Yes, with an AT-style controller, it's necessary to enter HDD parameters into CMOS.

If you don't care about preserving the data, do the Low Level Format - from BIOS, or some other software, eg. HDINIT from Checkit 3.0.

I'm not even sure the hard drive works so that's what I would like to find out.

I set up the BIOS as
Cylinders 977
Heads 5
Precomp 300
Landing Zone 0
Sectors 17
Total size 41MB

I'm not sure Precomp and LZ are correct. Found the data on Stason.org... however, this video here gives different parameters but I haven't tried them yet: https://www.youtube.com/watch?v=ALpJm_IduX8

The hard drive is spinning but during POST it starts to produce fast rhythmic noise maybe once every 1/4 of a second so two per second.

https://stason.org/TULARC/pc/hard-drives-hdd/ … -MFM-ST506.html

You'll find correct data there.

Unless that's an RLL drive and controller...

https://stason.org/TULARC/pc/hard-drives-hdd/ … -RLL-ST506.html

wierd_w wrote on 2025-09-11, 14:58:

https://stason.org/TULARC/pc/hard-drives-hdd/ … -MFM-ST506.html […]
Show full quote

https://stason.org/TULARC/pc/hard-drives-hdd/ … -MFM-ST506.html

You'll find correct data there.

Unless that's an RLL drive and controller...

https://stason.org/TULARC/pc/hard-drives-hdd/ … -RLL-ST506.html

The first link is where I got my info from but no dice. Also, landing zone is blank for some reason. Maybe the hard drive has auto-park?

MFM rarely has such, as they are stepper based.

CSC Hard Drive Bible gives the same numbers.

Landing Zone doesn't matter - that drive has "Auto Lock Mechanism", which automatically moves heads to the proper place on power-down, no matter what's in CMOS.

Intriguing!

How does it actuate the head sled at poweroff I wonder..

wierd_w wrote on 2025-09-11, 15:53:

How does it actuate the head sled at poweroff I wonder..

I believe that at power-down, the spindle motor becomes to act as generator - its inertia is enough to provide the power necessary for the stepper to move heads.

As nobody explicitly mentioned it: The 16-bit controller card will conflict with IDE interface cards (except the 8-bit XT-IDE variant). If you use a multi-I/O card, disable the IDE port while the 16-bit MFM controller is installed.

Half-Saint wrote on 2025-09-11, 14:52:

The hard drive is spinning but during POST it starts to produce fast rhythmic noise maybe once every 1/4 of a second so two per second.

Does the drive make this noise even with no controller cables attached, just with power? If yes, the drive is likely dead.

Rhythmic noise, that sounds like the click of death, or the zip-zap of death, sounds like the drive is trying to ready, seeking out of park, return and repeat.
Some drives perform a more varied chatter, but if it never settles down, it sounds like it's bad, maybe bad servo track - and that's damn strange for a MFM drive to use a separate servo track, ah servo track and voice coil motor, so that explains the autopark, crazy modern technology for a MFM and the servo track would not be recoverable by a low level format

Ah, I can see it now - MR535 indeed uses voice coil...

Well that explains the autoparking..

wierd_w wrote on 2025-09-11, 17:51:

Well that explains the autoparking..

In this case - yes.

Though there was also plenty of stepper drives with auto-park - see eg. the ubiquitous ST-251.

I'll try to make a recording of the sound it makes. No matter what I do to the parameters, I can't boot off of it and I get thesame "C: hard drive failure" error during POST.

When I powered it on the first couple of times, I could only hear the spinning but after a while I realized that the hard drive was connected to the floppy connector by mistake. I'm also not sure, does it matter which secondary connector it's connected to on the controller side?

After I connected the cable to the correct 34-pin connector, the hard drive LED lit up and it started making some normal startup sounds. However, after a few seconds (after RAM count), it started to do that noise I described in the beginning.

Half-Saint wrote on 2025-09-11, 20:05:

I'm also not sure, does it matter which secondary connector it's connected to on the controller side?

You mean the 20-wire ribbon, right?
Then of course - the drive jumpered as first must be connected to the first connector on the card.

See also the diagrams here - https://www.minuszerodegrees.net/5170/cards/5 … abling_info.htm

I uploaded the sound recording and a photo of the card and the hard drive dip switches. I did not change any settings on the dip switches. I tried connecting the 20-wire ribbon cable to both J2 (on the left) and J3 (on the right). Now, either the drive parameters are wrong or something else is... I can enter the Hard Drive Utility in BIOS where I have several options including Low Level Format.

When BIOS parameters are empty, the drive doesn't make any sounds except for the spinning (startup hdd.mp3).

That drive is probably toast.

As for how to cable... here's hoping vogons doesnt clobber my bad ascii drawing. (Posting from a phone, no image editor)

1
2
3     1     2      1     2
4   [||]  [||]   [||]  [||]
5   [||]  [||]   [||]  [||]
6   [||]  [||]   [||]  [||]
7     ^     ^    [||]  [||]
8     |     |    [||]  [||]
9     |     |    [||]  [||]
10     |     |     ^     ^
11     |     |     |     |
12     |      L --- )----+--->HDD2
13     |           |
14     L-----------+---->HDD1
15

(Hopefully that looks right more places than my phone)

The cabling is done as 2 pairs. One long, one short, for each pair. I have labelled these with [1] and [2] respectively. The leftmost of each type is for HDD1, and the rightmost of each type is for HDD2.

The clicking in your sound file is the sound of the head seeking back and forth looking for the servo track, which it cannot find. This is a common mode of death in voicecoil based designs.

Stepper based designs have longer lives if cared for, because the head is moved in fixed increments from the motion of a stepper motor, and the entire disk surface magnetic layout can be recreated with a lowlevel format.

The downside is that they need precomp and landingzone, and need to be parked before poweroff.

I just discovered the most likely problem... destroyed 20-wire ribbon cable! It was chewed on by rats or cut by a knife so I now need to make a new cable or salvage the old one by moving one of the connectors.