Canon Canola L100A Calculator Refurbishment

A recent addi­tion to the desk­top cal­cu­la­tor col­lec­tion is the Canon Canola L100A desk­top 10 dig­it gas dis­charge dis­play cal­cu­la­tor. This unit was pur­chased as a work­ing cal­cu­la­tor, and was in good shape except the heavy yel­low­ing of the beige upper case..

The unit arrived with it’s pow­er cord wrapped sep­a­rate­ly, and was fair­ly well packed. Upon open­ing the pack­age, it was obvi­ous that the cal­cu­la­tor came from a cig­a­rette smok­ing envi­ron­ment. So the first order of busi­ness was to dis­as­sem­ble all the major com­po­nents and thor­ough­ly clean all the case­work, keys and pow­er cord with my favorite all-pur­pose cit­rus clean­er. The cir­cuit boards were cleaned with 91% Isopropyl alco­hol. There was quite a bit of flux residue on the boards from ini­tial hand sol­der­ing of wires, and quite a few areas where the main board had been reworked to remove sol­der bridg­ing and to touch up con­nec­tions from the orig­i­nal wave soldering.

Canon Canola L100A main cir­cuit board after ini­tial cleanup

The L100A is com­prised of three main mod­ules con­sist­ing of the main board and dis­play, key­pad assem­bly and con­nec­tor, and the AC pow­er con­nec­tor, switch, and trans­former assembly.

Canon Canola L100AAC pow­er assembly

My L100A was fac­to­ry set for 115 volt 60 Hz AC input. The three wire input recep­ta­cle routes the AC line leg to the pow­er switch and then to cir­cuit board con­nec­tion J1, which then con­nects to the fuse with the oth­er end of the fuse con­nect­ed to trans­former wires K6 and K7. The neu­tral leg of the con­nec­tor is tied direct­ly to cir­cuit board con­nec­tor J2 which then con­nects to K2 and K3 trans­former wires. Transformer wires K1 and K8 are sol­dered to the cir­cuit board, but are on iso­lat­ed pads. Pads K4 and K5 are linked togeth­er but have no wires attached.

The trans­former pri­ma­ry wind­ings in this con­fig­u­ra­tion mea­sure 126 Ohms. The two yel­low leads on the trans­former sec­ondary are the high volt­age sup­ply for the dis­play and have a resis­tance read­ing of 319.4 Ohms. The two grey trans­former sec­ondary leads are the low volt­age sup­ply and mea­sure 11.46 Ohms. Both sec­ondary’s attach to the main board thru a 4 pin con­nec­tor.
Measured cur­rent drawn at 120 VAC input is 0.062 Amps with all dig­its dis­play­ing 8’s. My mea­sure­ments show a max­i­mum of 5.8 watts and an idle wattage of 4.9 watts after the dis­play was cleared.

Canola L100A — key­pad and con­nec­tor assem­bly with new felt

The key­pad is fair­ly sim­ple with two teal col­ored clear keys, a 0–9 with dec­i­mal point sec­tion, 4 basic math func­tion keys, and an equals key. The key­pad is attached to the main board with a 20 con­tact edge card con­nec­tor with 13 wires. I end­ed up installing new cable lac­ing, as the old stuff was very dirty and start­ing to unrav­el. The key­pad had some foam strips under the keys that had dete­ri­o­rat­ed, so it was removed and replaced with some adhe­sive backed felt strips

Canola L100A — key­pad cir­cuit board and reed switches

The keys use a mag­net and reed switch that clos­es when the key is depressed. The cir­cuit board uses a 2 X 11 matrix for con­nect­ing the switch­es to the main board. There are only 18 switch­es, so that allows for sev­er­al spares. Transistor TR2 is con­trolled by pin 25 of the TMS1824 IC and alter­nates selec­tion of the two columns in the key­pad matrix (b1 & b2).

Canola L100AAC to DC fil­ter sec­tion with new capacitors

The AC pow­er assem­bly con­nects to the main board via a con­nec­tor with 4 pop­u­lat­ed pins. The high volt­age sec­ondary is con­nect­ed to two diodes D2 and then is fil­tered by the C2 ver­ti­cal­ly mount­ed axi­al capac­i­tor pair rat­ed at 4.7 uF and 200 volts each. The low volt­age sec­ondary con­nects to 4 D1 diodes in a full wave bridge con­fig­u­ra­tion, and then is fil­tered by C1 a 470 uF at 25 volt capac­i­tor. The HV fil­tered out­put is 297 volts and the LV fil­tered out­put is 20.3 volts DC.

You may have noticed the dif­fer­ent label­ing of the com­po­nents by type and not by indi­vid­ual com­po­nent. So all 4.7 uF @ 200 V axi­al capac­i­tors are labeled C2, both HV diodes are labeled D2 because they are the same, there are 11 each C8 capac­i­tors. Only dif­fer­ent val­ues are giv­en dif­fer­ent des­ig­na­tions. This threw me off a bit when I was putting togeth­er my list of replace­ment capac­i­tors, with two capac­i­tors labeled C2 and three of them labeled C3.
Both 200 volt C2 capac­i­tors were show­ing signs of leak­ing, so I decid­ed to replace all of the 50+ year old elec­trolyt­ic capac­i­tors. Most of the capac­i­tors were Shoei brand­ed, with one of the C6 capac­i­tors being an orig­i­nal Nichicon unit.

Canola L100A — removed elec­trolyt­ic capacitors

Original capac­i­tor and replace­ment list:

  • C1 x 1 — 470uF @ 25V radi­al 7.50mm lead spac­ing replaced with 
    • 470uF @25V radi­al Nichicon UPX1E471MHD1TN
  • C2 x 2 — 4.7uF @ 200V axi­al replaced with 
    • 4.7uF @250V axi­al Vishay 500D475M250DC2
  • C3 x 1 — 33uF @ 6.3V axi­al replaced with 
    • 33uF @ 10V axi­al Nichicon TVX1A330MAD
  • C5 x 2 — 1uF @ 100V axi­al replaced with 
    • 1uF @ 150V axi­al Nichicon TE1500-E3
  • C6 x 3 — 4.7uF @ 25V axi­al replaced with 
    • 5uF @ 50V axi­al Nichicon TE1303-E3
Canola L100A — rear of main board view under display

The small­er axi­al capac­i­tors are locat­ed under the dis­play assem­bly, but are easy to replace with their polar­i­ty clear­ly labeled on the board.

Canon Canola L100AJRC Septanix dis­play module

The dis­play mod­ule was made by the (Japanese Radio Corpo­ra­tion) JRC and is a Septanix 237A1 . The dis­play is sim­i­lar to a Burroughs Panaplex dis­play, but is made up of indi­vid­ual dig­it mod­ules in what looks to be a ceram­ic case with a glass top cement­ed to it. There are 12 dig­its on the dis­play, but only 11 are used by the cal­cu­la­tor. The left most dig­it is used for the sign and error indi­ca­tion. The unused dig­it is cov­ered to pre­vent unin­ten­tion­al infor­ma­tion from being dis­played.
The dis­play bezel had what used to be an open cell foam gas­ket that has since degrad­ed into a gooey mess. I was able to clean the dis­play with 99.9% Isopropyl alco­hol which eas­i­ly removed the residue, but it was still quite a mess with black gunk every­where. I have some very thin felt that was used as a spac­er between the glass and the bezel to replace the foam.

Canola L100A — main board chipset and key­board connector

The L100A uses two (Large Scale Inte­gra­tion) LSI (Inte­grat­ed Circuits) IC’s man­u­fac­tured by Texas Instruments specif­i­cal­ly for Canon Inc. The TMC1824 and TMC1825 are both (Dual-Inline Plas­tic) DIP cas­es with 40 pins each, using a 10µ PMOS (P-chan­nel Metal–Oxide–Semi­con­duc­tor) process. This chip-set was Canon’s fourth series of cal­cu­la­tor chips man­u­fac­tured by Texas Instruments.
The cal­cu­la­tor also uses inte­grat­ed dis­play dri­vers includ­ing two KH6248 IC’s for the dis­play anodes one KH6249 IC for dri­ving the dis­play seg­ment cath­odes.
The KH5305 tim­ing mod­ule is used for clock gen­er­a­tion and its out­put is dis­trib­uted to both 40 pin IC’s.
Component date codes and the qual­i­ty con­trol stamped label, set the date of man­u­fac­ture at June of 1972 for this unit with ser­i­al num­ber 217253. The Nichicon capac­i­tor had the old­est date code from the sec­ond week of 1972 mak­ing it 51 years old as of the writ­ing of this blog post.

Canon Canola L100A oper­at­ing with bezel frame removed

The bot­tom case shell is fit­ted with a shield that has a Gold Iridite Chromate con­ver­sion coat­ing applied to it. The shield is ground­ed to the AC pow­er cord ground thru the AC mod­ule assem­bly which is mount­ed above it with four screws. The main cir­cuit board is also ground­ed via the two raised shield tabs and the dis­play mount­ing frame screws.

Canon Canola L100A — Gold Iridite coat­ed shield with bezel ground wire

The AC mod­ule installs on top of the shield along with the insu­la­tor sheet used to pro­tect the main board from acci­dent­ly short­ing to the shield. There is also a small ver­ti­cal­ly attached insu­la­tor that pro­tects the back edge of the main board from con­tact­ing the AC mod­ule frame. The green wire is attached to a tab at the front of the shield and is lat­er attached to the met­al bezel shield.

Canon Canola L100A — Base with AC mod­ule and insu­la­tor sheet installed

This cal­cu­la­tor uses a now obso­lete plug found on Canon and Monroe (Canon OEM) cal­cu­la­tors. It is a three pin rec­tan­gu­lar con­nec­tor with the cen­ter pin con­nect­ed to ground and cal­cu­la­tor chas­sis shield. This cal­cu­la­tor also has a small pop-out han­dle so that you can eas­i­ly car­ry it around.

Canon Canola L100A — back of unit han­dle extend­ed with AC plug and jack

It took a while after order­ing the capac­i­tor replace­ments to receive them due to some inclement weath­er affect­ing the north­ern USA, so I decid­ed to do some fur­ther research on how the cir­cuit board was wired togeth­er. One of my favorite ways to decode dou­ble-sided cir­cuit boards is to back­lite the board and pho­to­graph it and then use my image edit­ing pro­gram to enhance the front and back traces. I use an edge detect fil­ter that works good on cer­tain board types, and a neg­a­tive fil­ter that almost looks like an x‑ray image was tak­en of the board. This helps me to fol­low the traces across most areas of the boards, but is lim­it­ed by areas with large IC’s and mod­ules that block the light. I then have to resort to con­ti­nu­ity test­ing of the traces in those areas.

Canon Canola L100A — main board bot­tom traces
Canon Canola L100A — Backlit cir­cuit board with a neg­a­tive filter
PinTMS1824PinTMS1825
7VDD ‑8.54 volts9CP clock
8VSS com­mon23VDD ‑8.54 volts
9CP clock24VSS com­mon
10SCT030KC invert­ed
25TR2 Collector33VDD ‑8.54 volts
29KC invert­ed35SCT0
38TB336VGG ‑15.35 volts
40VGG ‑15.35 volts40VSS com­mon
Canon Canola L100A — chipset labeled pin listing

The CP clock is a 100 kHz sig­nal with a neg­a­tive 14.2 volt pulse that is 2.62 μs wide.

SCT0 is a 970 Hz square wave at minus 8.4 volts.

TB3 is a 25.1 kHz neg­a­tive pulse at minus 8.4 volts (no screen capture).

Now that all the elec­trolyt­ic capac­i­tors had been replaced and the cal­cu­la­tor test­ed, it was time to try anoth­er “RetroBright” pro­ce­dure on the upper case.
This time I used 12% hydro­gen per­ox­ide solu­tion for 26 hours. The first after­noon was some­what shady and cool, so I used an aquar­i­um heater in an out­er water bath to keep the solu­tion warm. The con­tain­ers were brought in overnight with the heater set to 86 degrees F. The fol­low­ing day was bright and sun­ny which kept the solu­tion vig­or­ous­ly bub­bling through­out the day with­out the heater.
Overall I give the results 95% with a few areas a light shade dark­er than the rest. All of the yel­low­ing was gone, and I think the results are an arti­fact of the injec­tion mold­ing process for the plastic.

Refurbished Canon Canola L100A calculator

The L100A cal­cu­la­tor is a well built unit by Canon, but mar­ket­ed as an entry lev­el cal­cu­la­tor. It does­n’t have per­cent­age, square root, or mem­o­ry func­tions but what it does do, it does very well. For an alge­bra­ic 10 dig­it four func­tion basic cal­cu­la­tor it is a rock sol­id unit.

Leave a Reply

Your email address will not be published. Required fields are marked *