Test Equipment Anonymous

.

Another acquisition from the swap meet was this HP 85A Professional Computer; while quite limited even compared to most hobbyist PCs of the early 80s, it has a very powerful scientific command-set in its HP-BASIC command interpreter, coupled with HP-IB and other I/O as options. I'd say the best way to describe it would be as "The Ultimate Programmable Calculator".

There are several things that "always go wrong" on these; the printer belts and tapes and tape drive capstan rubber bits all die from age, and the keyboard plastic bits crack from age & embrittlement & morons banging on them.



First thing I found on opening the case was that someone had bought replacement belts, but evidently was a bit gorilla-fisted trying to install them; this screw stanchion for the paper advance stepper was borked.



Not one of the things that "always goes wrong" was one of these paper tape spindles had gotten lost, and this cracked flange that supports the printhead drive cog. Again, I suspect "gorilla-fisted" as the cause; there's also a mounting-screw flange right there with the ear broken off. These bits all mount in place using rubber well-nuts/blind-nuts which had pretty obviously been overtightened for a long time.



After sleeping on it, I contrived to give it a artificial limb by gluing a 3DP bit on it. First step was to score around the broken-off stub so I had a reference to grind flat with my Dremel...



...yielding this which gave me a surface I could use as reference to design a prosthetic for repair.



And here's the printed part being trial-fitted. I need to substitute a m3 x 18mm self-tapper for the original screw; sadly I have nothing with this oddball hex-head, so will use a common Philips button-head screw.



While designing and printing that part, I also designed and printed 2 parts to make a substitute for the missing mate to this paper roll spindle...



All it took was a few licks with the emery board to make the inner spindle turn silky smooth...



...it looks funny because it was designed to be printed flat for best strength from FDM process. Blame Cura for the dickballs.



Perfect fit; the paper now rolls smooth and square. The next problem to solve was that cracked flange; I decided to sleep on it as well.



The next morning I awoke with a plan; after CA-gluing the broken screw-flange back together so I had it for a reference, a couple hours of fettling with a scrap of steel saved from a PC chassis yielded this repair plate. It is designed such that it pulls that crack in the flange closed as it is slid on; I applied CA in the cracks before epoxy-ing it in place.

mnem
Continued in next post...

Here we can see the repair plate epoxied in place...



...and the bottom side along with that spacer now epoxied down.



Final assembly here; I'm dead chuffed with how well that plate fits and works...



...and here you can see how that flange and plate work together with the printhead drive cog.



And finally, the money shot: I typed in the little bit of code that Dave used in his video; it's a variant of the HP Demo program shown in their brochure:

mnem

So, over several days after I got it home, I took a deep dive into CuriousMarc's restoration resources...

At first, I contrived to print up an O-ring hub like the ones he first made in this video, as I had already found those NOS DC2000 tapes above. This involved a rabbit-hole of Frustion360, re-engineering my 3DPrinter (it was already in the queue, but desire to use better plastic for my hub brought it to the forefront), and designing the part as well as a jig to drill/tap screw holes.





Then, just as I was reviewing my progress on a near-finished design, I realized that I actually cared that this approach would make my tape drive only work with DC2000 tapes, as I might later want to sell the 85A on once I finished refurbishing and playing with it.
fuck.

I realized I'd need to do the current "best" method, marc's extended-capstan mod as seen in the video above (85A Segment starts at 09:20); however, as I don't have access to a lathe (yet), I'd need to find another means to get that little bit of metal.




Fortunately, I was at least able to salvage the design time spent on the O-ring hub; that was easily modified to make a part that could be printed in stainless steel by JLCPCB. And due to their minimum part dimensions, I got 3 parts for my $9.15, aligned along a common shaft to cut apart with my Dremel. I did make one iteration with screw-holes, just in case. A bit of epoxy and a custom 3DP clamp and it fit perfectly; after curing overnight I was ready to try again.

I spent a few days playing with my CR-6SE (it supports 0.08mmLH on the 0.2mm nozzle profile, where the CR10Smart only 0.12mm) trying to print a usable tire out of TPU, but after a dozen or so iterations, I still couldn't quite get there. I would need to suck it up and try my hand at the Plasti-Dip process, even though I really didn't like how wibbly-wobbly the drive surface of the finished product looked in marc's video.






The result, after several attempts to get the feel of laying down a film of Plasti-Dip with a putty knife, was actually pretty promising... I learned to accelerate the process quite a bit with a hot air gun, temperature monitored by holding a hand right behind the slowly-spinning hub to ensure not enough heat to damage the polycarbonate timing rotor at the base of the capstan hub.






Then I made the mistake of trying to measure it too soon after it had dried and made dents in the drive surface of the newly-cast tire.






In desperation, I tried rolling it ever-so-gently on a sheet of shelf glass to iron out the dents... and made the wibbly-wobbly almost perfect! After curing overnight, it measured 10.6mm; a wee bit smaller than the nominal 10.7mm, but marc's attempt also measured a wee bit smaller and he states it works very well.





To make it fit in the tape-drive chassis, I had to sand down approx 1mm of the end of the tire for clearance... it's close, but more than enough.





Like marc, I had to file the opening in the drive chassis and PCB to make clearance for the tape-access door on the cartridge; it really was a trivial job with the Dremel. I worked over the nozzle of my shop-vac clamped to the edge of my bench to reduce fiberglas dust contamination.





Here's a closeup of the tire and the drive wheel in the tape cartridge. You can't see the tape very well here as it's all below the drive wheel; the mechanism doesn't work like a conventional tape capstan/pinch-roller. Rather, the capstan turns the drive wheel, which in turn pulls a belt looped around both tape spools inside the cartridge. This arrangement allows for very fast and precise linear tape advancement, while at the same time automatically tensioning both spools.

More details on how this mechanism works can be found here: https://www.hp9845.net/9845/tutorials/s ... index.html

and here: http://www.hp9825.com/html/dc100_tape.html

These articles are a great read; this mechanism really is a ingenious bit of engineering from the golden age of electro-mechanical design. :thumbs_up:





With the motor oiled (a couple drops of PTFE-bearing low-vis synthetic oil from my RC racing daze) and reinstalled, the final task is to install this 1K8 resistor across the original recording current resistor to increase record current, making this drive compatible with the newer DC2000 tape composition. This mod is supposedly backwards compatible with the original tapes as well.





And here's the money shot!

Capstan rotor turning very slowly to obviate any wobble (this motor will turn reliably right down to approx 0.8V or so); I'm dead-chuffed with the results. My next installment will involve making jigs and modding the DC2000 tapes with perforations for the optosensor that detects the ends of the tape as seen here. That will be... interesting.



mnem
it goes round... and round... and round...

This too.

Reserved for Keyboard resurrection.

This one too.

Reserved for general faffing aboot and conclusion.