Opening a Spice machine

Yes, I know I'm LCD kinda guy, but in a moment of weakness I agreed to help out with a spice machine. I have one missing it's battery contacts that is otherwise dead, so it's time to find out what's inside. I know this is not going to be pretty (shudder).

I've looked through the archives and cannot find anything on separating the lower case halves. I remember reading about a trick with dental floss but can't find the posting. While the new compressed format archives saves bandwidth, I haven't figured out how to search them easily.

Can anyone point me to the Rosetta stone of spice opening? Thanks all.

PS: Heh Norm, this might be your only chance to say "I told you so", or just "LED's rule" :^)


Hello, Randy;

have a look at the images below. They are not in scale but will help you understand how do Spices are interlocked:

There are some techniques to open the case, and one of them involves using dental floss. As you may see, what you have to do is slightly and carefuly fasten both halves in the bottom while you pull the back half backwards, so it slides over the small plastic "tab" (is it correct?). The dental floss creates a "path" so the back half slides safely. I use a small plastic cut from new HP-blister case. There are plastic surfaces not thick enough to injure the plastic "tab".

Hope this helps.

Luiz (Brazil)


Randy --

The "dental floss technique" for opening Spice units

On of these days, I'll write an article, complete with my pictures and Luiz' illustrations (he's already assented.)

-- Karl


Like Randy, I remembered something vague concerning dental floss. When I got curious about my working HP-34c, I just wasn't curious enough to overcome my squeamish uncertainty about simply forcing things apart.

The diagram and instructions will enable me to take a look.

Thanks, Luiz and Karl!


It can be done without deltal floss.
Remove the screws under the battery cover. Then take one half with the left hand, the other one with the right one and pull firmly. Don´t care I have opened a lot of them and I have never broken anything.

Good luck


Yes, just pulling the case halfs apart can work and in some really stubborn cases is necessary. But there is a MUCH safer way to try first. Basically you remove the two top screws and lift the top half of the case up enough to clear the battery contacts. Then place your thumbs in the battery compartment and fingers along the logo strip on the top half of the calculator. Then you squeeze the machine so that the bottom half of the case is slid forward and the internal latch is released.


Hey, Karl;

as North-Americans say: anytime, anywhere. You and anyone else are granted to use the drawings. They are so specific that I cannot see another use for them.

I'll correct the drawings (just few "adjustments") but I'll keep the file type and size: they are somewhat small (4K and 34K) and I think this is an advantage.

I did not keep correct scale, because the interlock is, in fact, closer to the "edge", a bit more to the right in my drawings. I thought it would be easier to understand if I exagerated this distance a bit.

Thank you for your (and others, too) support.

Luiz (Brazil

Edited: 7 July 2003, 11:28 p.m.


Well, it's open and it's not pretty. Thanks all. David's push/pull method worked fine. Now on to the ugliness of what embodies a Spice.

It is a 32E that is missing it's battery and contacts. No biggy, that's why there are bench power supplies. 3 volts applied shows no sign of life. About 0.5 ma off, 1.0 ma on. Definitely no LED activity here. It is a compression/solderless unit. It has been apart and ravaged before, as there are three ugly solid wires soldered from the battery connector onto the circuit board, bypassing or at least backing up the flex circuit. Everything there is in order, the wires go to the proper locations. The power switch board pads were trashed with extraneous solder balls for no apparent reason. That cleaned up reasonably well, at least now the switch is smooth.

The first thought since it has been apart is: Are the chips in the proper locations? The two 8 pins are at the top and bottom left positions with nothing in the middle location. The top chip is 1MA1-0209, bottom, 1MA304. The 40 pin is AMI 1820-2162. Correct so far?

The next question is: Should I solder them into place? The bottom carrier has some indentations where the chips are. Could the contact be that bad?

I've now got a pile of parts that once was a calculator. Putting it back together mechanically is easy. Doing that so the pieces will work as the engineers at HP intended back whenever it was they designed this monster is very questionable.

Anybody out there with the lumps, bumps, bruises and warts to help me improve my chances at getting this thing to work? It just makes me love the 41 all the more...


Hey, Randy;

welcome to the Spice's world. (no music, please...)

I do not own an HP32E, but I see that the 8-pin IC's codes "look" strange; well, as David Smith called our attention many times, seeing these codes leads me to think that it's possible to find RAM-like codes in these 8-pin IC's and no ROM-like code. I placed all data I found (there is some missing data, but I'll find it and post it, too) in one single table (credits given)

ID# "2" CT      CPU ID #              1         2       3       OBS
31E N SL 1820-2105-B (7922YY) ---- ---- 1MA301
31E Y S 31E 2438 00031 ---- 1MA15002 (1)
33E Y SL 1820-2105 (7822EX) ---- 1MA10204 1MA10003
34C N S 34C/B-2102 00034-69110 1LB902 1MA10014 1MA115 (1)
37E Y SL 1820-2122 (7819MO) 00031 no_label 1MA10008 (2)
38C N S 38C/B-2102 00038-69110 1LB903 1MA30006 ---- (1)
38C N SL 1820-2162 (7944DY) 1LB903 1MA305 ----
38E Y SL 1820-2122-A (7850MC) 1MA107 1MA10013 1MA10005
38E Y SL 1820-2122-A (7916CS) 1MA105 1MA113 1MA30003
(Ellis posted these ones)
34C Y SL 1820-2162 A (7906) 1LB5002 1MA1 14 1MA10015 S#1934A
34C N S 1820-2162 (8047) 1LB902 1MA1 14 1MA115 S#2051S
(Randy posted this)
32E ? SL 1820-2162 1MA10209 ---- 1MA304

I see we're growing. Soon we'll have a considerable database that will help others to check their Spice's safely.

About the non-working condition: I had an HP31E that could not charge batteries. After removing the flex circuit completely and placing wires, it worked perfectly fine. Have you inspected or measured (continuity) the flex connections? If you have broken trail(s), it's more likely to happen what you described.

I have no HP32E, but the HP31E seems to have the power board. If you need, I can take some measurements, but let me tell you Ellis' post about the HP34C and its power supply is a must-read.


Luiz (Brazil)

Edited: 9 July 2003, 12:45 a.m.


Randy you're doomed. The ESD alone is going to blow those chips away. I have written some views about dealing with these units ONLY with anti-ESD precautions. You'll almost certainly catch "2 disease" from the ESD, in my opinion.

I have an article posted about 34C, if you want to do research, you might want to read that over. You see, I have tried soldering chips down..... and the observations about whether to try it are in the article.

To make SPICE units better, one thing I would definitely recommend is gentle trimming of certain plastic clips with an X-Acto knife, so that they release easier.

This includes the 6 clips that hold the calculator module down. And, it includes the various surfaces on that frightful full-width latch at the bottom edge of the calculator.

The X-acto knife can be used to advantage, for future disassemblies ...... of course, you have to know how much is enough. If you trim away too much then you wreck it.

There are actually 2 full-width locations, both on the bottom-shell of the calculator, that can be trimmed to advantage for future disassemblies of that lower latch.


Hi, Norm;

I remember I read in one of the first posts about Spices (yours?) about solder the IC's in a solderless unit. I want to know if you used conventional soldering iron and if the flex material (sort of plastic) melted when soldering the chips. Did you use any particular technique? There is an article about using soldering flux and microwave oven to produce regular and safe solder. Have you ever tried this?

Just curiosity.

Best regards.

Luiz (Brazil)


I don't have a solderless unit (yet), or I might be tempted to try this . . .

I've fixed a couple of TI's recently with conductive copper fluid from a Permatex "rear window defroster repair kit". The tiny bottle of liquid seems to be alcohol-based, is loaded with very fine copper filings, and dries almost instantly.

I suspect that, if the a chip could be held in position and the circuit not flexed or moved, a tiny drop of this fluid could be introduced separately into each lead/pad joint and surface tension/capillary action would spread it around. After drying the stuff should help insure more reliable electrical bonds.

Buuuut . . . I haven't tried it in this application, so it (or some other conductive ink/glue/paint) might be worth considering, but I've no experience to back up the notion.


Hey, guys;

based on Paul's idea, why not to "paint" the flex tags and the IC's terminals with conductive ink? That would be safe and would allow future replacements and maintenance, what I take as the best advantage for solderless units.

Best regards.

Luiz (Brazil)



Yes, the article I posted described that there is a tendency to see melting of the plastic underlay. In fact, it was still quite useable, but nevertheless there was some melting, which was worrisome.

Yes, it was a precision temperature-controlled iron,
digital readout, using standard solder, set as low as is fair and reasonable. One could get into specialty low-melt point Indium solders, BUT, those stick lousy so you are back to where you started. You're triple-doomed if you don't have a precision controlled iron.

THE CONCLUSION is that the Luiz "clean it up and oil it and put it back together" is right on-target.

No, I dont think U want to use conductive epoxies. If they get a hairline crack, you are in trouble once again. You have no assurance that they will bond mechanically.

Back to the "keep it solderless" then what oil? A little unsure there, some kind of "tv tuner cleaner", etc. Just enough oily stuff to keep the oxygen from coming in and creating fresh corrosion.

I say keep it like the manufacturer intended. This is true for many situations, including car repairs. The more U modify, the more trouble U got. If you keep it solderless, U have the option of swapping chips later if there is a problem. Solder it, and its all over. Just clean it up enough so that it works OK. The article I posted says that the little 8-pin chips are seen to form a bluish oxide, which may contribute to not making a connection. Probably the oil would reduce the tendency to oxidize.

I use a "Leica Zoom 2000" microscope, a very low power (x7 to x30), yet exceptionally convenient inspection microscope. Just set it to 7 power. Flip the chip upside down and put it on the stage, then U start working on those chips to make the blue oxide go away, and see fresh metal.

What do you use? I dunno, maybe the finest wet/dry sandpaper U can buy at Home Repot.

In all of this, you are at grave risk of wrecking the chips w/o the MAXIMUM possible ESD protection. I would suggest some of those "chip clamps" from the early 1980's, that had a clip every 0.1" and you clamped it onto the chip. Short all the pins together on the clip. Now you've shorted out every pin on the chip, and can work on its oxidation problems quite freely.

Remember that if all pins are shorted together, ESD damage is impossible.

That STILL leaves you with 100 other issues about dealing with ESD, at other moments in time. But if you want to work on each pin of each IC, try those clamps, shorted out.

Put it back together at the end, avoid ESD trouble to avoid "2 disease" and it should be fine, even as a solderless unit. That's what I learned by trying to solder one.


Greetings, when time permits, I have continued to try to solve technicalities regarding spice calculators.

The key cushion pad between the keys and the domes, I am still trying to locate a good substitute. Thanks for those who supplied samples. I am now dealing with a die-cutting shop that is making reasonable effort to match the material.

Firstly, its NOT "Nitrile 60 rubber" because that stuff stinks too much. U dont want your calculator to stink like a tire shop do you ?

And, its NOT the pool liner at Home Depot. That's made of PVC and has no dry lubricants.

The dry lubricants appear to be important. Under the microscope, the white powder is STRONGLY embedded into the black silicone (I'll say silicone until I learn what it really is). Whereas rubber products from a gasket shop has a superficial talcum powder that rubs away easily.

I think the dry lubricant is important . We are still trying to match the material. If anybody KNOWS a supplier for the material, now is the time to talk.

The genuine HP item is about 0.011" thick . I think it
could go thicker... maybe 0.015" or even 0.020" thick.

Where I tend to be going, is for a few hundred bucks, they will make a tool and DIE CUT the material and supply me with a boxful. Folks, dont forget the industries where this stuff really comes from (rubber feet folks, paying attention?) you get the material and have it die cut.
They want a few hundred bucks to prepare a tool (a bunch of razor blades) then they stamp away... So I'd say that
you may spend more time foolin' around with scissors than
just getting the real thing (and it wont look homemade either). Then sell the extras to people who need the same item.

After all, one good HP calculator costs a few hundred bucks anyway, why not get the genuinely tooled parts you really need, at least for things as simple as die-cut rubber.

- Norm


We've got a lot of issues with the spice units,
including "2 disease", what batteries to stuff in, how to service a solderless unit, how to open them, how to lube the slide switches, and maybe how to trim the plastic clips so they dismantle a little easier.

HOW ABOUT THE low battery indicator?? When I use "AA" Alkaline, whats noticeable is that the indicator seems to light up long before the battery is near the end-of-the-road.

Anybody else notice this? It gets brighter as you illuminate more segments of the display.

I put the batteries on the scope, I saw there is 0.1 volt of noise on the batteries, pretty consistently. It's the chopped loads from the display multiplexer. That right there is enough to perhaps make the low battery indicator less reliable (adds 0.1V of slop).

Where I would go here........ I think the low battery indicator may have a threshold that is TOO HIGH A VOLTAGE and furthermore may be too RAPIDLY SENSITIVE. It should be filtered to turn on only to the AVERAGE BATTERY VOLTAGE, and maybe have a lower threshold. Its enough screwed up that I dont feel happy with it.

ANYBODY GOT A SCHEMATIC? Not a marginal one, but a thorough one... say for an HP-34C. With an exact schematic, I could see what's the deal with the low-batt indicator and clean up the circuit a little (maybe add a cap, or change a resistor).

FACT: If you add 1000 uF across the battery, the low battery light is less likely to illuminate. This means the current spikes are tending to draw down the battery on a spiked microsecond basis, and triggering the low-batt indicator.

- Norm


Well, the low battery indicator was designed to warn you about when a rechargable would fail, not an alkaline. Rechargables of that era would hit the lower voltage level and then take a steep nose dive.

Many people use rechargables for their Hp48's and complain about this problem. What you are seeing is the low charge lite designed for a rechargable, when your alkaline still has as much electrical capacity of the older rechargables with a full charge. Your redesign would probably work for your alkalines, but ruin its use for rechargables if you ever wanted to return to them.


I have a couple of machines with a black rubber sheet that does NOT have any white powder. They are slightly sticky and several keycaps lift with up with it. But, it works just fine... mayby even better than the powdered ones since the keys tend to not float around as much. Probably no need for a special powdered liner.

I also found a third type of HP liner material. It is a lot like the yellow foam rubber material but is green and has a slightly higher density.



Some time ago I saw a 37E that had been serviced and didn´t work. The CPU was rotated 180º. After placing in it´s position it worked. May be is worth you try it. I don´t think you will damage the CPU because it didn´t happen with that calculator.

Good luck



I'm not sure, but I think maybe "them are fightin' words"! (Kind of like "bring them (or was it "it"?) on!")


I know it isn't graphical, but the netlist I posted includes every net that contains more than one device pin. All unlisted device pins are no-connects.


Hi Jon,

Already looked at that. All pin 1's are toward the top of the unit. I was curious about the empty hole in the middle of the 8 pin devices - I didn't know if the previous fixer forgot a part. Wouldn't be the first time. I later noted that it probably never had a part in that position as the pads where free of indentations unlike the positions that held parts. I'm sorta stuck. Everything looks fine, no oxidation or creepy crud anywhere. Not sure what to try next.


Hi, Ellis;

you know what? I remember seeing these words in a soft situatuation, but now you made me remember that in other situations it was exactly like this: a call for fighting.

Karl, I hope you did not understand the wrong way... I was trying to let you know that you can use the drawings anytime you need, anywhere you want to.

Thank you, Ellis.

Luiz (Brazil)


On a lot of Spice machines the low battery light will flicker a bit when modern batteries are installed and fully charged. I think the duty cycle of the PWM is shortened to the point where it fools the circuit into thinking the battery is low.


yeah but with a schematic, maybe I could add a filter cap or change a resistor and then it would work fine. That's why I'm interested to look into this further, as time permits.

Problems like this should be fairly trivial to correct..... its really kind of embarassing to HP that they did not accomodate disposable Alkaline batteries in the spice units from day #1 ...... since they were available at the time ..... and since they last so much longer than a charged-up NiCad.


Based upon the number of machines out there that would need a new sheet, it will probably be much more cost effective to have them laser cut. Lower set up charges and the ability to do smaller production batches.


Even though there is no corrosion evident, have you cleaned the mating surfaces? Some people say to clean with alcohol, and this might make sense to remove invisible residue from plastic outgassing, etc. but I don't know if it will remove invisible metal oxides that could be causing high resistances. Personally, I like to use the "Pink Pearl" pencil eraser that has been recommended for a long time to clean the gold plated contact fingers of plug-in option boards. When I reworked my unsoldered 34C recently, I used an eraser to clean all the pads of the flexible PCB, the pads on the power supply flex circuit where it mates with the main flex PCB, the mating surfaces of the LED display module pins and the bottoms of the IC pins that contact the main flexible PCB. Then I used a rubber bulb (I'm too cheap to buy canned air!) to blow air to sweep away the eraser crumbs. Also I used the eraser to clean the slide-switch contact surfaces on the flexible PCB and the the little moving metal contacts, too, before lubricating them with Radio Shack Lube Gel. The switch contacts looked like they had never been lubricated - same on my soldered 34C - and on the unsoldered one there was some serious battery corrosion on the slide switches, which I cleaned off first. Basically, I just used a sharp probe to scrape the corrosion down to something shiny, then used the eraser to polish it up.


Occasionally Fry's has two can packs of "air" for $0.99! Usually limit one pack. Pay cash, take to car, repeat until x<=0.


Forgot to mention while at Fry's pick up a Caig "Gold Pen". Like a magic marker filled with anti-corrosion stuff. Kinda pricey, but seems to work well on Spice connections, classic keyboards, etc. They also make something called Deoxit pens. Probably filled with the same stuff, but it is red and costs a bit less.

Possibly Related Threads...
Thread Author Replies Views Last Post
  HP Prime : Galton's machine Mic 0 807 11-15-2013, 04:23 AM
Last Post: Mic
  opening hp 15c for repair FORTIN Pascal 4 1,497 10-12-2013, 10:35 AM
Last Post: FORTIN Pascal
  CLASSIC/SPICE KEY LEGENDS Mike T. 2 1,078 08-21-2013, 04:53 PM
Last Post: Mike T.
  Convergence Dream Machine anetzer 11 2,432 04-12-2013, 07:16 AM
Last Post: David Hayden
  HP Spice Legacy in the former CCCP (MK-61) Mike Morrow 17 3,464 04-07-2013, 08:25 PM
Last Post: db (martinez, ca.)
  O.T. - A pristine machine Palmer O. Hanson, Jr. 4 1,252 08-08-2012, 12:16 AM
Last Post: Dave Shaffer (Arizona)
  The Turing Machine Comes True Gilles Carpentier 1 926 06-26-2012, 09:26 AM
Last Post: Rory Molinari (USA)
  (OT) Lego Turing machine BruceH 9 2,260 06-22-2012, 08:14 AM
Last Post: Bill (Smithville, NJ)
  Working around broken Spice battery connections Les Wright 7 1,994 06-06-2012, 06:50 PM
Last Post: Mike T.
  Woodstock/Spice Keyboard Cleaning Revisited Matt Agajanian 0 758 05-12-2012, 03:43 PM
Last Post: Matt Agajanian

Forum Jump: