Basic Troubleshooting Tips for Old Calculators

by
Alex Knight

Articles


Disclaimer - following are some tips that I've put together based on my experiences with troubleshooting some early-model calculators. Please use common sense when trying to clean, modify, open up, or otherwise repair an old calculator. The author assumes no responsibility for the accuracy or applicability of the following information, use this information at your own risk. Sorry, the lawyer made me say this.

So you've just picked up an early '70s calculator and you anxiously put batteries in it, turn on the power switch, and nothing happens. AAARRRGGGH!!! Don't give up yet on the machine, however, there are some simple things that you can try to possibly bring it back to life, without having to enlist an electronics expert.

Perhaps the most unfriendly element to calculators that have been sitting unused for a while is the oxygen in the atmosphere. Oxygen attacks the metal parts in the machine and can form an insulating oxide layer on contact points that prevent proper operation. In normal operation, this oxide usually gets cleared by use of the machine, but it builds up over time and can be a problem with a machine that hasn't been used for several years. Leaky batteries may also cause corrosion on the battery terminals, therefore, the first place to start your debug process is with the battery terminals on replaceable-battery machines. If there is visible corrosion it should be removed. Even if there is no visible corrosion the battery terminals should be thoroughly cleaned to remove any oxide layer that may have formed. I usually use a mild abrasive such as a pencil eraser or emery board, or gently scrape the contact points with a small knife blade or nail file. This is something that you need to use your own best judgment for, taking care to make sure the battery terminals make good contact with the battery but don't remove more metal than necessary, and don't risk ruining the calculator permanently.

Once this is done, put fresh batteries in the battery compartment - you also want to make sure the batteries themselves are clean. Some machines are finickey about how the batteries are inserted, particularly ones with 'spiral' contact springs - make sure all the batteries are making proper contact with the terminals on both sides. Sometimes the the terminals lose their springiness and may need to be pried outward with a small screwdriver or pliers to make good contact with the batteries. Now try turning the machine on. If the machine doesn't work, re-check the orientation of the batteries, the condition of the terminals, and the proper contact between the batteries, and try again. I've had several machines that were so particular about the batteries that after simply taking the batteries out and putting them back the machine started to work.

If the machine still doesn't show signs of life, the next area to try debugging is the ON/OFF switch. The contact points in the switch are also subject to oxidation. Try moving the switch very slowly from OFF to ON and back, looking for signs that the machine is trying to turn on. If there's still no response try putting slight pressure into the switch while moving it back and forth, to help try to wipe away some of the oxide. With machines that have a shaft on the switch that is long enough to hold with your fingertip, you may also want to try slightly pulling on the switch while moving it back and forth. You may find that the machine starts working with the switch part way between OFF and ON but not when all the way to ON - you may have to live with this but at least the machine works!

Another area of weakness exists on some machines that have a jack for external power. In some of these jacks there is a spring contact that connects the battery when no external power supply is plugged in. This is generally the case only with DC external supplies used with replaceable-battery machines and not with AC transformers used on many rechargeable-battery machines. This jack can fail due to corrosion or due to the contact losing its springiness. Unfortunately the calculator has to be opened up to try to fix a problem with the jack, which you may or may not want to do.

If you decide to try to fix this, open the calculator carefully and examine the part of the jack on the inside of the machine. You want to look carefully at the jack to see if it has three wires going to terminals on the jack and a springy metal contact that is supposed to short two of the terminals together when the external power plug is not inserted.

If you see this arrangement, you want to check that the shorting spring is making good contact. If not, then try to restore the springiness with an appropriate tool. Also, make sure that the contact points are free of oxide. I've done this by gently scraping the contact points with a small knife blade or a nail file.

Another thing to look for while you have the machine open is to see if it uses connectors between two or more parts of the electronics - oxide can develop on the connectors and cause the machine to appear dead or erratic. Sometimes the connectors may be near the battery compartment and corrosive materials from leaky batteries can cause problems with those connectors - I've seen this on several early Rockwell machines. If you're adventurous you may want to pull the connectorized boards apart, gently clean the connector pins, and put the parts back together - but if you do so be sure to take precautions against static electricity, otherwise the electronics could be damaged.

Copyright @1998 Alex Knight

May 14, 1998


Comments added by Nicholas Bodley  (June 25, 1998)

Good thoughts! I'd suggest a few points:

*Before* you put in the batteries, be sure you know which way they are supposed to go in. Putting them in backwards could destroy the innards, if the calculator isn't designed to allow that. (Reversed batteries definitely will destroy a diode in an HP-48, and opening up an HP-48 is not at all casual.

(See the HP-48 FAQ; there are probably some comments there. Go to www.engr.uvic.ca/~aschoorl/faq/ ) (Yes, that is "schoorl"; he's been tremendously helpful to the HP-48 community)).

Try scraping as a last resort, if there's any likelihood that you'll be removing a layer of electroplating intended to protect against corrosion.

Electroplating thickness, for gold, at least, is measured in microinches or micrometers--very thin! For instance, a brass (or steel) spring might be plated with nickel. In the short term, you'll restore operation, but you could risk accelerated corrosion, long-term. However, some contact springs (not so often for batteries, but more likely in external-power connectors) are made of corrosion-resistant alloys.

Consider careful use of a contact lubricant, which can protect surfaces against corrosion, and help maintain good connections. Except for switches with metal contacts, the lubricating properties are incidental. There is a costly substance that has been sold under the tradename Stabilant-22 that enhances conductivity of connections. Go to a good home audio store to get some, and be prepared for a stiff price tag. Don't expect it to necessarily do wonders if there are no other signs of life.

Comments added by Mark Glusker   (June 25, 1998)

Another product that I have used successfully is DeOxit and PreservIt, made by Caig Labs (www.caig.com). This isn't necessarily better (or cheaper?) than Stabilant-22, but it may be more readily available. I have used this on some recalcitrant HP Woodstock keypads with great results, soaking a stiff piece of paper with the liquid and sliding it inbetween the circuit board and the rubber dome mat (after removing several of the heat stake posts at the edge of the keypad to gain partial access). As Nicholas points out, this may not be advisable on rubber dome keypads like the Woodstock HP's, but in this case it helped salvage my considerable $2 investment in an HP-22.

When it comes to collapsible-dome contacts, it's safer not to use sprays that contain solvents. The black button is electrically conductive, and is likely to be made of a carbon-loaded elastomer. (Elastomers are "springy" materials that can be stretched and will regain their shape; rubber is the best known, but lots of elastomers (especially foam) are not rubber.) I'm not really qualified to recommend much; cleanliness is quite important, though. Do *not* scrape the interdigitated contact-finger patterns!

Corrosion protection is very important. I'd say that Stabilant-22 probably doesn't belong here, nor does lubricant.

Spray contact cleaner really ought to do some good, but be careful so that it goes only where you want it to. Better to apply it with a swab than its by using its own pressure; you get better control that way. Try Radio Shack, in the USA.

Especially with calculators that have internal multi-pin connectors, contact cleaner and lubricant should do good. Move them back and forth a few times before the cleaner evaporates.


Opening up handheld electronic devices is a whole topic in itself.

One helpful little secret is that a screw might be hidden underneath a label. Removing the label without defacing it is a real art, in my opinion.

I'd say that you need to find a solvent that will soften the adhesive, but not attack the label nor the plastic of the housing. (It would be nice to have an organic chemist friend!) Then, you'll need to reattach the label with an effective adhesive that holds, yet permits future removal. (Rubik's cube sticker experience can help! Btw, don't rearrange stickers. Disassemble a cube (no tools needed), and reassemble in the solved state.)

Another matter, which is *very important*, is knowing how to replace the screws! But first, a few words word on removing screws. If you have a worn cross-recess screwdriver (Phillips, in the USA (and Europe?); Purasu (?) in Japan), at least try grinding off the point (you could use a sharpening stone, but don't create a groove). Much better yet, invest in a very good set of screwdrivers. Worn cross-recess screwdrivers should be saved for their tool steel...

Any well-designed device will have screws all of the same length. Once you remove your second screw, compare its diameter and length to the first. Keep comparing, and make a sketch for any exceptions. Remember that cats love to see how quickly they can hide those things that look like peculiar bugs.

Comments added by Mark Glusker  (June 25, 1998)

Keeping track of screws is a big problem, especially with mechanical calculators that use many different screw types. Even with LED calcs, as Nicholas points out, there may be a screw that is a different size than the others to keep it from shorting out against something. When I'm taking something complicated apart, I usually put the screws loosely back in the exact same bosses that they came from, just to keep track of them. When reassembling, it's pretty easy to look at the holes in the piece to be reassembled and figure out which screws hold it back in. I find this to be much easier than having a small container with a lot of different screws in it, particularly if I am taking something apart one day and not reassembling it until several days later.

When you remove a screw from a plastic structure, unless the plastic has a molded-in metal insert (or some other metal part), the screw will have special threads. They are "high" and "narrow", spaced quite far apart, and sometimes are made with alternating smaller and larger diameters.

Keep in mind that when the screw was first inserted, these narrow threads deformed the plastic to create mating threads in the plastic. The slight springiness of the plastic ensures a good grip.

Here's the important part: When you replace the screws, try to remember to re-use the threads in the plastic. If you don't, after a few insertions, the plastic in the hole will be shredded, and the screw won't hold. Push the screw gently into the hole, and turn it backwards while pushing gently inward. You should expect to feel a bump when the threads align. Only now should you tighten the screw. Experience and judgment will tell you how tight it should be. (For those with really-big tool budgets, there are torque screwdrivers!)


It seems that the majority of housings have latches to hold them together. These are structures unique to modern plastics that resemble, in their geometry, an ordinary room-door sliding latch bolt that lets you push the door closed, but which also keeps the door closed. Latches are really great for housing designers, because they are "no-cost" features that allow housings to be snapped together really quickly. Properly designed (they almost always are), they hold really well. Compared to a latch bolt, there are no moving parts, they don't stick out as far, the angle is different ("flatter"), they are much "wider", and depend upon the spring quality of the housing material. Assembly creates a temporary harmless deformation, just before the snap.

It also seems that some latches, maybe many, can be disassembled with minimal harm, so that they can be reassembled successfully. However, ** I could be very wrong on this point!** Comments welcome!

Comments added by Mark Glusker  (June 25, 1998)

I think this is generally true, however, it definitely helps to know how the snap was designed to work. Some calculators, particularly low cost models like the Novus Mathematician, have snap details that are easily broken if you take them apart incorrectly. These models were never designed to be taken apart, but can be disassembled with care. Even models designed to be serviced, like LED HP's, usually have a particular procedure that should be followed to open the case. Usually, one side of a snap detail is designed to deflect and the other side remains more or less stationary. If you know which side is meant to deflect, it's much easier to release the snaps by applying gentle inward pressure on that half of the case, localized at the snap location. Usually the plastic case is flexible enough to allow the snaps to be released one at a time (otherwise the case would be rigid enough that it wouldn't need the additional snap in the first place). The problem with randomly prying the halves of a case apart (not that Nicholas was suggesting such a thing, mind you) is that you are applying force in a direction the snap detail is designed to resist. I have seen a lot of damage done this way - many plastics are too soft to resist being gouged by a metal object prying them apart. If you attempt this, I would try to use as wide and straight a blade as possible to distribute the force (e.g. putty knife, not small screwdriver).

Often it's hard to know exactly where the snaps are, but sometimes you can see a small "witness line" near where the snap is located. A separate piece of steel is needed in the injection molding tool to create the undercut for the snap. This separate piece needs to move out of the way before the plastic part can be removed from the mold, and thus where this separate piece meets the main part of the tool, a small line can be seen on the finished plastic part.

Sometimes you can see the witness line in the bottom of the split line between the two halves of the case. It's hard to describe, but under a very bright light it would appear as a faint ridge of plastic running perpendicular to the split line, or perhaps a mismatch in the texture over a distinct area. Once you think you've located the snaps, it's much easier to concentrate your efforts in those areas. I'll have to look at my calculators at home for some specific examples.

I've designed quite a few plastic housings in my career, and the "one-time" snap details like those found in hand held calculators are usually reserved for parts that either never need to come apart, or can be easily replaced if damaged. Remember that for a high volume manufacturer, a plastic part, like a calculator housing, probably costs less than a dollar. Screws and their assembly would cost more money than just slapping two plastic parts together with snaps. It doesn't make life easier for industrial archaeologists like ourselves, but it was a necessary step in the evolution towards today's under-$10 disposable calculator.

Mark Glusker, glusk@sgi.com

I'm not an expert at releasing latches, but fooling around with junked housings can teach a good bit. You can sometimes press inward at a gap between housing parts to help disengage a latch. Sometimes, a latch doesn't seem to be damaged by brute-force separation. A tool like a table knife, with a rounded edge, could be quite useful. Consider getting a Macintosh computer "case cracker" tool. It might be very useful.


|* Nicholas Bodley -  *|* Electronic Technician {*} Autodidact & Polymath
|* Waltham, Mass. - *|* -------------------------------------------------------
|* nbodley@tiac.net -*|* The personal computer industry will have become
|* Amateur musician *|* mature when crashes become unacceptable.



HISTORY ALBUM MUSEUMS COLLECTORS RESOURCES

X-Number World