Hi
I do a bit of calculator collecting and mostly the machines I have in my collection I've repaired one way or the other so normally I've no bigger troubles here almost no matter what's wrong but the HP spice series seems to be the exception.
I got the following two half to/or fully defective ones. A 38E from '79 and a 32E from '82.
Lets start with the 38E. The zero-key does not respond - the keyboard is made of flexible plastic and so is the "PCB" the four chips are not soldered. I've tried cleaning the contact but no luck. I can short the traces near the contact point to the key which results in a zero so no defective ICs. Otherwise in great not very used condition and all other keys work. Do you think it's possible to repair the keyboard with conductive paint? Something tells me it's expensive and perhaps other keys are on their way to eternity too and because of construction they're as good as unreachable unless they're edge keys.
The 32E on the other hand is very well used, display glass not intact but based on past experience I might be able to fix it acceptable. However I cannot get it to work - it seems just dead. I wonder if the inverter could be the case. Is there a common fault in these I can look for? I've tried to measure for shorted transistors but no obvious defective one as far as I can tell from my multimeter on diode setting. Also the 32E is with a normal PCB and constantly messing with the inverter put stress on the soldered ribbon cable.
Another possibility. Which of the two do you consider the most interesting/valuable model? One has a nice PCB with decent "klixon" or what to call them style buttons, the other one an intact glass. The 38E worked flawlessly when I got it but some very occasional light use later the zero is not working. So my heart says the 38E even though I'm mostly a mathematical guy. Also I loved the date arithmetic. Btw. so far I can tell it's functionally and performance wise identical to the original/normal 12C. But It can very well be you say the 32E is more scarce so should repair that first.
My first though is to take the PCB from the 32E. It seems identical to the non solder "PCB" of the older 38E and it even got the room for the 3 small ICs while the 32E only populates two. So unsolder the ICs replace with those from the 38E. Reuse inverter from 38E (also because it's battery end isn't damaged by slight corrosion). Put the PCB in the 38E. As far as I can see closely comparing the outer shells of the 38E and 32E they seems 100% identical so it should be very doable - I have no trouble using a soldering iron. Yeah I "loose" a defective 32E which I btw. was given for free since I love old junk but gets a 38E with higher quality keyboard it should have had from the beginning. Oh and odds are the display is intact of the 32E so if buying one with defective segments it can be repaired - defective ones are cheaper ;) - exactly why most of my calculators have required some sort of repair or at least mostly a total clean w. total disassembly and shine up to look great again.
They must have had a lot of warranty issues since the newer is way better than the older. Without having tried the 32E functioning I believe it's almost on par with Ti SR-50(A) - functionally identical but of course the materials can differ. One can only wonder why many (all?) of their newer calculators use keys basically identical to the older 38E. IMHO that design should have been ditched at the design state. A much better choice would have been conductive rubber like many other things today and from the far east back then - with hard outer keys for quality appearance - it's much more durable and when they make trouble a clean is usually all that's needed.
Edit:
One last question. The big IC - I believe that's the CPU? Are they identical on the 38E and 32E? If so I can just use the one already on the 32E if it's not defective. Leaving one unsoldered for a spare part. That is of course if there's no risk the CPU may damage the ROMs from the 38E. Uf only I could get some life into the 32E...
Edited: 8 Feb 2012, 4:14 p.m.