HP 9825a

[Paul Dale]

(03-23-2017 04:27 AM)sleibson Wrote:  The HP9825 printer story is here:

An extremely interesting read. Thanks for posting it.


- Pauli

[Craig Bladow]

I have been enjoying a recently acquired 9815a and the printer will print a few lines and then stop advancing paper, when the paper stops moving, the sound coming from the mechanism diminishes.

I've attached a photo of the output.

Does the mechanism need to be cleaned and lubricated or is there something else that needs repair?

   

[Geoff Quickfall]

Standby for photos of the diassembly.

There are two lubrication points only, each side of the axle for the daisy (thumb) paper advance wheel.

Next mechanical point are the rubber dampening (bounce pads). One of the pads has a screw adjuster which regulates the height of the bounce. It has probably compressed as well as the other three pads. These regulate the bounce of the impact electromagnet causing the next advance of the printer.

Third is the paper advance roller, mine is hard as a rock but others have discintegrated. There is a fix in the archives that MAY apply, as it is for the 9820:

Printer platen fix

I have to do more research on that.

Geoff

[Craig Bladow]

I've replaced the capstan material on my 9815a tape drive and have inserted a NOS tape with a date stamp from 1990. When I try to mark the tape I get an "End of Tape" error. I've advanced the tape manually, [/align] and don't see any clear spots where the coating may have fallen off.

[cruff]

(04-01-2017 06:36 PM)Craig Bladow Wrote:  When I try to mark the tape I get an "End of Tape" error. I've advanced the tape manually, [/align] and don't see any clear spots where the coating may have fallen off.

Was the tape inserted in the rewound position? You didn't say you removed it from a sealed package, but I assumed so. Did you try this with part of the cover off so that some room light may have reached the tape hole sensor? What happens if you attempt to rewind the tape?

[Craig Bladow]

(04-01-2017 09:09 PM)cruff Wrote:  Was the tape inserted in the rewound position? You didn't say you removed it from a sealed package, but I assumed so. Did you try this with part of the cover off so that some room light may have reached the tape hole sensor? What happens if you attempt to rewind the tape?

The tape came in a sealed package. If you look at the bottom of the tape, the the tape had two holes in it in visible where door on the left that flips open, I manually spooled those on to the left reel which was mostly empty to start with. Yes part of the cover is off. The "Rewind" command results in a "Tape End" error also.

[Craig Bladow]

I replaced the cover and still get "tape end" for Mark and Rewind commands.

[Geoff Quickfall]

Hello Craig!

I am in Beijing and as usual the Chinese will not allow me to PM or send emails. Will be home Monday.

Send me an email using the email button below as I have some ideas.

Some questions:

You mention two holes: the holes in the tape are as follows:

- during a rewind the first "single hole" is encountered telling the 9825 that you are near the beginning of the tape, a second "single hole" is then encountered which stops the tape correctly. If the 9825a continues to rewind it encounters pairs of holes side by side (pairs). These are the EOT holes and result in the EOT message. Physically spool back to the single hole sequence and attempt a read/write at that point.


Did you condition the tape? A full forward followed by a full rewind which adjusts the tension and removes possible wrinkles. Both bad tension (loose to wrinkles) can result in EOT or a travel past the two "single holes".

Geoff

[Craig Bladow]

Email sent.

[Craig Bladow]

Found these in my 9815a today. Removed before working on tape drive!
   

[Geoff Quickfall]

Ha, maybe the previous owner thought that paying homage to the 9825a tape drive subset of the 9825a Gods would improve the EOT outcome!

I received email but cannot send until I leave Beijing so standy.

Geoff

[Craig Bladow]

Troubleshooting the "End of Tape" Error so far:
1. Positioned single hole in tape near the drive wheel. No change. Note that I have not yet heard the motor do anything.
2. Inspected both tapes I've tried for missing coating, which might lead to "End of Tape", both look like they have their coating intact. (Except where the drive band has sat for the last 27 years, but I've manually rolled past that section.
3. Slid an opaque piece of paper behind the cartridge to block the hole sensor light. Still get "End of Tape".
5. Tried to rewind with no cartridge present, received a "Cartridge Out" error, so cartridge sensing appears to be working.
4. Perused HP's 9815a patent and Tony Duell's schematics. Next step will be to trace the Hole sensor signal to see if there is a component or circuit fault.

[Geoff Quickfall]

Ah the puzzle, sorry the fix isn't that easy. beginning to sound like hardware or PCA and not the ubiquitous tape failure.

Geoff

[Craig Bladow]

Here is a photo of the culprit. Good news is that the photo transistor and op-amp appear to be working. I think a 2N3904 or similar will do the trick as I couldn't find the direct replacement for this transistor.
   

[Geoff Quickfall]

Well Craig,

my reconditioned drive is exhibiting the same faults. It functioned flawlessly for Larry during his conversion but now in my 9825a I get a good boot on the 9825a, ie no tape errors. Cartridge out message when empty, but no matter what I do, error 43.

Most aggravating. Wondering about motherboard now.

Geoff

[cruff]

(04-05-2017 07:34 AM)Geoff Quickfall Wrote:  Cartridge out message when empty, but no matter what I do, error 43.

I spent about a month working on the disassembly of the 9825T firmware, which is very similar to the 9825A firmware in its patent listings.

For the tape drive error 43, this is caused by bits 3 (power off), 1 (servo failed), or 0 (unexpected BOT/EOT) of the tape controller's status register (R5) being set. These will be placed into bits 15, 13 or 12 of the ‘errwd’ (RAM base page word) by the ‘stsaf’ routine of the tape driver firmware.

At power up, the this will be caused by the probe of the tape drive during the implicit ‘ldp 0’ executed by the Extended I/O ROM at initialization time, and in addition by any explicit tape commands you may issue.

The first version of my disassember, documentation about what I have found and the disassembler output for the mainframe ROMs and most of the extension ROMs can be found at the
VintHPcom Groups.io mailing list site.

[Craig Bladow]

Replacing the transistor has fixed the permanent "End of Tape" error on my 9815a!

The motor moves the tape briefly for a rewind command but then I get the "Cartridge Out" error.

Removing and reinserting the tape will clear the error until the motor begins moving.

By applying downward pressure on the tape I was able to have it respond to a "Mark" command.

I suspect that my replacement capstan material is too thick and/or the tape is not fully latching.

The capstan material I'm using is 8.9 mm od, 6.3 mm id so the thickness is 1.3 mm.

Any suggestions?

[cruff]

I thought the outer diameter of the roller should be larger than that. I have used the procedure talked about in these Museum and Vintage Computing Forum discussions to rebuild rollers.

Clamp the motor securely, power it with somewhere between 5 to 12 VDC, a variable power supply works nicely. I found about 10 V provided enough torque to not stall. Use a new Xacto knife to trim the "magic tape" to somewhat wider than the width of the hub so that it will stay on the hub during use, but is not too wide to rub on the cartridge housing. If you don't do this, the repaired roller will eventually be spun off the metal roller during use. Use 80 or 100 grit sandpaper or a diamond nail file to reduce the diameter to between 10.7 - 10.9 mm (0.425" nominal) and to clean up the edges you trimmed with the knife.

[Craig Bladow]

(04-07-2017 12:05 PM)cruff Wrote:  I thought the outer diameter of the roller should be larger than that.

I stretched the silicone tubing over the capstan. This failed as it walked of the capstan during operation.

(04-07-2017 12:05 PM)cruff Wrote:  I have used the procedure talked about in these Museum and Vintage Computing Forum discussions to rebuild rollers.

I tried the alternate to "magic tape" which also rolled off the capstan. After re-reading the posts you linked to I realized that I didn't stretch the tape in the direction perpendicular to the roller, and didn't leave extra hanging over the edge so will try again with a 2"x1/4" piece of tape.

[Craig Bladow]

In the continuing adventures....
The second try at a silicone tape fix for the capstan seems to have worked. However I managed to spool two cartridges off. So that means that the end of tape sensor is now failed in the other direction.