• Please review our updated Terms and Rules here

IBM-5160 hard drive:Lemons to lemonade

Exceter

Experienced Member
Joined
May 17, 2019
Messages
286
Location
Vermont
A while back I purchased an IBM 5160 that had an IMI-5012H hard drive in it but, unfortunately, it didn't boot or spin up. After a quick examination of the computer, I found that the hard drive was making this noise: https://youtu.be/0TqyR2eXevQ

I then noticed that some of the screws on the lid were missing.
missing_screw.jpg

I decided to remove the lid and look inside because the drive was already making a horrible noise and it appeared that someone had been inside it before. This is what I found:
missing_platters.jpgbroken_heads.jpg

There's no way to fix this drive without replacement platters and heads and there's no way to get those parts without taking them from another drive and, at that point, you'd just use the other drive. I was a little discouraged at first and this might have been the end of the story but I noticed something interesting about how this drive was made. The drive itself is cradled within a frame which has rubber shock absorbers. When taken apart, the frame looks like this:
empty_frame.jpgshock_absorber.jpg

It occurred to me that I could fabricate an aluminum adapter plate which matched the hole patterns of this frame and of a modern 3.5" drive. After a little googling and measuring, I came up with this design:
cad_sketch.jpgcad_render.jpg

I sent the design out to get cut by a waterjet. Unfortunately, my measurements weren't perfect and I had to do a little trimming and drill a new hole(1 of the 4 holes is offset). However, with those minor modifications out of the way, I'm very pleased with the result. It keeps the exterior of the computer looking original, the new drive gets added shock absorbtion, and I'll even be able to use the hard drive light by connecting it to the HDD activity pins on an XT-IDE.
assembled_top.jpgassembled_bottom.jpg

I don't feel bad about repurposing the frame because the drive was 110% dead and because everything I've done is reversible, if necessary. Though, I can't think of a reason that I'd want to reverse it. I am now left with what's left of the drive and a controller which I can't bring myself to throw away. If anyone has any use for them as spare parts, I'd be happy to send you the controller and what's left of the drive for the cost of shipping. PM me if you're interested.
leftovers_top.jpgleftovers_bottom.jpg
 

Attachments

  • cad_sketch.jpg
    cad_sketch.jpg
    13.5 KB · Views: 1
  • cad_render.jpg
    cad_render.jpg
    7.9 KB · Views: 1
I assume that you'll replace the original with a 3.5" IDE drive, which means that you're also going to get an XTIDE (or similar) controller.

At this point, why not dump the hard drive entirely and use an XTIDE with a CF card?

(FWIW, there are prefab 5.25" to 3.5" drive adapters still being made.)
 
I assume that you'll replace the original with a 3.5" IDE drive, which means that you're also going to get an XTIDE (or similar) controller.

At this point, why not dump the hard drive entirely and use an XTIDE with a CF card?

(FWIW, there are prefab 5.25" to 3.5" drive adapters still being made.)
There's nothing to dump because a previous owner had completely removed the platters from the drive. I will definitely be using an XT-IDE to replace the original controller. I decided to re-use the frame because it has an original IBM branded full height bay blocker which wouldn't be compatible with modern adapters and I don't think they make this style bay blocker anymore. I don't want to use anything that's half height for aesthetic reasons. As for why I don't use a CF card, it comes down to personal preference. I'd prefer to have *some* hard drive seeking noises even if they aren't the same or even as loud as the original drives.
 
By "dump the hard drive", I meant dump the idea of spinning rust entirely. :)

However, it's your box, so have fun--but I'll be darned if I can hear a 3.5" drive made in the last 20 years seeking over the fan noise.
 
You could even do that with a small MCU attached to a speaker and is gated off the drive select. No need for TSRs.

Come to think of it, you could devise an MCU-based unit that shows the current head and cylinder, as well as read/write and noise. Include a CF card and it's a complete solution!

Given my yearning for the "good old days", how about a unit that periodically squirts hydraulic fluid on the floor? :)
 
As for why I don't use a CF card, it comes down to personal preference. I'd prefer to have *some* hard drive seeking noises even if they aren't the same or even as loud as the original drives.

I've been pondering the idea of hooking a piezo buzzer to the activity LED feed on my XT-CF card, but based on how quickly that light flickers I suspect the resulting noise wouldn't sound a whole lot like the real thing. (Some probably-flawed head math actually suggests the resulting sound might be a tone peaking an octave or two over middle C, if the "track to track" stepping time has any relation to the speed at which the light flickers. if it's just based on drive select... it's still just going to be pretty synthetic. And *brief*.)
 
Given my yearning for the "good old days", how about a unit that periodically squirts hydraulic fluid on the floor? :)

Then I could make a mint from selling replacement fluid cartridges! I'll be rich! RICH!!

Because I'm a software person, I only think in terms of software. Back of the napkin design: A TSR could keep about 20KB of long, medium, and short seek audio samples in RAM; then hook INT 13h AH=0 through AH=7 and make the appropriate short/medium/long squawk based on head velocity. Since a CF card is fast enough to make several requests while the same noise is still playing, the TSR would need to do nothing if audio is still playing.

This is quite silly.
 
Last edited:
You've just given me a terrible idea to create a TSR that simulates ST-225 seek noises out of a Sound Blaster.
That's not a terrible idea at all. I'd download it.

You could even do that with a small MCU attached to a speaker and is gated off the drive select. No need for TSRs.

Come to think of it, you could devise an MCU-based unit that shows the current head and cylinder, as well as read/write and noise. Include a CF card and it's a complete solution!
It's ideas like this which make me wish I wasn't a software guy. This would be a beautiful XT-IDE mod. Maybe someday I'll learn enough about hardware to take a crack at it.

...Given my yearning for the "good old days", how about a unit that periodically squirts hydraulic fluid on the floor? :)
LOL
 
Jim, if you dig back into the SIMTEL20 archive, I suspect that you'll find a TSR that provided an on-screen HD activity indicator, so the basic framework is already present. Back when, I wrote a TSR that wrote a log to hard disk of hard disk activity. I wanted to see what MSDOS was doing for large data transfers. It was an interesting puzzle--how to write to a DOS file in the middle of DOS file access by another task. :)
 
you'll find a TSR that provided an on-screen HD activity indicator, so the basic framework is already present.

Plus Hardcards flash a plus sign on the screen when the HDD is accessed. I'm assuming that was to be done in the BIOS, so something similar could be added to XT-IDE, right? Without the memory footprint of a TSR?
 
I don't remember that from my original computer. Maybe I didn't have the TSR to go with my Hardcard?

It looks like it's light.com for the plus sign and there is a sound.com that makes a noise. Now I guess I'll have to plug the thing in and see if the software is required.
 
It looks like it's light.com for the plus sign and there is a sound.com that makes a noise. Now I guess I'll have to plug the thing in and see if the software is required.

Would that TSR work with an XT-IDE or would it require an actual hardcard?
 
Would that TSR work with an XT-IDE or would it require an actual hardcard?

I got excited about this mention of the Plus hardcard and just a few minutes ago tracked down an image of the installation disk. Tried running "light.com" and "sound.com" on my Tandy+XT-IDE and while they don't produce any errors when I run "light on" or "sound on" they also don't seem to actually load or do anything, at least on my machine. They may well rely on either hardware or bios code specific to the hardcard.
 
Back
Top