• Please review our updated Terms and Rules here

Help with an HP tape drive?

jmetal88

Veteran Member
Joined
Jul 23, 2010
Messages
878
Location
Derby, KS
I have an HP 16-track tape drive, model 9144, that I'm trying to get working. The first thing I did was replace the capstan rubber wheel with an o-ring, as specified here:

http://www.hpmuseum.net/display_item.php?hw=257

It seems to have no difficulties advancing or rewinding the tape now, but after it makes it through a long sequence of whirring (and I've tried three of my tapes so far) it comes back with a fault. Back panel display says 'F3', then if I press 'Display Results', it will cycle through the following:

U0, A3, A2, A1, 23, U2

Looking at the hardware reference, I can't tell whether that means U2 (the Controller) has a fault. It's displayed in the results, but has no fault code following it, so I'm assuming no?

The rest of them seem to mean there is a suspected fault in either the Read/Write circuitry, the Device Dependent Controller (dunno what that is?) or the Drive Mechanism, or any combination thereof. The failed test is Test 23, a read-write test where it tries to write a value to a 'test area' of the tape and read it back, and those errors are returned if the value it reads back is different than the value it wrote.

Does anyone have any suggestions on what to start looking at in terms of fixing this issue? I'm sure I could work it out myself eventually (if it is fixable, that is) but it would be helpful to have advice from someone who has worked on these before, perhaps. One thing I did notice is that the head stepper potentially has issues, as in it was initially trying to move all the way to the bottom position and then keep going, emitting a grinding noise. It has actually stopped doing that at this point, though.

EDIT: Hmm, I do have a couple of write-protected software backups here. Since it won't do the read/write test on a write protected tape maybe I can at least see if it will read an existing tape this way.
 
Last edited:
Okay, after further testing, it seems like the drive is capable of reading write-protected tapes that are already written (up until it gets to any of the damaged bits, anyway). So, drive mechanism may be fine, read circuitry is probably fine, possible problem in write circuitry? Either that or perhaps my tapes are just too old to reliably record new data.
 
I'd verify that write current is present if possible. QIC's generally start shedding their oxide when they are no longer useable but the plugged heads would also interfere with read operations.
 
I actually don't know whether I want to work on the tape drive now. It'd be fun to play with, sure, but I don't think I have any practical use for it, at least regarding the writing portion. As for the commercial software I have, it looks like the bulk of it is in unreadable condition, and the one tape that may be intact actually seems to be for a 300-series machine and refuses to install on the 700-series I have.
 
Before you give up on the "unreadable" commercial software, consider sending it to me or Al K or Chuck G for recovery.
 
Hm, do you really think you'd be able to recover it? The tapes are in HP's proprietary 16-track format, and for each of them I can read the start of the tape and the first few files, and then the read fails.
 
I've got a 9144A here I can try to read them with. I can't figure out why it won't write anything yet, but it does certify tapes. It's plugged into a 300 series machine, so it's even appropriate. Let me know; I'll cover postage.
 
I want to try a few more things first, I think.

I just peeled off the o-ring I had glued on and put on a thick rubber washer instead, and it seems to have improved things somehow (I'm guessing because of the slightly larger outside diameter). I haven't tried to complete a tape read yet, but I actually got it to a 'pass' condition with a non-write-protected tape after just three insertions, so things are starting to look better. I'll have to fully test it after work. What it did just now was that, on the first insertion, it failed the same way as always. On the second insertion, it passed the first read/write test (23) but failed the second (24). Then on the third insertion it stopped the tape with a 'pass' condition. Crazy.

EDIT: Although it refused to let me unload the tape for some reason. Hmm.
 
>I can read the start of the tape and the first few files, and then the read fails.

I am not familiar with computer tape drives per se but serviced many tape drives in my career as a technician. If the capstan drive is not properly seated or wrong take-up tension,
you will get tape crawl up or down the capstan/head and failure as described above. Also plugged or worn heads may read ok but write fails. Writing requires better contact between
head and tape than reading. Also, pictures of the tape path mechanism are worth a thousand words ...

Larry G
 
>I can read the start of the tape and the first few files, and then the read fails.

I am not familiar with computer tape drives per se but serviced many tape drives in my career as a technician. If the capstan drive is not properly seated or wrong take-up tension,
you will get tape crawl up or down the capstan/head and failure as described above. Also plugged or worn heads may read ok but write fails. Writing requires better contact between
head and tape than reading. Also, pictures of the tape path mechanism are worth a thousand words ...

This is a QIC cartridge drive with PU tension belt. Really not much to make ti work but for a capstan pressing against the cartridge's drive wheel. Not at all like video or audio drives, since the data encoding is pretty much self-clocking.
 
I have to note a correction for one of my earlier posts. I said earlier that all three of my commercial tapes read up to a point and then stop, but after re-testing them, that isn't true. Two read up to a point and stop, and one doesn't read at all (I'm wondering if it's even the right format for the drive, but I don't know how you tell the difference between that issue and the tape being generally unreadable). Of the two that read, I can get about 1 megabyte off one before it fails and about 4 megabytes off the other before it fails. The failure points are pretty consistent, so I suspect the tapes are each just bad in certain places.
 
Oh, and if anyone's curious how the drive is doing with a blank tape since my previous status update, it will pass the self test occasionally now, but trying to actually write anything to the tape from the computer after it passes turns into an outright failure.
 
I've got an ADIC 530/1 drive that takes the same DC600HC/DC615HC Iotamat-formatted tapes. Also a 16/32 track drive. I suspect that the ADIC and HP drives are related--both large boxes for what they do.

The interesting thing is that the ADIC drive doesn't contain an optical sensor for BOT/EOT holes--it uses the formatting to tell what's what. Otherwise, it's pretty much a QIC-36 interface drive. There's a 6502 buried inside the drive and an EPROM with a 3M copyright on it.
 
I've got an ADIC 530/1 drive that takes the same DC600HC/DC615HC Iotamat-formatted tapes. Also a 16/32 track drive. I suspect that the ADIC and HP drives are related--both large boxes for what they do.

The interesting thing is that the ADIC drive doesn't contain an optical sensor for BOT/EOT holes--it uses the formatting to tell what's what. Otherwise, it's pretty much a QIC-36 interface drive. There's a 6502 buried inside the drive and an EPROM with a 3M copyright on it.

Same with this drive. No optical sensor even though all the cartridges still have the window and mirror. It just reads the formatting marks to figure out where it is.
 
I decided to try to see how much I could read off one of the school's old system backup tapes. Turns out I can get 12 megabytes off the one I tried before it gives up and says it can't read anymore, and looking at the reels (running the drive with the cover off), that's a significant portion of the tape.
 
HPMuseum.net has a page discussing problems with a belt internal to the tape cartridges going bad and how to replace them.(See link below)

"Over time, these quarter-inch cartridge tapes become problematic. Our experience is that about 50 percent of tapes up to 15 years old are readable. 35 percent of tapes between 15 and 20 years old are readable. Less than 10 percent of tapes over 20 years old are readable (almost no tapes are readable after 25 years)."

http://www.hpmuseum.net/display_item.php?hw=266
 
HPMuseum.net has a page discussing problems with a belt internal to the tape cartridges going bad and how to replace them.(See link below)

This has been beaten to death several times on cctlk. The simplest thing to do is steal a belt from a working cart.
12mb will be around the time the tape serpentines. If it's shedding, the oxide will pile up on the head and glue the
tape to the head when it tries to reverse direction dumping the oxide onto the tape, which then has an unreadable
block on every track. This is the failure mode for most old serpentine tapes; a bad block at the beginning and end
of the tape.
 
There are a few indicators that can give you a good hint as to the recoverability of a QIC tape--in other words, signals as to your success--or failure.

Even before opening the cartridge, I look at the brand. While both "3M" and "Scotch" brands were used for tapes made by 3M, I've had very little success in recovering old blue-label "Scotch" when compared to "3M" branded tapes of the same vintage. I don't know what the reason is, but that's my experience.

After you've opened the cartridge, take a close look at the two wheels at the top of the assembly as well as the capstan in the front of the cartridge. If you see dots of white or black stuff on them, you're mostly likely going to have a tour through sticky-shed hell. If there are uneven deformations in the part of the capstan that extends from the cartridge, the tape was having problems long before it got to you, as those are marks of the tape being stuck and the drive wheel wearing depressions in them.

Not all sticky-shed problems can be resolved with baking--in particular, the binder can often bleed through the oxide coating. That's nasty news. Sometimes, using a thin film of D5 lubricant can get you out of trouble, but it's not a cinch, by any means.

Mostly, the best solution, is to fire up the flux capacitor and go back in time and shoot the guy who thought this stuff up.

I've got some Scotch DC600A carts from 1985 to work on. They're obviously sticky; the belts have long disintegrated on their own and there's shedding on the internal wheels. I'm not optimistic and will do my best, but I'm not enjoying this.
 
Back
Top