The Museum of HP Calculators

HP Forum Archive 20

[ Return to Index | Top of Index ]

9830 tape drive question
Message #1 Posted by David Ramsey on 17 Aug 2011, 11:49 a.m.

I recently snagged a beautiful 9830a from a long-time HP reseller in San Jose. The machine is in spectacular cosmetic condition; it looks as if it came out of the box just a few months ago.

It has the 8K memory upgrade and I grabbed a slew of ROMs as well as an equally-snazzy 9866 printer.

Everything works great except the tape drive. I can read the pre-recorded Math and Stat packs reliably; the system test cassette reads perhaps 20% of the time, and although I can successfully MARK the two NOS HP tapes I got with the calculator, I've never been able to write anything to them and read it back without getting an ERROR 59 (checksum error) or ERROR 58 (misc tape error: door open, STOP pressed, clear leader problem, etc.) In the latter case there is no obvious physical problem I can see, and yes, the internal tape light works.

I've cleaned the head with alcohol and wiped it down with a microfiber cloth.

Now, this could just be cruddy 35 year old tapes. I've ordered new tapes and a head deGausser which I should receive next week.

The tape drive makes a continuous "tap..tap..tap" sound as the tape is read or recorded at normal speed. With the top of the computer off, a finger on the driving spool motor will feel each "tap" distinctly. Is this sound normal? And does anyone have any suggestions for things I could try?

Re: 9830 tape drive question
Message #2 Posted by Tony Duell on 18 Aug 2011, 4:37 a.m.,
in response to message #1 by David Ramsey

No, tht sound is not normal, and may be part of the problem.

Unlike audio cassette recorders, the HP9830 tape drive doesn't use a capstan and pinch roller, it just drives the spools. The tape speed is not constant therefore, but that doesn't matter as the data encoding is essentially self-clocking.

Anyway, there are 2 motors inside the drive on a rocking beam, which is moved by a pair of solenoids, one each side. Each motor spindle can be forced against the rubber tyre of the drive spindle. My first guess is that these tyres, which are simple O-rings, have decayed.

I would take out the tape drive, open in up and look at it.

Start by removing the top cover and the keyboard, then undo the fixing screws and lift the tape drive up and out, it's plugged into an edge connector on the backplane. Take out the screws on each side and ease off the mounting bracket. This carries the connector PCB, and the various parts of the tape drive are connected to this by 'plugs' which are actually little PCBs fitted with pins. Carefully unplug all 4 of these.

There's a metal bracket across the back of the motor assembly. It's held on by 5 screws (2 each side, and one in the centre). Take this off, and slide the motor assembly out too. You can now see the drive wheels and the tyres. There's a sliding brake plate which is sprung against the tyres when the drive is stopped, and this might have made notches in said tyres.

Normally you don't have to remove the drive spindles themselves, but if you want to, take off the front cover next (1 long through-bolt, slide out the window before it falls out, and unscrew the ground wire). Then take off the leaf spring from the back of the drive spindle (1 screw), the circlip on the front side, and the plastic washer. The spindle will slide out rearwards.

My notes suggest a size BS219 O-ring is the right part to use.

Feel free to contact me privately for more information if you need it (and for that matter about any other aspect of the 9830 and 9866).

Re: 9830 tape drive question
Message #3 Posted by David Ramsey on 18 Aug 2011, 4:22 p.m.,
in response to message #2 by Tony Duell

Why, yes, there is a small notch in one of the O-rings:

That probably accounts for the sound. Guess I'll order some more O-rings. I'll let you know if that makes a difference.

Re: 9830 tape drive question
Message #4 Posted by David Ramsey on 19 Aug 2011, 11:41 a.m.,
in response to message #3 by David Ramsey

Pending the arrival of new O rings, I swapped jthe existing rings to each other's wheels.

The rings were soft and flexible, indistinguishable from new except for the notch in one of them. Perhaps the computer was stored in an inert atmosphere.

The tapping noise is greatly reduced, although still present. I can now reliably read the systems test cassette, although there's still no joy marking and recording and reading back from a blank tape. About half the time, doing a MARK 3,5000 results in the tape slowly moving all the way to its end, and a subsequent ERROR 58. On the occasion that the MARK operation appears to succeed, attempting a STORE operation will result in ERROR 59 (checksum error); if that succeeds, the error will occur on the LOAD operation.

Hopefully this is bad tapes or a head in need of deGaussing.

Re: 9830 tape drive question
Message #5 Posted by David Ramsey on 19 Aug 2011, 10:07 p.m.,
in response to message #4 by David Ramsey

Dang. New tapes and deGaussing had no effect; although I will replace both O rings, I suspect that won't have any effect, either. Further suggestions solicited...

Re: 9830 tape drive question
Message #6 Posted by Tony Duell on 20 Aug 2011, 4:30 a.m.,
in response to message #5 by David Ramsey

OK, it seems like the reading problem was due to the O-rings, since you can now read the pre-recorded tapes you have.

This also pretty much means that the head is OK. As you probably know, the head has 2 windings both of which are used for reading and writing. And much of the electronics must be working too.

The tape interface is one section of the machine which was supposed to be repaired to component level in the field, so the service manual (available from the Australian museum site) has some troubleshooting information. But I will try to provide some more here.

The unit writes 2 tracks. A pulse on one track is a 0, a pulse on the other track is a 1, and a pulse on both together is a byte marker. Tape words are 9 bits long, the middle bit (and thus in the same position relative to a marker no matter which way the tape is moving) is used to indicate a file header.

I don't think the 9830 verifies after writing or marking, so it's perhaps not surpising you don't get errors there.

You want to start on the Read/Write PCB, 09865-66563 (blue and orange handles, of course). When doing a write operation, look at TP2 and Tp3 with a 'scope or logic analyser. These are the outputs of the write encoder circuit. You should see pulses on both of them, as described above.

Now connect the 'scope or LA to the otuputs of the read amplifier. That's the outputs of U4a and U4b -- page 81 of 'my' schematic. Read a good tape to see what to expect, then try to read one of your tapes. See if either signal is missing (or garbled).

If one channel is bad, it shouldn't be too hard to trace the problem. Possibly the write driver circuit on the tape drive head PCB (page 87 of my schematics).

Re: 9830 tape drive question
Message #7 Posted by Tony Duell on 20 Aug 2011, 4:39 a.m.,
in response to message #6 by Tony Duell

It's also possible the write clock is running way off frequency. When marking a tape (or similar), look at TP1 on the tape controller PCB (09865-99652, blue and red handles, page 79 in 'my' schematic). It should be about 3kHz.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall