Repairing HP-41 keyboards

[Michael]

The divide key does not work on my HP-41CX. All other keys seem to work properly. Is there an easy fix for this?

Will the keypad from any other 41C or 41CV work or does it have to be a 41CX for all functions to work properly?

Thanks for your replies,

Michael
mcshack1@attbi.com

[Vieira, Luiz C. (Brazil)]

Hi;

the worst part is actualy cleaning the contacts. Under each key there's a small metal cap that bends everytime you press a key and it "snaps". This way, the inner top of the cap touches the surface of the keyboard. It happens that the cap's border are already touching one of the poles of the key contact, and when it is forced down (key pressed) it touches the other pole and closes the key contact.

The most common problem is related to dust. If dust gets inside the cap, it will be between the cap and the mid contact. I did not find the time to draw it, but I hope my description helps.

The good news is that there is a small hole exactly aligned with the middle of each cap. If you open the calculator and locate the hole under the key with bad contact you'll have a chance to clean it. I use a few brush bristles soaked with some sort of cleaning liquid (isopropyl alchool) and tied together as a bundle. I carefully insert them inside the (mentioned) hole, then I rotate the bristles and remove them. I repeat this operation about three to four times. If it does not recover the key's contacts, maybe you have another sort of problem.

About keypad compatibility: there are two basic types of HP41 if we consider their inner parts. Look at the image below (MoHPC original image)

The left LCD refers to the first series, called fullnut. Their keypad is fully compatible with each other.

The right LCD reffers to the halfnut series, newer ones. They have everything built in one single board: LCD, chipset, keyboard... Like an all-on-board computer mainboard. In this case, there's a need to remove and resolder parts if you want to use a "sane" keyboard with bad chipset: you need to remove the bad chipset and resolder another. The worst of this all is taht the "chipest" for the HP41CV and the HP41CX are not exactly compatible. They are electrically compatible, but there is only one chip that is identical: the processor itself.

Have you checked which HP41 is (are) yours?

I hope this helps for now. If not, let us know.

Success.

Luzi C. Vieira - Brazil

Edited: 18 June 2003, 8:18 p.m.

[Michael]

Thanks for the quick reply. This unit is the half-nut version with rounded corners in the display. I think that you may have right and left reversed in your message? I had heard that breaks usually took out an entire row of keys, this is only the divide key that is affected, so I am hoping that this is good news and that the key needs to be cleaned as you describe. It sounds like the keyboard exchange idea is probably out, but maybe not necessary anyway? Thanks!

[Vieira, Luiz C. (Brazil)]

That's the third time I write the same message (I lost the first two and I cannot even understand what happened). I'll shorten it this time.

As you have a halfnut, try cleaning the key's contact. Look at the image below.

You'd need to remove the chip in the middle and both the LCD and the hybrid chip (top of the image) an place them in the "donated" keyboard. I think it's worth trying to clena the key's contacts first. And you're 100% correct: I wrote the message without looking at the image, so I actualy reversed left and right references; thank you.

Best regards.

Luiz C. Vieira - Brazil

[Randy Sloyer]

In my experience, a single key that does not function has about a 99% chance of being restored through cleaning if:

1. The key has the same amount of travel as all others.
2. The key has the proper "snap" - that is the metal dome that forms the moving part of the switch has not fatigued.

While Luiz is technically correct, the last thing (in my opinion) you want to do is to take keyboard assembly apart as shown in his photos. He is a brave man with several 41's to spare. I would consider it only when every method of cleaning has failed and you know the switch is beyond help without drastic measures. If it is that bad, you'll probably also need a replacement dome from another unit.

Here is (sorry Luiz, I stole your great photo) an image of where you can measure with an ohmmeter to check the operation on the divide key. I purposely point you up to the X key for the one contact so you do not interfere with the divide key center contact. Contact resistance of a good key is less than 1 ohm. If you have really sharp, pointy probes, do not stick them directy into the plated through holes for fear of damage to the dome in the X key. The hole to the left is the common point for the four keys of the bottom row (div,0,.,r/s) and has no dome above.

I just finished cleaning a CX that had 3 dead keys. It was so full of eraser dust that it was a wonder that any of the keys worked at all. You can start by using canned air with a extension tube and direct the blast directly into the key contact hole after doing a good surface dusting. You'll be amazed at the dust bunnies that have grown inside your machine.

Good luck, post again if you have any questions.

Edited: 19 June 2003, 12:22 a.m.

[Vieira, Luiz C. (Brazil)]

Hi, Randy;

I'm flattered for the fact these pictures are of any use for you. Use them as well, for as long as HP does not invite our bosses to close our sites with their images...

You know, I wrote the message three times mentioning how hard it would be exchanging keyboards in a halfnut and that, based on this fact and in the somewhat simple cleaning procedure, cleaning should be performed as many times as needed. Your post is a lot better than mine, thank you.

The halfnut you see in the image (scanned) is the one that recognizes only the first 64 registers. After SIZE 000 it reads [.END. REG 63 ]. If I use external memory modules, after a SIZE 000 the calculator goes nuts and shows the weirdest registers quantities I've ever seen in HP41: [.END. REG 92 ], [.END. REG 101], [.END. REG 05]. Each time you press [PRGM] you see a different amount of available registers. After a few [PRGM] press, the caculator goes completely crazy and orders [MEMORY LOST]. If you let it without memory modules, everything goes fine. I had an original HP41C fullnut case and I built this halfnut there. Now I have a custom HP41C halfnut... too bad it's fake.

That's the reson I played brave and removed the keyboard, I was trying to find the problem at the other side of the mainboard. The best chance to scan a halfnut's guts and generate a few images, right? I'm glad you found this one usefull, I intend them to be.

Best regards and thank you.

Luiz C. Vieira - Brazil

[Mike]

I agree, if you take the keyboard apart, you might (likely will) have loose keys when you get it back together.

I have some photos here that show this disassembly.

41 Keyboard disassembled

[Michael Meyer]

Oh no! The technology failure is spreading!

Maybe Theodore Kazinski wasn't completely wrong. (Bad, sick joke).

To better tech days,

Michael