Well I just finished working on the APF Mark 1 calculator so I wasn’t looking for another calculator project right away, but I stumbled upon this Burroughs C 3207 calculator with several issues for a price that was too good to pass up. And it’s a vintage 14 digit Nixie tube calculator Built in late 1969.

I have been looking for a 12 or more digit Panaplex display calculator for quite a while, and I thought a Nixie tube calculator would be way out of my budget.
I am quite familiar with several of Burroughs mechanical calculators and their series of computers, but I knew nothing about their electronic calculators. So when this unit became available, I really did not know what to look for, or to look out for.

Some of the issues I noticed before I purchased it were a non-functional 11th digit, hazy scratched display lens, damaged power cord, and overall wear and cleanliness of the unit.
A quick web search before the purchase was of little help, and most of the partial information available, was scattered across many sites and different model numbers.
Hopefully I can pull some of that together in this blog post.

A bit of history on the Burroughs electronic calculators can be found on the DoPECC “Burroughs Calculators — Overview” website. The Burroughs 3200 series of electronic calculators were built by Hayakawa Electric Co. in Japan, later known as Sharp Corporation. The C 3200 series of calculators seem to be based on the Sharp Compet CS-23 calculator with some changes to the case, display and keypad layout.



It was time to start some of the obvious repairs and cleaning. First item on the list was the damaged AC power cord. This unit was configured for 120 volts at 50/60 cycles, and uses 0.235 Amps and 25 watts according to the ID plate. The power cord was very frayed and held together with many wraps of fiberglass packaging tape which also had failed.
I wanted to retain as much of the original calculator as possible, so I ended up cleaning and reusing the original power cord as the only damaged area was where it entered the unit.
After cutting off around 6 inches of cord, the wires were re-soldered to the line filter and new heat-shrink insulation added.

The next item on the repair list were the 50+ year old electrolytic capacitors, mainly located in the power supply.



The larger capacitors were in fair shape with ESR readings a bit above normal, but not bad for this age of capacitors. The smaller capacitors did not fair as well, with high ESR value on the 22 uF at 35 volt unit, and with the 1 uF at 300 volt grey capacitor completely open circuit.

I ended up replacing all the axial electrolytic capacitors in the power supply with new Nichicon and Vishay units of slightly higher voltage ratings, that I had in stock. The 0.1 uF at 450 volt non-polarized capacitor was also replaced.
This power supply provides ‑24 volts, 200 volts, and 0 volt GND to the small card edge connector on the display assembly.
There is an additional 1 uF at 150 volt radial electrolytic capacitor on the display board that I didn’t have a replacement for, that will have to wait until my next Digikey order. It tested in fair condition, but I don’t want to wait too long before replacing it.


Now that the power issues have been addressed it was time for some deep cleaning. If I had to guess, this is only the second time that this case has been opened. The first time being a quick temporary repair of the AC power cable by pulling the cable further inside and re-clamping it. I am also fairly sure no cleaning was performed during that repair. There were packets of dust and debris 3/4″ deep in the front corners of the case.
Overall the main and display circuit boards were fairly clean, with most of the lint and debris around the outer edges, display bezel, keypad, and screened vents.




The calculator is comprised of four major assemblies consisting of the display assembly, main logic board, keypads, power supply and wiring harness.
The display assembly was easy to remove by un-screwing a metal clamp between the two card edge connectors on the back of the main board. After removing the left connector on the main board and smaller power connector on the display board, the assembly is attached with four plastic push-pull locking pins, and lifts straight up.


The main board is held in by the previously removed metal clip and the rubber lined metal power supply bracket at the front of the case. I found that it is much easier to remove the keypad wiring harness connector on the back of the main board just to get it out of the way before sliding the board back and lifting it up to remove.

The next step for disassembly is to remove the power supply and wiring which I originally performed out of sequence. The transformer is attached with four large screws, but to access two of the screws, the fuse block needs to be removed first. The odd thing I noticed about the fuse when removing it to gain access to the block mounting screw was that it had wire leads soldered to the terminals of the fuse block. Something I had never encountered before.
With the transformer now loose the power supply bracket is held on with 5 screws, one of which secures the pass transistor bracket.

All that is left now is to un-screw the keypad bracket from the top case. The bracket holds the individual key modules, decimal point selector switch and power switch.

The tinted plastic display bezel and red plastic thousand indicators are held in by a metal bracket and three screws. The bezel on this unit was heavily scratched and hazy from many years of improper cleaning.

I used some Novus #2 plastic polish, and a hour of hand polishing to semi-restore the bezel to near new condition. There were a few deep scratches that were mostly removed, but still visible upon close inspection. Luckily the pretty glow of the Nixie tubes distract most casual viewers from looking too close.


The dirtiest part of the calculator was the keypad assembly. There was a mix of lint, dirt, food, skin flakes, hair, and oil; my very definition of gross. I would estimate that at least 50% of my time cleaning was spent on that assembly. The keys use a moving magnet to activate a reed switch for each key, except the “F” key which uses a microswitch.
Even though there are no contacts in the keypads, I ended up using DeoxIT D5 for cleaning the mechanical key assemblies, and it did an excellent job.
AC power is routed through two shielded single conductor cables to the top mounted power switch.

After all the cleaning was completed, the quicker process of reassembly commenced. There were no issues with putting the calculator back together, as it is mostly the reverse of disassembly. I did find it easier to reconnect the keypad wire harness connector before fully installing the main board, which is the opposite of disassembly.
It was time for the first power up of the calculator, and after all my usual pre-checks, I used my Watts Up Pro power monitor to verify normal current draw for the calculator.
No major issues with power and the Watts Up meter indicated 25 VA which closely matched the ID plate wattage specification.
There was a small amount of initial concern as the display indicated a random assortment of numbers, some with multiple digits on in the same tube, but this is normal as the registers power up un-initialized and requires the “C” clear key to clear the registers. The same goes for the memory function and requires the “I*” key to be pressed.
In clearing the memory, the memory register is transferred to the display register before actually clearing the memory so if there was un-initialized garbage in the memory, there is now new garbage in the display register again requiring another press of the “C” key.
The quickest initialization sequence is to first press the Memory Clear key, and then the Clear key after power up.
The 11th digit Nixie tube was now working with no issues, so I have no clue as to what I did to fix it, or if there was originally a problem with it at all.

While the Burroughs C 3200 series calculators seem to be based on the H.E.C. / Sharp 22C and 23C designs, the keypad layout and naming that Burroughs specified is significantly different. Here is a rundown of what the keys are and their function.

The CD Clear Display key clears the current display register.
Used for clearing figures mistakenly set.

The K Constant key is used for carrying out calculations by use of a constant. Push to lock the key. Push again to unlock the key.

The RC Recall key, exchanges the contents of the X register (operand being displayed) with those of the W register (second operand).
Useful in cases where the divisor and dividend in a operation need to be interchanged.

The F Floating point key is specific to the Burroughs calculator, it sets a floating point mode of up to 6 decimal places.
Push to lock the key. Push again to unlock the key.

The C Clear key, clears and initializes all the contents except the memory M register.

The Red Minus key, orders subtraction. It starts the calculation after changing the sign of the operator.

The += Plus Equal key, derives sum, product, and quotient.

The Division key orders division. The key lamp turns on when the key is pressed. The key lamp clears upon completion of the operation. If the Constant key is activated the lamp remains on until the key is deactivated.

The Multiplication key orders multiplication. The key lamp turns on when the key is pressed. The key lamp clears upon completion of the operation. If the Constant key is activated the lamp remains on until the key is deactivated.

The I * Clear memory key, recalls the contents of the memory to the display, and then clears the contents in the memory M register only. This also clears the memory indicator lamp.

The <> Memory recall key, summons the stored contents in the memory to the display. There is no change to the stored contents in the memory M register, or memory lamp.

The Memory minus key, subtracts displayed figures from the contents in the memory. (there is no change in the display.) The first memory entry sets the memory lamp.

The Memory plus key, adds displayed figures to the stored contents in the memory. (there is no change in the display.) The first memory entry sets the memory lamp.

The Tabulation slide switch (0 ~ 6), Specifies fixed decimal places.

I did end up making a temporary modification to the Multiply key, because it was not lighting the indicator lamp inside the key. I verified that there was a 22 volt change on the key indicator terminals when the key was first depressed.
The solution is a red LED with a 2K ohm resistor in series, then attached across the indicator terminals. This hangs over the back side of the multiply key housing and projects towards the center of the keycap, lighting it up.
I would like to find out if the indicator bulb is replaceable. It is most likely a 24 volt incandescent bulb. There is no access to the indicator from the back side of the housing, and my initial attempts in removing the keycap without applying too much force were futile.
Update: The keycap indicator has been partially repaired C 3207 keycap indicator repair

Located under the rear AC power cover is a 45 pin interface connector that is connected to the keypad wiring harness and the main circuit board. Along with the keypad, some other signals on the connector are; clock phases, Bit and Digit timing, VDD, GND, IDLE, nPXI, divide / multiply key indicators, along with many other signals.
The H.E.C. / Sharp 22C version of this calculator had an external accessory, the “Memorizer 60” which was an automatic programmer that plugged into this connector and allowed key sequences to be learned and then played back. Most likely it would also work with this Burroughs calculator version.
I really would like to find a copy of the operations manual, and schematics for this model, but I will enjoy using it in the meantime.

If you noticed anything I should have included about the model 3207, let me know in the comments below. Thanks!