TZ30 / TK50 / CompacTape adventures

[Mal]

I've been playing around with a DEC TZ30 CompacTape drive this week - so far without much success.

The drive itself seems to be working fine. It's installed in a MicroVAX 3100 Model 90 running VMS 5.3. It seems to be a fairly robust design - no rubber rollers anywhere in the tape path, and no drive belts. The two spools are direct-drive and the only other items in the tape path appear to be a couple of stainless-steel capstan/tacho assemblies. The to-and-forth jogging that the drive does when a tape is inserted is quite interesting to watch.

The big problem I'm having is residue build up on the heads that causes the drive to be incapable of reading a tape. For example, if I MOUNT and attempt to do a DIR on the DEC VMS 5.1 distribution tape shown in the attached photo, I get to see the DIR listing of about the first 9-14 files, but the DIR operation then aborts. The tape can't be read again successfully until I clean the heads with a cleaning probe bathed in isopropyl alcohol. The third picture shows the fouling that occurs on a single read attempt. The same volume of residue continues to accrue on each attempt, even after about 10 attempts.

Presumably I'm dealing with disintegrating media here? Is this a common problem with TK50 tape (CompacTape)? If so, is there a known work-around (baking the tapes etc)? Or should I write this off as being a futile exercise?

I'm keen to get this working if I can as I have several hundred tapes to explore (some of which may have material worth archiving) and a number of early-90s VAXes with CompacTape drives installed.

 

[glitch]

Ah, the TK50...

I decided to try and get going with some TZ30/TK50 drives and TK50 cartridges when I acquired my MicroPDP-11/73, since it came with a TK50 and controller. I had moderate success with the TZ30 under Linux, but no luck with the TK50, unless jamming and unspooling tape cartridges counts! I did find that heating the TK50 cartridges greatly reduced the amount of residue on the tape heads. I simply placed them on top of a radiator for a while -- this was in winter, so we had the heat on.

My personal opinion is that archiving is practical if you have a TZ30, and worthwhile if you have tapes that haven't been archived before. I would not attempt to use TK50 tapes as a means of actually doing useful work though -- far too much headache, and if you already have a TZ30 you should be able to substitute a more reliable SCSI tape drive. I've been using old DDS1 4mm DAT drives/tapes with my various machines that want tape, until I can get time to get my 9-track Pertec interface drives up and going. DDS1 4mm DAT is cheap, available, and seems to survive pretty well. I have a number of NOS and used tapes that all operate fine, and a handful of well used tapes that were intermittent.

 

[Al Kossow]

Ah, the TK50...
My personal opinion is that archiving is practical if you have a TZ30, and worthwhile if you have tapes that haven't been archived before.

That has been my experience as well. The TZ30's big advantage over the TK50 is it is easy to pop out the head stack to clean it, which needs to be
done after every tape. Because of the construction of the cartridge, they aren't easy to bake. With 3M 1/4" cartridges, they come apart and go together
easily so you can bake them en mass in a commercial food dehydrator.

TK50 cartridge media is very sticky at this point. Without some media treatment, you will glue the head to the tape and tear off the oxide after the first pass and the tape
tries to reverse direction.

 

[Mal]

Al, Glitch,

Thanks for the feedback. Very helpful. I will try putting a little warmth into a couple of the TK50 tapes, but I fear they are probably beyond help, other than perhaps for a one-pass archiving attempt.

I've spent more time on this today using a batch of used CompacTape II tapes. I played around with about 8 second-hand tapes that (I now know) were formatted in TK70 format. My results with those tapes in a TK70 drive were much better. No stuck tapes and no read or write errors (though I had a couple of tapes that took a full 25 minutes to complete a DIR command). I was also able to demagnetize the start of the tape, so I could then INIT them in the TZ30 in TK50 format (I discovered today that the TZ30 outright rejects CompacTape II tapes that are in TK70 format). They then operated in the TZ30 without any faults or failures. After about 4-6 hours of experimenting, there was a slight but not significant residue build on each drive.

Glitch - I'm interested in your Linux experiments with the TZ30. I tried using dd to copy VMS 5.5 onto a TK50-format CompacTape II using the TZ30 tonight on an Ubuntu machine. It's a 13MB file. The TZ30 seemed to be doing a write-verify pass on each block (lots of frequent direction-reversal going on). I interrupted it after 13 minutes. By then it had only copied 270 blocks (135kb) - about 1% of the file. Did you have any success copying images on to the TZ30 without these frequent direction reversals?

I'll also try some 1/4" tapes soon, but because I have machines that don't have SCSI, I'd also like to be able to do software installs from the TK50/TK70 (or maybe I can learn how to do a VMS install over Ethernet - haven't worked that one out yet).

 

[bear]

Because of the construction of the cartridge, they aren't easy to bake.

Really? I've had totally usable results just chucking them in the dryer whole, three per tray arranged so each holds another's dust flap open. I've done more than a hundred like this and can get through five or so tapes on average before needing to clean the heads. More if I bake them longer.

But you're right, if you don't do anything to the carts they will totally jam up the drive in nothing flat. And THEN you're in for a world of hurt. (BTW it's just that the motor drive logic won't tolerate any kind of drag -- in most cases if you insert a piece of paper between the tape and the head when this happens, you'll at least be able to get the drive to unload the tape again without a whole lot of swearing.)

Mal-

Frequent direction changes indicate you're not keeping the tape buffer full, and it's in start-stop mode. This is bad for everybody. Increase the output blocksize to 8k, which is what VMS BACKUP is going to expect anyway.

 

[NF6X]

Thanks for suggesting the paper trick, bear. My first attempts with a TZ30 drive (stupidly, without baking the tapes) were disastrous. I had to unload and re-spool the tapes in hard mode; that is, manually, after disassembling the stuck cartridge in order to carefully spool many yards of tape back out of the drive. If and when I decide to play with TK50 tapes again, I hope I'll remember the suggestions in this thread.