• Please review our updated Terms and Rules here

RLL drive sampling project

Al Kossow

Documentation Wizard
Joined
Sep 1, 2006
Messages
3,471
Location
Silicon Valley
The interest in Seagate ST-27x drives revived my intererest in a project I've wanted to do for a while, to collect flux samples of RLL hard disks. A friend of mine got
a Maxtor XT2190 that appears to have been written with an RLL contoller. The problem is there is very little information on what that encoding actually is, so for the
last couple of days I've been formatting drives with various AT controllers that I own and have been collecting samples, along with making a list of controller read
performance. The performance is pretty bad, especially the WesternDigital controllers.
Data is up here:
http://bitsavers.org/projects/hd_samples
 
Al,

I have reformatted an ST277 (60 Mb RLL drive) to think that it is an ST251 (40 Mb MFM drive). I just used XXDP to low level format it on an RQDX3 specifying ST251 parameters and I know that some people have tried reformatting ST251 in RLL to come up as 60 Mb drives. As far as I recollect ST277 is a selected ST251 formatted and certified for RLL use.


Seems to work fine. It an alternative to reformat IBM drive I have in my Micro PDP11/73.

Peter
 
I've got the DTC and OMTI RLL controllers as well as the WD 1006V-SR2. The latter's performance seems to be fine, as your figures show. But it's been years since I've fooled with ST506 interface drives on these controllers. Ultrastor was apparently the best of the crop, but I don't have one.
 
I have a flux samples dump from a ST251 I formatted with a ACB4070 controller many years ago. It is from a ABC1600 and has a DNIX file system. I thought I would adapt the Davids mfm reader any day, but it hasn't happened yet.

thanks! one of the things that started this is I have a V185/ACB4070 combo out of a 3Com 3Server. I went back and made a transition file of the disk this morning and there is nothing on any of the tracks, so I guess I need to look at the read channel on the drive to see if anything is even coming off of the heads.
 
An RLL drive you can perfectly LLFormat to MFM, the other way round not preferable due to dense issues.
An RLL format squeezes the information more tightly onto the disc and increases the effective areal density by 50%.

Manual Saegate ST277R with Graphs specs and Flux density.
And find ST412 Manual with the Track format shown. Seagate ST412 OEM Manual 1982.

A good site with also Drive Information MFM and RLL, REDHILL MFM-RLL
Also ST277R manual:
Recording Method
----------------
The ST251 and ST252 are designed for operation with the ST412
interface with MFM encoding at 5.0 Mbits/sec. data transfer rate.
Operation of an MFM drive with an RLL controller is not approved by
Seagate and will void the drive warranty.

The ST277R and ST278R are designed to operate with the ST412 Inter-
face with Run Length Limited (2,7) encoding at 7.5 Mbits/sec. data
transfer rate.
 
People's comments seem to fundamentally not understand what I'm doing here. I'm perfectly aware of the differences between MFM and RLL encoding, their bitrates and limitations using RLL encoding on media/read channels designed for MFM. The point of this is there is no documentation on what the RLL flux-level formats are from controllers that shipped on PCs so you don't have any idea on how to write an RLL decoder for them or be able to tell from a flux-level recording what an unknown disk drive was formatted on. I've also discovered that there is low-level interchangability
between Western Digital MFM formatted drives and a couple of other controller cards that didn't use Western Digital's chipsets, the Everex EV-346 and NDC5525.

UNRELATED -- CHANGE THE "Write something..." box to be the same width as what you see here so the line wrap isn't screwed up.
 
Last edited:
Al, you've probably already done this, but where I'd start is with the RLL controller chip itself, say the WD50C12. The datasheet seems to have a pretty good level of detail, at least as a starting point. Just remember that even if the encoding is (2,7) RLL, that the recording is still NRZ.
 
An RLL drive you can perfectly LLFormat to MFM, the other way round not preferable due to dense issues.
An RLL format squeezes the information more tightly onto the disc and increases the effective areal density by 50%.

Manual Saegate ST277R with Graphs specs and Flux density.
And find ST412 Manual with the Track format shown. Seagate ST412 OEM Manual 1982.

A good site with also Drive Information MFM and RLL, REDHILL MFM-RLL
Also ST277R manual:
Recording Method
----------------
The ST251 and ST252 are designed for operation with the ST412
interface with MFM encoding at 5.0 Mbits/sec. data transfer rate.
Operation of an MFM drive with an RLL controller is not approved by
Seagate and will void the drive warranty.

The ST277R and ST278R are designed to operate with the ST412 Inter-
face with Run Length Limited (2,7) encoding at 7.5 Mbits/sec. data
transfer rate.

As far as I remember the ST251 was identical to the ST277R drive hardware wise. The ST277R was selected ST251 drives that were in certain specs. Since what is interesting when storing stuff on a harddrive is how close flux changes occur not the data rate prior to coding meaning that the rate of flux transitions is something else than the data rate prior to encoding. RLL differentiate between a bigger set of flux lengths thus storing more information for each flux transition.

The problem that occur then is that jitter in the drive speed would affect the flux length. My understandig is that the selected drives had a lower motor speed jitter which giving better margin for RLL coding.

it is perfectly possible to format any ST506/ST412 drive with RLL but it might be hard to read it back in some cases.

Back then it was quite common to format ST251 drives in RLL. In most cases it worked just fine and you gained 50% capacity. The rumor was that the selection also was based on demand. If there were big demand for 251 drives simply more of them went into that bin even though they might have passed as 277R.

Now thirty years later the drive might not be as good jitter wise as once uppon a time and it possible that is harder to read it.
 
Magnetic coating had a fair amount to do with things, also. Plated media tended to work better than older "powdered rust" coatings, particularly when the particle size is larger.
 
Back
Top