• Please review our updated Terms and Rules here

IMI5012H hard disk specs needed!

sombunall

Experienced Member
Joined
Feb 18, 2010
Messages
154
Location
Montreal, Canada
I have an International Memories Inc. 5012H hard disk that I want to exercise. I haven't used it in a few years. I had an 8 bit Xebec controller that it plugged into. During a recycling run, I recycled the controller because I thought I could just use BIOS type 1 in an AT BIOS if I wanted to because I thought all 10 meg drives used those settings. I am wiser than that now.

On the drive I wrote 305, 4, 17. I guess I must have gotten this from disk manager but I'm not 100% sure. I found this table on the internet:
http://alasir.com/books/hards/06-68.gif

But the fields are blank and it says it is 15 megs and not 10 megs?

I do see the 5012 is 306, 4, 17, WP=214, RWC=307

Which brings me to another thing I need explained. In an AT BIOS I could specify the settings for write precompensation (WP) but what about reduce writing current (RWC)?

I have a few controllers, one is the WD1002A-WX1, I think the super BIOS version.

http://stason.org/TULARC/pc/hard-disk-floppy-controllers/U-Z/WESTERN-DIGITAL-CORPORATION-Two-MFM-ST506-412-driv-221.html

I have 10MB heads=4 cyl=306 WP=0 RWC=153 on that controller. But will it work? Is there no way to specify RWC in an AT BIOS without a specialty Xebec controller? Assuming I could even find out what the settings are for this HDD?

ALSO! Carlsson has this drive http://www.vintage-computer.com/vcforum/showthread.php?3085-PET-collection/page3

attachment.php

attachment.php

attachment.php
 

Attachments

  • IMI5012H_bottom.jpg
    IMI5012H_bottom.jpg
    102.3 KB · Views: 1
  • IMI5012H_back.jpg
    IMI5012H_back.jpg
    99.2 KB · Views: 1
  • IMI5012H_top.jpg
    IMI5012H_top.jpg
    98.8 KB · Views: 1
Well, the RWC setting for this drive says that it doesn't use the RWC line, so don't sweat it.

Try it with the ST412 settings--it'll probably work just fine.
 
Every reference that I can find (including a few printed on paper) say this about the IMI5012:

Model 5012
Heads 4
Cylinders 306
Sectors Per Track 17
Landing Zone 0
Write Precomp 214
Reduced Write Current 307
Encoding MFM
Capacity 10 MB
Interface MFM

Since RWC > Cylinders, it's a no-op.

IMI's 15 MB drive was the 5018--same basic mechanism as the 5012, but with 2 more heads. Maybe the last

This would be about the time that Onyx and IMI merged to form Onyx+IMI (one of the least imaginative names by far) with that building on North First Street. Their first 5.25" drive was about the size of a shoebox.
 
IMI's 15 MB drive was the 5018--same basic mechanism as the 5012, but with 2 more heads. Maybe the last
Yeah, same drive with an extra platter; I've got some of each. They were available with different interfaces (and conversion kits) which caused some confusion, as did the different capacities resulting from different SPTs (note the 32 sectors in the data sheet).

BTW, sorry 'bout my snarky Google comment; the official specs are actually a little incomplete and misleading when talking about using it in a PC-compatible; no precomp stated and 32 SPT is not a PC-compatible format. Thanks for the full story, Chuck.
 
Ok thanks for all the help! I am planning to use a 16 bit WD1003-WAH in a 386 now. If an HDD has Cylinders < RWC then does the WP need to be specified in the BIOS at all? Also what is the difference between WP and RWC? Shouldn't they be on the same cylinder?
 
WP = Write Precompensation. Basically, the diea is that as you move toward the inside cylinders, track length gets shorter and you start to get the phenomenon of "bit crowding", where writing two transitions close to one another actually causes them to spread apart. WP adjusts the write timing slightly to compensate (the algorithm is a little too long to go into detail here). Depending on the quality of your controller's data separator, it may not matter.

RWC = reduced write current. On inside tracks, some of the issues with bit crowding can be ameliorated by reducing the write current somewhat. Generally, this isn't as effective as WP. On your drive it doesn't matter--and both numbers are manufacturer's best guesses.

I'd use the ST412 definition as-is to start with. Most manufacturers recognized the importance and market share of the Seagate 10MB design and tended to make certain that their own drives could be drop-in replacements. Just common sense.
 
Last edited:
I have one of those same drives, havent tried hooking it up, but it has BAD written on the top
so I don't have a lot of hope it will work :)

Mine has an IBM front plate, so I think they must have used these in the early XT's.
 

Attachments

  • DSCF0057.jpg
    DSCF0057.jpg
    93.9 KB · Views: 1
  • DSCF0060.jpg
    DSCF0060.jpg
    88.7 KB · Views: 1
  • DSCF0056.jpg
    DSCF0056.jpg
    96.8 KB · Views: 1
I have one of those same drives, havent tried hooking it up, but it has BAD written on the top
so I don't have a lot of hope it will work :)

Mine has an IBM front plate, so I think they must have used these in the early XT's.
I think you're right; one of mine also had an IBM front plate.

I supported a number of systems (Cromemcos) that used these drives for several years; I don't know whether it was a friction/bearing problem or a motor/driver issue but a few of them developed a problem with starting up, but not with the usual 'stiction' issue. The disk was free to spin, but the motor just didn't seem to have enough torque to get it up to speed quickly enough to avoid shutdown.
 
Those were the days... ;-)

BTW, another instance where Wikipedia seems to have it totally wrong:
http://en.wikipedia.org/wiki/Write_precompensation

Yes, it's wrong. Doubltless it will wind up in some kid's term paper too. From the IDEMA Standards documentation:

Write Precompensation — The intentional time shifting of write data to offset the effects of bit shift in magnetic recording.

I can probably cite another 20 references if pressed to do so.

Whatever idiot at Wikipedia wrote his "definition" didn't understand much about magnetic recording. Increasing the write current is exactly the opposite of what you want to do.
 
Thanks for clearing that up Chuck. I looked it up in Upgrading and Repairing PCs but it only has a paragraph in the glossary and doesn't mention RWC.

Those were the days... ;-)

BTW, another instance where Wikipedia seems to have it totally wrong:
http://en.wikipedia.org/wiki/Write_precompensation

I just read that and cringed. I hope it gets fixed. I actually feel slightly shocked that Wikipedia could do this. :eek:

So I tried the drive on the aforementioned 16 bit WD1003-WAH and it found the data right away. When it started the bearings made this crazy sound for the first few seconds. Second time I started it was quiet. I think to myself did I format it before with this card? Is that Xebec controller compatible with WDC? Also disk manager would not low level format. It just did a high level format. I am going to do a DM /I to force it to low level format. I am also going to add a fan to ventilate the bottom as it gets hot.

Which beings me to another question. They say (ok it was probably Chuck right?) you should low level format it when you change controllers, mount the drive in a different attitude or just after so many years to be on the safe side but why? If everything is written new including the partition table with a high level format then why should it matter?

Which beings me to Yet Another Question. I low level formatted a 60meg RLL drive and painstakingly entered the defect map manually. Then I low level formatted it and the bugger make noises and went very slow over said sectors anyway? A bug in an earlier version of disk manager?
 
Don't bother reformatting. I'll bet that the original controller with this thing used WD technology, no matter what the controller brand was. i.e., "If it works, don't fix it."

I take it that you mean "...and then I high-level formatted it" (i.e. you used the FORMAT command). Yes, this is perfectly normal. The whole idea of the defect map is that bad sectors be more-or-less permanently marked that way, so you don't try to use them. There's a "bad sector" flag in the sector ID address header that can be set, so that any attempt to read or write the sector will fail. FORMAT still has to find these, so you get a bunch of retries.

Later on, when drives started getting smarts, bad sectors were "diverted" or re-mapped to spare areas. Most SCSI drives do this, as the SCSI standard pretty much mandated that every sector on a drive be readable and writable. But MFM and RLL drives are pretty much "brain dead".

That noise you hear on startup may be due to the little copper grounding clip that contacts the end of the spindle. Sometimes these start to vibrate with a most annoying whine. Another cause may be a dragging spindle brake (I don't recall if the IMI drives used one or not).
 
Don't bother reformatting. I'll bet that the original controller with this thing used WD technology, no matter what the controller brand was. i.e., "If it works, don't fix it."

I take it that you mean "...and then I high-level formatted it" (i.e. you used the FORMAT command).

Well I just used disk manager and it uses a kind of format for high level formatting. The thing is it did not even check for bad sectors it just wrote a new partition table, in advanced mode. It's version 6.02 if that matters.

OK then what do you think about this statement then Chuck:

Older drives needed to be re-low-level-formatted occasionally because of the thermal expansion problems associated with using stepper motor actuators. Over time, the tracks on the platters would move relative to where the heads expected them to be, and errors would result. These could be corrected by doing a low-level format, rewriting the tracks in the new positions that the stepper motor moved the heads to. This is totally unnecessary with modern voice-coil-actuated hard disks.
http://www.pcguide.com/ref/hdd/geom/formatLow-c.html
 
Well, my attitude still is "if it isn't broken, don't fix it".

A 10MB hard drive is very forgiving--if you were working with, say, a Priam 519 with 1224 cylinders (a much narrower track), it might make sense. So take that recommendation with a grain of salt. If your drive is showing a lot of errors that aren't on the defect list, go ahead and give it a shot.

One of my systems here is still running with a 36MB Quantum Q540. It has never been low-level formatted since it was put into service 25 years ago. It still works fine.

Now, please note that I think that anyone who doesn't back up their hard drive periodically is an idiot. So do the right thing and quit worrying.
 
Last edited:
Well, my attitude still is "if it isn't broken, don't fix it".

A 10MB hard drive is very forgiving--if you were working with, say, a Priam 519 with 1224 cylinders (a much narrower track), it might make sense. So take that recommendation with a grain of salt. If your drive is showing a lot of errors that aren't on the defect list, go ahead and give it a shot.

One of my systems here is still running with a 36MB Quantum Q540. It has never been low-level formatted since it was put into service 25 years ago. It still works fine.

Now, please note that I think that anyone who doesn't back up their hard drive periodically is an idiot. So do the right thing and quit worrying.
Couldn't say it any better than Chuck (as usual). I think the issue about periodically doing a low-level format isn't so much a question of the tracks on the platters moving, but wear and slack developing in the stepper and head mechanism, but as Chuck says don't worry about it unless and until you start having errors.

But in any case, if you're putting anything important on old disks like this (especially old IMIs) you should just assume that it's going to be dead the next time you turn it on, and back up accordingly.

And yes, AFAIR those drives did have a little solenoid brake with a felt pad radial to the flywheel and some (including this one) had grounding tabs on the spindle and some didn't; dry bearings could also cause noise, but these drives were never the quietest anyway.

Sounds like you're in business; have fun!
 
Well scandisk turned up 1 bad sector, presumably because I forgot to park it. I recorded the audio and EM of the whole thing, including spin up, errors, scandisk and spindown. That's another reason why I'm going through all my MFM drives now.

I also made a special mounting with a 92mm fan to cool the drives while testing. I used an old epson 5 1/4 bracket used for mounting those 720k drives with no face plate and meccano and the "enclosure" is very stable.

Wait a second. If format / disk manager has to mark the bad sectors after you enter the defect map then what's the point of manually entering the defect map in disk manager?
 
Wait a second. If format / disk manager has to mark the bad sectors after you enter the defect map then what's the point of manually entering the defect map in disk manager?

Because what FORMAT sees is the result of writing and reading back a sector. There's always the possibilty that a bad sector will be data pattern-sensitive and slip under the radar and be considered as "good" when it's not. By entering a defect list, the sectors are marked bad without a chance of accidentally being used.
 
Back
Top