5150 era floppy drives and marginal media

[jasonwpacker]

I have a situation that's hard to understand regarding floppy drives for my 5150s, so I am opening it up to the hivemind.

Across two 5150s, I have two full height (1 Tandon, 1 MPI) and two half height (Shugart) floppy drives. All four of them are great at reading diskettes that I have written in the past but are unwilling to format a single floppy - including those that are already formatted!

I have tried each drive with both machines in an attempt to eliminate variables. Different main boards, different floppy drive controllers, different cables, different PSUs. Diskettes from two different companies and in the case of BASF, from two different boxes, one of which I only just opened from the original shrink wrap.

The diskettes are fine, and can be formatted in 360k on a 1.2MB drive in one of my 386's, though then the IBM drives refuse to read them. Plugging those IBM drives into the controller on the 386, I can also format the floppies. I'm pretty sure this exonerates the media to some degree (though it could be weakening but not yet failed magnetic media).

What am I missing here? If it can read disks with existing data on them without issues, could it still be either a timing/rpm issue, or a head alignment issue? I've cleaned the heads and greased the gears. Any recommendations for next steps?

 

[djg]

No great ideas. What happens when you try to write to a floppy you can read? Can you read the file written? What is the error you are getting?

 

[krebizfan]

Also, what DOS version is being used? Some of the later versions tried to prevent reformatting an already formatted disk or saving unformatting information which may not fit. Maybe trying an unconditional format "/U" could force the format.

 

[RetroAND]

Stupid question: can the drives write anything at all at the floppy disk when connected to the 5150?

If not, there's something wrong with the write circuit, maybe at the controller card.
If it does, maybe it's a software problem.

 

[george]

It is not the media. Since the drives are tested in your 386 PC, do the 5150 floppy controllers work in that 386 motherboard of yours with the IBM drives attached? Have you checked the voltages on your 5150 PSUs, especially +5V and +12V rails that go to the floppy drives?

 

[jasonwpacker]

Thanks for all of the responses. Let me summarize:

1. I can write to an existing formatted floppy and read from it just fine, regardless of which machine or drive I'm using.
2. I have tried both MS DOS 3.3 and 6.22 with the same results.
3. When I put the IBM FDD controller into the 386, it worked just as well as the UMC chipset card it replaced, both for reading, writing and formatting.
4. The drives can write to floppies and read the resulting files when hooked up to either 5150, each with a different controller card.
4. I did check the 5 and 12 volt rails on both of the XT power supplies during the tests and they didn't change by more than a percentage point or two when there was activity. I did not monitor the signals coming from the FDD controller card over the data connection but for power I feel comfortable saying that it's well within normal ranges - 5.xx and 12.xx throughout.

In all of the tests, I am running format or "copy con" while booted from the floppy itself, regardless of DOS version, so the PC software is consistent from test to test. Firmware might be another question altogether, but for it to just start happening out of the blue, on two machines, is odd at best.

The only other consideration I didn't think of, nor would I expect it to be an issue, is that while they're still using their stock clocks, both have NEC v20 CPUs.

I think I'm going to put an 8088 back into one of the 5150s and see if that has any measurable impact.

 

[SomeGuy]

Try some different floppy disk formatting programs. Some formatters can bypass DOS and BIOS issues, and may give you more meaningful error messages. The formatter included with MS-DOS is very brain-dead.

ImageDisk has the ability to format individual tracks. Might give that a try. It also has a tool to test RPM and general alignment, but if it all works on the 386, I'd suspsect a software problem.

Also, run some system diagnostics, such as Check-It, to verify the general operation of the systems.

What DOS configuration are you using? Is there a hard drive or XT-IDE? Are there any device drivers loading at boot?

 

[jasonwpacker]

At this point, the systems had been stripped down to additional memory, video and the FDD controller. I did run checkit without finding anything out of the ordinary.

And apparently the issue was gremlins? because now at least the Tandon will format the same problematic disks, while 48 hours ago none of them would. Gotta love this hobby...

 

[RetroAND]

Blame Gizmo! Stripe was right!

 

[Plasma]

The diskettes are fine, and can be formatted in 360k on a 1.2MB drive in one of my 386's, though then the IBM drives refuse to read them. Plugging those IBM drives into the controller on the 386, I can also format the floppies. I'm pretty sure this exonerates the media to some degree (though it could be weakening but not yet failed magnetic media).

If you have a DD disk that was previously formatted in a DD drive, and then write/format it with a HD drive, it won't be reliable in any other drive until it's degaussed ("bulk erased").

 

[george]

If you have a DD disk that was previously formatted in a DD drive, and then write/format it with a HD drive, it won't be reliable in any other drive until it's degaussed ("bulk erased").

You are wrong, it will be reliable in any properly (including factory) aligned 96 TPI (80 tracks) drive being HD or DD.

 

[modem7]

If you have a DD disk that was previously formatted in a DD drive, and then write/format it with a HD drive, it won't be reliable in any other drive until it's degaussed ("bulk erased").

You are wrong, it will be reliable in any properly (including factory) aligned 96 TPI (80 tracks) drive being HD or DD.

See the '360K 5.25" disks on 720K/1.2Mb drives' section of [here]. Is that something that you disagree with ?

In the IBM DOS 3.3 reference, IBM includes, "IMPORTANT: If you write on any of these diskette types using a high-capacity drive, you may not be able to read the diskettes in a single-sided or double-sided drive." ('Single-sided' and 'double-sided' being IBM's short way of referring to the 48 TPI drives that it supplied in the IBM PC family).

 

[SomeGuy]

Good to hear it is working now. Some times floppy drives just need to be exercised for a while before they will work reliably.

 

[Plasma]

You are wrong, it will be reliable in any properly (including factory) aligned 96 TPI (80 tracks) drive being HD or DD.

You are wrong. The disk will only be reliable in another HD drive if the alignment is exactly the same. Which is highly unlikely, and doesn't apply to OP's situation regardless.

 

[george]

You are wrong. The disk will only be reliable in another HD drive if the alignment is exactly the same. Which is highly unlikely, and doesn't apply to OP's situation regardless.

You are wrong again as usual. Have you heard of tunnel erasure, and what it is for? Why don't you try with real floppy drives, these days they are cheap unlike back then.

 

[Plasma]

You are wrong again as usual. Have you heard of tunnel erasure, and what it is for? Why don't you try with real floppy drives, these days they are cheap unlike back then.

You are making a fool of yourself again, as usual. Just because you've done it before and it worked for you doesn't mean it's reliable. And it's definitely not good advice to give anyone else.

 

[george]

If you don't know what tunnel erasure is for, then consult some books and report back. If you can't, I will do it instead of you. I will explain why you are wrong on this matter, and my knowledge is not empirical at all, as you are trying to shift the focus of what I asked you, and with that very question I am actually giving a cue why you are wrong. So far, I have witnessed several wrong advices of yours in this forum. For the previous one in another topic I made only a hint, and you slightly revised your advice but it is still most likely wrong. The time and unnecessary purchases of the OP will prove that. I am interfering usually when others are providing very misleading information (almost always unintentionally).

 

[Plasma]

If you don't know what tunnel erasure is for, then consult some books and report back. If you can't, I will do it instead of you. I will explain why you are wrong on this matter, and my knowledge is not empirical at all, as you are trying to shift the focus of what I asked you, and with that very question I am actually giving a cue why you are wrong. So far, I have witnessed several wrong advices of yours in this forum. For the previous one in another topic I made only a hint, and you slightly revised your advice but it is still most likely wrong. The time and unnecessary purchases of the OP will prove that. I am interfering usually when others are providing very misleading information (almost always unintentionally).

I know what tunnel erasure is. As for the other thread, RAM and ROM are both memory. Your "hints" have nothing to do with my posts.

I suggest you take a break. Your obsession with me is unhealthy.

 

[george]

I know what tunnel erasure is. As for the other thread, RAM and ROM are both memory. Your "hints" have nothing to do with my posts.

I suggest you take a break. Your obsession with me is unhealthy.

I am all ears. Had you known what tunnel erasure is for you wouldn't have claimed this just a few hours ago:

The disk will only be reliable in another HD drive if the alignment is exactly the same.

 

[Plasma]

I am all ears. Had you known what tunnel erasure is for you wouldn't have claimed this just a few hours ago:

And yet I did. Why do you think that is? Have you considered that there may be variables which impact the effectiveness of the erase heads?