HP Forum Archive 15

[ Return to Index | Top of Index ]

How to measure HP41 card reader speed ? Message #1 Posted by Bernhard on 2 May 2005, 5:46 p.m. Is it possible to adjust the HP41 card reader speed without even having a HP 41 ? Such as faking a "motor start" signal and then measuring some signal(s) while a card runs through ?

Re: How to measure HP41 card reader speed ? Message #2 Posted by Raymond Del Tondo on 2 May 2005, 8:35 p.m., in response to message #1 by Bernhard If you have access to an HP-41 service module, higher version than version 1A, but lower than 2x, this should help because AFAIK the later 1x have some card reader diagnostics built-in.

How to adjust HP41 card reader without having a HP41 ? Message #3 Posted by Bernhard on 3 May 2005, 4:26 a.m., in response to message #2 by Raymond Del Tondo But I need to do it with the equipment in my electronics lab (have scopes, counters, etc.) alone as I don't have a HP 41. I want to wire that card reader to some non-HP equipment. Edited: 3 May 2005, 6:51 a.m.

Re: How to adjust HP41 card reader without having a HP41 ? Message #4 Posted by Tony Duell on 3 May 2005, 3:42 p.m., in response to message #3 by Bernhard From what I remember, the 8 pin IC on the card reader mainboard that isn't the HP41 ROM is nothing more than a 555 timer. It's wired as a monostable, the preset adjusts the time, and it's something to do with the motor speed control. What triggers it, I haven't figured out yet. It is connected to the 20 pin IC, which is the card reader interface. It may well be that that chip outputs a clock signal. Time for me to get a 'scope on one of my readers, I guess. Incidentally, for those who know the older card readers : The sense amplifier chip (on the little vertical PCB at the back) is the same as the one in the 67/97. But the 'motor control' section is not used to control the motor in the 41 -- IIRC it controls the card sensor LED. So don't start in that area.

Measuring card reader speed Message #5 Posted by Bernhard on 4 May 2005, 9:41 a.m., in response to message #4 by Tony Duell Hi Tony, thanks for offering your help. To be more specific, I have the reader which has the 1826-0322 amplifier chip (like the HP67), but I want to toss away the second board with the motor speed control and the other 20-pin chip (presumably a sort of CRC and ROM) and do motor speed control with my own circuits. The information I need is how long it takes for the card on a *perfect* HP reader to run through. I think this could be measured by connecting a Timer/Counter like the HP5327A to one of the mechanical switches in the reader and take a few time duration measurements while a card runs through. Any HP calc from the HP45 on could be used to calculate mean and standard deviation of the measurements and the data looks good, I'd love to learn the result from you.

Re: Measuring card reader speed Message #6 Posted by Tony Duell on 4 May 2005, 2:41 p.m., in response to message #5 by Bernhard You may have noticed that I mentioned a 'project with a broken 41 card reader' in another thread here. It sounds like we've got much the same idea -- I intend to link the 41 card reader mechanism to a microcontroller, etc... I was going to keep both PCBs. Desolder the ROM (which is dead anyway) and the interface chip from the lower board and solder wires to the right points. I have a schematic (if you want it, either buy the HPCC schematics CD-ROM or ask me nicely..) so I know what pins to connect to. I was planning on keeping the 555 timer circuit and driving it as the HP controller chip does. Fortunately I have a working 41 and card reader so I can make measurements on that... Needless to say anything I discover or do will be given back to the HP community (software will be GPLed as ever...) Anyway my bench is currently full of a dismembered 9845, which is going to take several months to sort out (it's one of the most complex machines -- with one of the most complex PSUs -- that I've ever had the pleasure of working on). So other projects are on hold for a bit.

Re: Measuring card reader speed Message #7 Posted by Bernhard on 4 May 2005, 3:21 p.m., in response to message #6 by Tony Duell Dear Tony - nice to hear you had the same idea, and a "dead" card reader, too. The mine had a gummy wheel and someone tinkered with it and I am quite confident the basic mechanism is still good, but likely the guy has tried to fix it by cranking at the speed adjustment pot, so I need the correct data. There can't be much logic in a CRC chip (1820-1751) so we may succeed to replace it with a small but fast microcontroller. This could be used to repair defunct HP67s or to pull the data from HP mag cards into a PC and from there to cyberspace. Please note my email address and send me a mail when you have the speed measurement from your good card reader.

Re: Measuring card reader speed Message #8 Posted by Raymond Del Tondo on 4 May 2005, 5:11 p.m., in response to message #7 by Bernhard Maybe there's a more simple solution to the card reader speed adjustment issue. If you have a working card reader you may also measure and adjust the electrical resistance of the potentiometer, once given the value of a working unit. I'm not quite sure, but maybe a 'nearby' adjustment could turn the motor speed into the allowed range. At least worth a try... Raymond

Re: Measuring card reader speed Message #9 Posted by Tony Duell on 5 May 2005, 6:12 a.m., in response to message #8 by Raymond Del Tondo To reply to both previous messages : We (Bernhard and I) want to 'bypass' the HP card reader controller IC, which I suspect may output a clock signal or similar to the 555 timer circuit. We're not just trying to repair a card reader. So knowing the resistance of the pot in a working reader won't help. When I was fixing a 65, I connected a logic analyser to the card reader chip read lines and read a card. This showed the signals were pretty darn slow -- a few kHz-ish. Any microcontroller would be fast enough to handle the raw data from the sense amplifier chip.

Cloning the HP card reader controller (CRC) Message #10 Posted by Bernhard on 5 May 2005, 4:08 p.m., in response to message #9 by Tony Duell Tony is right - any fast microcontroller could handle signals so slow. It could also listen to the ACT bus (ISA, SYNC, DATA) and emulate a CRC chip. This is not so trivial to do as there is a lot of traffic going on there. And we would need some true understanding what is going on inside the original CRC. Anyone on this list having some datasheet ? Joke intended ;-) but when I watch how those big corporations handle their proprietary technologies' documentation then it is easy to predict that progress in the microelectronics / computer field will soon come down to a crawl when more and more of those people who created that technology beginning with the 1960s go into retirement. Maybe some ex-HP-employee who worked there in the 1970s lurking on this list and having memories what is in the CRC could help.

[ Return to Index | Top of Index ]

Go back to the main exhibit hall