could just be because the old capacitors need to warm up before they can give enough current to the hungrier CPUs ? Runs stable once it boots.

Reply 9 of 17, by barleyguy

User metadata
Rank Newbie
Rank
Newbie

The reset switch symptom is probably the power good test. Before the machine boots, it tests the power quality. If it thinks the power is bad/insufficient, it will refuse to start. The reason it happens with the faster processor is probably because it draws more power. Hitting reset a couple of times gets around the power good test and allows the machine to boot.

EDIT: Are you sure your power supply is good? My recommendation would be to test the voltages with a voltmeter. With the machine running but the top off, put the negative lead of a voltmeter on one of the black wires going into the motherboard power, and then check each of the colored wires with the positive lead. Each one should be very close to either 5 volts, 12 volts, or 3.3 volts. If they are way off, like 4 volts, 6 volts, 13 volts, your power supply is too far off in voltage, which will cause the power good test to fail. This can also be affected if you are overdrawing the power supply.

Reply 10 of 17, by konc

User metadata
Rank l33t
Rank
l33t
maxtherabbit wrote on 2020-02-23, 21:38:

the other thing to check is the DMA controller on your motherboard. Run Checkit and do the system board tests

Checkit and all other diagnostics I've run pass. And that's the most frustrating thing about this, I have no other issue/indication to help locate the problem

Baoran wrote on 2020-02-24, 05:11:

I once had similar issue and it ended up being one of the pins of the chipset was bent and was touching the pin next to it. I would check out the motherboard with a magnifier to see if anything is visibly wrong with it.

Good idea, I was hoping to see something but after a careful inspection of the chipset it's very clean and definitely nothing bent/touching. Not even slight oxidation to suspect a pin is not making contact.

SirNickity wrote on 2020-02-24, 20:46:

Nothing mechanical means... no drive head movement? Or the motor doesn't even spin the disk?

If you are referring to my issue you're right, I didn't word it correctly. The disk spins, the light goes on but the heads don't move at all. You know when you don't have a disk in the drive and you try to access it and does that grrr sound before failing? It doesn't do that, no sound at all only the light.

Reply 11 of 17, by douglar

User metadata
Rank Oldbie
Rank
Oldbie
SirNickity wrote on 2020-02-24, 20:46:

Nothing mechanical means... no drive head movement? Or the motor doesn't even spin the disk?

Turns out my issue was the multi-IO card. I tried a 3rd Multi-IO/IDE card in the system, and it is booting from floppy now.

Tried a more powerful power supply, and I still have to hit reset to get the board to boot with a dx4 overdrive, so I guess that is what it is.

Reply 12 of 17, by Horun

User metadata
Rank l33t++
Rank
l33t++
douglar wrote on 2020-02-28, 02:47:
SirNickity wrote on 2020-02-24, 20:46:

Nothing mechanical means... no drive head movement? Or the motor doesn't even spin the disk?

Turns out my issue was the multi-IO card. I tried a 3rd Multi-IO/IDE card in the system, and it is booting from floppy now.

Tried a more powerful power supply, and I still have to hit reset to get the board to boot with a dx4 overdrive, so I guess that is what it is.

Glad you figured out the floppy issue. The DX4 issue could be a BIOS thing, some boards that supposed to support it just do not do it well iirc.

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Reply 13 of 17, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

Could be marginal bus timings or who knows what. I had a VLB build that, until I figured out the right card order, would sometimes boot, sometimes wouldn't. That's the thing about digital. It works or it doesn't, and unless you can see the waveforms on a scope or something, it seems kind of random. Maybe the reset circuit is just a little too quick on the draw for that particular CPU to be ready. *shrug*

Reply 14 of 17, by douglar

User metadata
Rank Oldbie
Rank
Oldbie

That motherboard is old and odd. I installed win98 se on it yesterday and at first I thought I could hear the HD noise, until I remembered that I was using a CF. Turns out the speaker was making faint chatter from the motherboard cross talk. It was much more digital sounding than the religious AM radio station I pick up on my land line telephone sometimes.

Reply 15 of 17, by Romain

User metadata
Rank Newbie
Rank
Newbie

Hello,

I've exactly the same issue with a MORSE KP 386 2.0 motherboard, in brand new very nice condition.

Just to add : no problem to boot an run with another HDD disk with already installed (Linux, DOS, Windows, etc) always no problem!

BUT when I come to A:\dir I've much times faillures of read (sometime dir ok but reading file KO - Abort, Retry, Fail, etc.).
And if I want boot with a floopy, it's work maybe .... 1 time on 30. Else, I've always messages like "Non system disk", sometime truncated like "Non syste" and system stop, etc 🙁

I've tryed with all my floppy drive collection (~12 units models, from the past to recent) and 3 IDE ISA different controller with differents chips. New unsealed floppy disk.
And of course a lot a differents cables en power supplys, RAM, VGA, FPU unplugged, timings, etc.
Many computer source for the floppy boot making, USB floppy disk, etc.

I'm going to be crazy because with the HDD, no problem !
But with the FDD ... it does'nt work properly.

@douglar : maybe you can give use the exact controller model to help us ? 😀

Maybe some news from the first post for an solution or idea ?
Thanksss

Last edited by Romain on 2022-03-07, 16:54. Edited 5 times in total.

Reply 16 of 17, by Romain

User metadata
Rank Newbie
Rank
Newbie

Some news about my case :

14Mhz crystal changed but without any effect :\

HDD boot and using : always fully OK
FDD boot : "Non-system disk" , etc (when the string isn't bugged by truncation or some random chars on the screen)

Reply 17 of 17, by Marco

User metadata
Rank Member
Rank
Member

I had similar issues with my floppy drive not working. The reason was to high isa bus speed and therefor also too high DMA clock speed. All due overclocking. All above 12 MHz was not welcome there

1) VLSI SCAMP 311 / 386SX25@30 / 16MB / CL-GD5434 / CT2830/ SCC-1&MT32 / Fast-SCSI AHA 1542CF + BlueSCSI v2/15k U320
2) SIS486 / 486DX/2 66(@80) / 32MB / TGUI9440 / LAPC-I