|HP-65 Repair/Restoration (Pictures) :)|
Message #1 Posted by Dan Lewis on 10 Jan 2013, 11:17 a.m.
I got an HP-65 for Christmas and I spent the last couple weeks repairing it and getting it ready to go back to work. I'm just posting this now because I haven't had the time to resize and upload all the pictures until now.
I'm posting this with the hopes that you can benefit from it and maybe learn how to repair your own HP classic calculator. I've also posted this elsewhere (on a non-HP calculator site) so some of the descriptions may seem trivial.
Each picture has a description below it.
Here can be seen the world’s first programmable handheld calculator, the HP-65, in all its glory! The funky red background it my anti-static mat. When I first got it, it was quite dusty and even a little dirty. Many calculators today have a “2nd” or “shift” key. This one had so many functions that it needed three shift keys! The “f” key performed the function written in yellow above or below each key. The “f-1” key performed the inverse of the yellow function (if available). For instance, pressing f-1 LOG would take 10 to the power of the number in the display (the inverse of LOG (x) is 10x). The g key performed the function written in blue on the front face of each key. These shift buttons allowed each key to have (gasp) four functions! When I first got the calculator, the “CLX” key didn’t work, and the “g” key could not even be pushed down. This was a result of crystals left by the Ni-Cd battery pack that had been left inside the calculator. Another interesting feature of this calculator is its user-defined or “soft” keys. The A,B,C,D, and E keys could all be redefined to do something else. When the calculator was not running a program from a magnetic card, these keys performed the functions molded in white above them. When the calculator was running a program, the user slipped a magnetic card in a slot above the keys (that would cover the labels already there). The soft keys would then perform the function written on the card. This is the reason that the soft key functions are included twice on the calculator; when running a program, the soft key functions could still be accessed using different keys.
Here’s a close up of the dust on the calculator. The display is also a little hazy, which is annoying (a little rubbing compound did the trick, and it’s shiny as ever now). If you look closely, you can see that the “R” above the D key is not closed in. Because HP knew these labels would have abrasive cards sliding over them, they molded the labels into the calculator instead of printing them on. This way, they could never wear off! A caveat of this molding process is that there could be no enclosed spaces.
Here is what leaving a Ni-Cd battery in for way too long can do. The copper terminals are plated with gold, an expensive but long lasting precaution. The positive terminal’s plating was completely gone and the copper was corroding. Those blue dots are the crystalline material left by the battery (and this is after the compartment was cleaned up!). It’s interesting that the positive terminal was eaten away, but the negative one is just fine.
To get at the screws underneath the label, one must, unfortunately, deface it. It’s alright though, the label can be straightened out with a spoon. In the serial number, 16 is the number of years since 1960, 05 is the week of production, and A means made in America.
At the left hand side can be seen the components of the card reader. The red and silver motor uses a worm gear to spin a wheel which pulls the magnetic card past the card reader (the silver and tan brick with lots of wires coming out of it). That weird blue stuff at the top is more of the crystals left by the battery. I don’t really know how they got in there, or under the key contacts for that matter.
Here’s a closer view of the worm gear that spins the wheel. The motor has a rather ingenious clutch assembly (the silver tube attached to the gear) that also needed to be rebuilt. It turns out that a WD-40 straw is the exact inside and outside diameter to fit the motor shaft, and the outer tube. This way, if the worm gear were ever stopped completely, the motor could still spin (while making a horrible squealing sound), giving you time to turn the calculator off before anything bad happened.
Here’s a close up of the blue substance. It had started to eat away at the gold plating and copper underneath. It sort of tasted like Jell-O (just kidding).
Here can be seen the underside of the keyboard, with more of this crap on it.
This is the remedy for the non-responsive keys: cut a 1/16 by 1/2 inch piece of fine grit sand paper. While holding one end with a pair of needle nose pliers, press down on the key contact lightly with a pencil (eraser side down) and make a few passes. Flip the sand paper over and repeat. The keys work with no problems now. More information on this procedure can be found here.
The keys needed a good cleaning with soap, water, and my toothbrush (I still use it, the 37-year-old dirt just adds flavor).
And HERE, is the INFAMOUS GUM WHEEL. When this calculator was developed, the card reader designer, Bob Taggart, used a machine-cut, polyurethane wheel to pull the cards through. You can read why he chose this here. As the material aged, it turned into nothing more than a gooey mess! This part almost always needs replaced in any HP-65, and this was half the reason I (Santa [my mom]) bought it, for the fun of replacing it!
Here can be seen the card reader contact fingers. These are small switches that tell the calculator a card has been inserted and to start the motor. They also let it know if the card is write protected and if it has passed through the calculator or not. The gold pads on the left side are the faces which the fingers contact.
Here is what puts those finger switches into action. The four nylon balls are pushed out of the way by the card and in turn push the switches to the gold pads on the other side of the keyboard. The little nylon wheel is a “pinch” roller that pushes the wheel into the card reader’s puller wheel. The trapezoidal slots on either side of the balls are spaces for leaf springs, which guide the card into the correct position.
Here you can see those leaf springs and the card reader head. The springs are VERY thin and VERY easy to lose.
This is the “brains” of the calculator. Its multichip CPU had 7, 56 bit registers. The whole thing can be separated from the circuit board-plastic backbone “sandwich” and put aside for safekeeping.
Here you can see the bridge board that contains the card reader circuitry. The crappy soldering job is a result of the wires breaking off after the board was moved around one too many times. It’s tough when you need a 30 watt pen and all you have is a 140 watt GUN.
The bottom of the bridge board has the same connectors as everything else, allowing for easy connection. This is the connection that you should disconnect and not touch again until you are ready for reassembly. There are screw holes that are obscured by it.
I fashioned a new battery terminal out of copper by hammering some 14 gage solid wire flat. Not bad for my first metal forming try. It’s not gold plated, but it works! The bends are critical to its elasticity.
Here’s a view of the battery terminals from the rear end. They are ultrasonically welded to the back bone. I attached mine, and stiffened up the other one, with some good old ethyl cyanoacrylate (superglue).
When I went out to get a spark plug for our snowblower, I asked the man if he had any fuel tubing in stock. We sized it up, and I came home with a 1” piece, for free! Needless to say, this solution was too good to be true and didn’t work because the tubing was not concentric. This resulted in the cards being pulled through at a non-uniform rate, which the reader did not like.
These are the business ends of the “ON/OFF” and “W/PGRM/RUN” switches. A common problem with these switches is that their metal parts (shown here) wear out and the switch no longer “clicks” into place. You may not think it annoying, but have you ever flipped a switch that didn’t snap? It’s actually quite frustrating. The solution is to bend the metal upward slightly so that the switch (and its associated “click” tang) gets pushed up into the front face of the calculator harder, thus giving a satisfying click. The one on the left is unaltered; the one on the right has been bent slightly.
After the fuel tubing didn’t work, I had an epiphany. What if I could just make a new wheel out of silicone? It’s simple, just drill a ¼ inch hole in some wood, fill it with silicone, and shove the gear in! Well it worked, but the ¼ inch hole was a bit too big and the wheel was too large to fit a card through. I tried again with a smaller bit, but it was still too big. I tried to shave some off but with no luck. So, I gave up, and fixed it the “real” way and ordered the proper sized o-rings from ebay.
Here’s a picture of the o-rings installed and ready for action. The card reader now works like new!
Here’s a picture of the finished calculator, ready to go back to work. After being stored in a closet for 20 years, I’d say it’s happy to be working again (at least I am that it is). All in all, it didn’t need that much work, and what did need done is nothing I couldn’t handle. I learned a lot while doing this project and I enjoyed every minute of it.