HP-41CV Keyboard/Behavior Weirdness

[Paul Brogger]

I think I've read the repair & reset articles about the HP-41, but am not sure whether my unit's behavior will be addresed by those techniques.

When first turned on, it seemed to be operating o.k. (With the batteries in the first time, it came up with "MEMORY LOST".) Fine. Then I thought, "I wonder how this thing programs?" I push PRGM, and the first display looked like a programming interface -- a line # in the 230-ish range, with some symbol that looked appropriate, but which I can't remember.

I started to SST, and things got weird. there are two or three three-digit numbers (generally, I think, in the 233-9 range) arranged on the screen, and as I SST, they shift to the right and change values . . . Very bizarre user interface, if that's what was intended.

Anyway, after having done this, I go back to normal mode and the keyboard is all messed up. I push a "1" and get a "9" -- I push the "2" and get a "." -- very quickly the cursor shifts to the exponent digits on the right -- bizarre non-alphanumeric characters appear . . .

I tried the "reset" techniques described in the MoHPC article -- the simultaneous back-arrow/enter/ON thing seemed to work once -- the keyboard started behaving normal-ish. But when I went back to PRGM mode and started SST (some people never learn!) the same result happened, and the keyboard is messed up again. (That, and the back-arrow/enter/ON technique hasn't worked a second time.)

Other things I've tried: -- varying degrees of tightness on the screws; -- cleaning the contacts with 91% isopropyl alcohol; -- shorting the outside battery contacts. (If prayer is recommended, I might consider "getting religion" . . . )

I'm faced with a partial refund situation here -- if the "soldering-the-cpu-to-the-keyboard" approach might help here, I would keep the calculator. If it looks like PRGM memory or ROM is messed up, locking the calculator, then I suppose I'll cut my losses.

Any other suggestions/links?

(BTW, this is Fullnut S/N 2408S41073.)

Thanks . . .

[Diego Diaz]

Hi Paul,

I'm not really one of the experts you can find here, anyway I'll do my best to help.

First, most of the 41's failures comes from two main issues: 1) corrosion, not only in the battery contacts but inside the calc. 2)contacts between the processor board (you said it's a Fullnut) and the keyboard as well as between this and the battery-I/O block.

My advice is to take the 41 apart and carefully clean every one of the above mentioned (some diluted acid to clean corrosion and some IsoPropylAlcohol to remove the acid residue), next is to pay atention to the zebra conector (the one that is between the processor board and the keyboard) some were made of a rubber tube with thin golden wires around, try to "fexibilize" them again pressing'em gently with your fingers (again, clean'em with IPA prior to reinstall).

Needless to say that poles and screws must hold both halves tigth, so if some are loosen or broken it also have to be repaired, you can find here in the MoHPC some good methods to achieve this task.

Hope you get you 41 running soon ;-)

Kind regards from the Canary Islands.

[Paul Brogger]

Folks:

I'm getting ready to connect my HP-41CV CPU and Keyboard via soldered wires as detailed in
Stefan Vorkeotter's article, but I'm confused on (at least!) one point:

He mentions a "middle row" of four contacts which need to be connected in addition to outer rows of 14 and 16 contacts. He also talks about "three flexible CPU board to keyboard interconnects" . . . (He is apparently describing an HP-41CX, if that's important.)

My HP-41CV has two rows of gold "zebra stripe" squishy connectors, held together with what seems to be clear-ish mylar, helped by a black plastic stiffener, and all sitting over & around the two lower screw posts. There's no way the middle row of four contacts (which do seem to be on the CPU board, but not the keyboard) are currently being connected to anything with this arrangement. (My keyboard seems to have two rows of 16 contacts each.)

Is the CX setup an older arrangement, and are these connections no longer maintained in the CV? Should I solder wires to these four even though there appears to be no connection for them currently? (I'm thinking not . . . )

And, does anyone have a torque spec for these all-important screws? (I'm going to buy a teeny little torque wrench . . . )

[Diego Diaz]

For your last post I have supposed that my "cleaning advice" did not fix the weirdness of your 41.

Regarding those infamous four "middle row" contacts, I've (re) read the 41's Service Manual throughoutly and have found nothing about such item.

May Stefan could explain this in detail if you can contact him...

Back to your 41, you said it has two rows 16 contacts each, wich is consistent with what's stated in the Manual, so you'd only need to solder the resulting 32 contacts in total.

Nevertheless, as soldering is a one way trip, I'll keep on trying to recover the golden zebra stripe functionality or even look for some substitute, there are hardware shops where you can find a sort of "all-rubber" contact stripe (as those used to connect LCD's to PCB's) you can cut to the desired length, so you can keep the "modularity" of your 41.

Again my best wishes from the Canaries.

[Paul Brogger]

Diego:

Thanks for your advice and help. Your responses have been helpful, and I appreciate you taking time to offer advice. (And "from the Canaries" sounds so exotic!)

No, everything I tried with the gold squishy thing didn't seem to help much. What gave me hope was that, after trying those things, if I squeezed the bottom end of the assembled calculator just right it would start to behave normally -- keypresses resulted in the right action, and I could even do a CAT 0 and see the functions scroll by. That convinced me that the problem was indeed a flakey CPU/Keyboard connection, and that if I eliminated that variable, everything would come together . . .

Well, 32 pieces of wire-wrap later (have you ever tried to strip 1/8" off the ends of a half-inch piece of wire-wrap wire?) it mis-behaves exactly as before! I press "1", I get "9"; I press "2", I get "."; I press "3", it's as if I pressed "EEX"; I press "Roll Down", I get (of all things) "ASN" . . . Pretty disheartening!

I put the pieces in a basket, and I'll try again this weekend. The next thing to test is the continuity and separation of all my connections, and whether, with the connections soldered, "squeezing" helps. If so, I'll have to assume I'm transferring stress to the CPU board, keyboard, LCD connections or power/module strip -- stress which somehow makes things right. Then, I suppose I'll have to see if I can find out what that something is . . .

I had a chance to get most of my money back (I'd disassembled the calculator before testing it -- WRONG!), so now I'm feeling like the gambler who's lost his bet. Ironically, all this fiddling with the calculator has made me want a working unit even more . . . I'd always thought 41's were kind of ugly-looking -- and I now have a very low opinion of the way things are lashed together inside. But the keyboard feel harkens back to the Spice, Woodstock and Classic models -- it would have been nice to have such a hefty unit to "go to" when doing calculations at the office . . .

Thanks again -- I'll consider undoing the wires and trying some zebra stuff, but of course the gold pads are now covered with solder, and I'm now doubting that the keyboard/CPU connnection was ever the problem in the first place.