• Please review our updated Terms and Rules here

DEC Rainbow HDD problem

Well, it is possible that track zero is bad, but you do also have the head postioner stepper sticking also. It is possible that the stepper drivers are not driving hard enough, but I doubt that one of the pole drivers is completely not firing, because you say that the motor still wants to turn when it gets stuck. You could watch the poles being driven on a scope (if you have a 4-channel scope) to see if one is weak. I think the stiff grease is still holding things up though.

Lou
 
I don't have a 4 channel scope so measuring that looks like it is not possible. I will keep moving the motor back and forth some more.

If this does not fix it are there any other things I could do? Can I use a different type of drive with the same 2 cable interface or is the electrical interface different for different 2 cable drives? Also if the drive ends up having some bad sectors do you think that freezing the drive will help any? I have had it make some modern drives function again, and it even eliminated some bad sectors on an old half height Seagate drive.

EDIT:
I think that the drive may actually work. I attempted a re-format (I do not know if it was low level or not), and the head seemed to step properly going upwards, but when it wanted to return it started to act up again. I think I just may need to move the flag back and forth some more.
 
Last edited:
I tried to run a format again. The disk made it up to the maximum range away from the photogate just fine, but when it tried to return, it would jam. Would the grease jamming the motion affect it in only one way or should I start looking for a different problem? Could the electronics that drive the stepper motor be malfunctioning? Since the format did not fail at the start I think that is a good indication that track 0 is not actually bad (could I be incorrect in this assessment?).
 
I tried reseating the ribbon cable that connects to the stepper motor to see if there was a weak connection. The next time I powered the computer on, instead of saying system load incomplete it said that it was a non-system disk. I was also able to get farther along with the format, but I still had to help the stepper motor return.

The screen flashed then the computer was frozen and nothing on the screen with all lights on the keyboard lit up. I did not see the lights on the back of the system to see if they indicated anything. I restarted the computer and I received message 2: main board. I forgot the diagnostic lights that were illuminated, but they matched the message. I then turned off the system and unplugged the HDD. I waited a few seconds then turned the system back on and it was fine. Should I worry about this, or do you think that it was just a passive glitch?

I am beginning to think that the problem might lie with the motor or the driver for the motor. What can I do to check to see if the output of the driver is good?
 
Rainbow... That screen flash business and lockup sounds like an over stressed power supply.

If you're formatting the drive to exercise it - won't it end up "non-system" if you're successful?

Help me out here.
 
I am formatting the disk both as an exercise and to see if the surface is still okay since I have been having trouble with it. And yes I believe that it would say non-system disk after that. The notable thing about it is that before it always would system load incomplete, even after previous format attempts. All the formats that I have done on the drive end up being unsuccessful because of the stepper motor trouble.

I will double check to see if the power supply is the one rated to handle the drive.
 
Ok, sounds good. I must confess that it took me a bit to remember you were talking about a Rainbow here. (My head went to XXDP at first)

I have one of these drives myself but haven't spun it up in quite a while.

Lou's idea of getting the mechanics loosened up is probably what's needed. I'd prefer to exercise the drive to health, just by letting it seek from one end of the platter to the other. (maybe 24 hrs?) I wonder how difficult it would be to write software to do that?

Formatting wouldn't be my first choice, but you gotta use what you have.

I know Lou has repaired drive electronics gone bad, but I think it was in the spindle driver section.

I have often worried about lubing these older sealed MFM drives. How to clean them of old grease, and what to use to replace it with.

Back in the day, I had an IBM 20MB MFM drive in an original PC-AT go bad. I replaced it with a pair of ST-251s. One to expand the system with, the second to hold nightly backups. So now I had a bad drive to mess with.

The 20MB drive with it's data now lost (yes... I too tried to format it as a diagnostic) began to make such horrible noises I just had to see what the problem was. I opened the platter chamber to discover the drive's internal filter (a molded plastic and paper part) which had been stuck to the inside roof of the cover by a sticky tape adhesive, had dropped loose at one end and was rubbing on the surface of the top platter. The inside was filled with plastic dust from the abrasive action of the platter against the filter.

After lamenting some lost data, I figured - what the heck. I took extreme caution to clean everything, including the filter and re-assembled the drive to see what would happen. I think I screwed the filter back in place and sealed the screws with loctite.

I couldn't believe it was serviceable again, but it was. Formatted without errors (once I filled in the bad block table by hand) It was still working when I left the company a couple years later. (LOL - just not in my machine)

Since then, the prospect of opening one up to clean and lube things isn't nearly as scary as it once was. I guess it's just a question of how desperate you are, and how screwed the thing is if you don't try.

I've tried, without success, to find out what techs used to perform this kind of service when it was being done back in the day. It would be good to know now, even what the specs are.

For the grease - a non-evaporating, hydrophobic lube. I'm thinking some kind of silicone stuff?

To remove the old stuff, I expect some kind of zero residue solvent like break cleaner. (Concerned about the plastics though - are bushing parts metal?)

Is this all to scary?
 
Okay. I will just keep trying to get the drive to move back and forth. I don't want to risk damaging the drive so I will not take it apart. I will continue to try and exercise the stepper motor.

If it does end up not being just stiff grease on the motor, where should I start looking on the drive's board?
 
Old Computers said:
If it does end up not being just stiff grease on the motor, where should I start looking on the drive's board?
Click to expand...

Look at the pole drivers for the stepper motor. Look at the power to the coils to see if they are all being driven equally as hard.

I like RSX11M+'s idea of exercising the drive's heads for a long time. Spinrite on a PC would be good for that.

The RD51 is pretty robust from the standpoint of tolerating being opened in the field. If you open it in a very clean room of your house, and don't touch the platters, you will be fine.

Lou
 
Drive Maintenance Reverse-Engineering

Drive Maintenance Reverse-Engineering

With respect to Sealed "Winchester" drives:

I guess my point in all that previously was: Individually, we're ultimately going to try all these techniques because - What else are you going to do with a failed drive? (door stop?)

So if we can pool our experiences, it may help us figure out which of these things work, and which don't. Particularly in the area of lubricants, where there seems to be so little information (and so much of the problem), but in other areas too.

Some thinking along these lines:


  • Why does grease congeal?
  • What role does moisture play in the process?
  • Does grease outgas a solvent over time?
  • Does mechanical movement "reactivate" or "de-congeal" grease?
  • What are it's lubricant properties after being "freed"?
  • Would increased temperatures be of benefit in freeing shafts and bearings with congealed grease?
  • Can a solvent or oil help "reactivate" congealed grease?
  • What lubricants are compatible with each type of bearing? (plastic, rubber, bronze, etc...)
  • How to identify which lubricants won't evaporate or spread over time?
  • How do platter shaft lubricants differ from those needed for head positioning mechanisms?

Unsealing and Resealing considerations:
  • How clean an environment do I need to open a drive?
  • How can a work area be best prepared?
  • What are common sources of dust?
  • How can dust be removed from a drive's insides?
  • Is it useful to clean a drive's internal filters?
  • Possible filter replacements?
  • Gaskets or sealants?
  • can mechanical orientation help during cleaning and resealing?
  • Inspection and diagnostic techniques?
  • Tools and precautions? (magnetism?)
  • Supplies?
  • Reference material?
  • Records?

Perhaps the most important thing we can share is our experiences, but also is the internal makeup of each drive type. Knowing what one may expect in opening and while being inside (by make and model) could be of tremendous help to a community if sharable. I can think of no better beginning than a photographic or video record of anyone making the attempt.


We also need to develop a clear and concise "Don't Do" list - particularly suited to nubes.

Gee this is starting to sound like fun.
 
Well, I'm a noob as far as dismantling hard drives is concerned, but I would absolutely love to see any of this happen. I've got a couple of pseudo-dead ESDI drives I'd love to take a crack at.
 
It would be quite wonderful to have a repository of HDD info. These old drives are getting quite old and it would be nice to be able to maintain them for longer. Thankfully they seem to be well made, to have lasted this long.

For exercising the drive what controllers would work in another computer? Would another Seagate controller work? Or was SpinRite ported to the Rainbow?

For measuring the pole drivers and the power to the stepper motor what do I do? I have never worked with a drive at this level before. Would I just take the volt meter and measure the pins going out of the board to the ribbon cable that connects to the motor? If that is the case what is the ground? Please let me know if the measurement is different.
 
I've already exchanged PMs on the idea of collecting bad drives to begin this work. What follows is an excerpt from one of those.

For my purposes, marginal drives may ultimately be more instructive than totally dead ones. However, the first activities are best done with drives that are doorstops.

Ideally, it would be best to disassemble and document at least one of each of the drive types used by DEC in original equipment. For this task, totally dead drives would suffice as long as they are complete, and are not mechanically damaged. These will serve as a reference set. Whether repairable, or not doesn't matter at this stage.

That experience will result in a set of instructions on how to disassemble drives without harming them, and provide a basis to proceed with work targeted at restorative investigations.

Next, a catalog of issues needs to be developed, and diagnosis of each so that remedies can be attempted and judged. For this "marginal" drives are needed where testing can be meaningful. Some of this can be collected remotely with software we will write. That will work for problems we already understand. For inscrutable cases [or ones that require the HDA to be opened] disassembly and inspection / diagnosis in a cleanroom environment will be required.

Ok, so that's the outline.

However, there is a downside. This would be that no assurances can be made about the outcomes, particularly for those drives involved in the early process. Undoubtedly, there will be damage caused by our efforts. We have to assume some of it will be fatal to the hardware, and unnecessary if we had wisdom to avoid it at that time.

However, if we don't try - if we don't work through these issues, we can be certain that all drives will eventually fail and none will be salvaged.



If we expect to benefit from this work, we're going to have to take risks. It also means getting voluntary participation. There might be some who could economically benefit from the work eventually. But the work cannot begin on that basis.

I guess we need to develop a general framework for participation that recognizes these items have value. I can't quite get my head around how that would work though.

Do you have any suggestions?
Click to expand...
 
I think that RD51 should certainly be the first drive to go through this highly documented teardown. It's interesting, I have had similar discussions with a fellow I met from alt.sys.pdp11 for a few years myself. He and I have focused on RD51, but have not given any up for science. We have done our best to try to study this drive without actually taking it apart (other than taking the top cover off) (I even brought one to work to look up inside the motor with a fiberscope to see if I could get to the bearings that way). We have also had discussions about grease and most importantly how to get something to rejuvinate it to the bearings where it's needed. This last point has been particularly troubling on RD51's which tend to screech and howl from the spindle bearings. We are at the point where the platters likely have to come out to get to the bearings.

All that may be a getting a little ahead though. As has been suggested above, some structured planning about goals may be in order.

Lou
 
Last edited:
Old Computers said:
For exercising the drive what controllers would work in another computer? Would another Seagate controller work? Or was SpinRite ported to the Rainbow?

For measuring the pole drivers and the power to the stepper motor what do I do? I have never worked with a drive at this level before. Would I just take the volt meter and measure the pins going out of the board to the ribbon cable that connects to the motor? If that is the case what is the ground? Please let me know if the measurement is different.
Click to expand...

I have an AT ISA bus seagate controller that I have in an old dec PC425 that works nicely for this duty. As for the stepper motor drivers, I would look at them with the scope. Then you can look at pulse height, width, and timing. You will need to identify which motor lead is common (ground) and which go to the windings. I don't have an RD51 out on the bench at the moment. I have the feeling though that we'll be looking at MFM drives on the bench this winter.....
 
Okay. I will work on that. I have made another development. When I attempted to start up the system again to run another format the system gave me message 2 again, and the lights were the same pattern (all lights lit up except for 5 and 7). I disconnected the power to the hard drive and turned the computer back on, and it was just fine. I am wondering if the HDD is drawing too much current or something in the power supply is going bad. It has only happened twice to me (the first time I also tried removing the HDD), but I think it is enough to conclude that something is up with this.

I have another computer with an ISA Seagate controller, but I think that it is an RLL controller. If I remember the drive in that system has R appended to the name. Could this controller handle the ST-412?
 
Hello All,
I am following closely this very interesting file, mainly about maintaining / fixing OLD DEC H.D.
I would like to "join the team", if something goes on on this subject.
BTW, right now, I MAY have the opportunity buying an ST-506, working state not known ( but does NOT seem junk ).
In your opinion, what would be a fair price for it ??
 
Gerard,

That drive was known in the dec world as RD50. I don't think that many ended up in systems. The RD51 (ST412) was wildly popular however. If I saw one I would spend no more than $20USD, so that if it suffered from the loud spindle bearings I would not be too sad.

Lou
 
Thanks, Lou.
Well the price is already $20 **BUT** I think that *IF* if it suffer from spindle bearing noise, *THEN*, for no more that $20, it may be a good candidate to "start the job" :
I mean :
- Carefully open it and see how one may exchange the bearings

*ELSE*, it will find a place in my "coming soon ( I hope )" working sytem.
 
You must log in or register to reply here.
Back
Top