| Manufacturer: | Canon |
| Model: | Canola EP151 |
| Year: | 1971 |
| Form: | Desktop |
| Functions: | Basic four, rounding, accumulator |
| Number of Digits: | 15 |
| Display Type: | PRI |
| Logic Technology: | SSI,MSI,LSI |
| Memory Technology: | MSI |
| Diodes: | 206 |
| Transistors: | 57 |
| Principal ICs: | SN3900/4500 Series |
| IC Complement: | Texas Instruments SN3920 (24), Texas Instruments SN3925 (24), Texas Instruments SN3931 (13), Texas Instruments SN4552 (17), Texas Instruments SN4553 (6), Texas Instruments SN4554 (56), Texas Instruments SN4556 (9), Toshiba TM4105
Texas Instruments TMS3314
Texas Instruments TMS4134
(152 total) |
| Tech. Data Level: | 2 |
| Tech. Data Source: | RE |
| Tech. Data Pages: | 21 |
| Tech. Data: | Schematic (pdf) |
|
While the mechanical calculator market declined into oblivion around the early 1970s, there was still a demand for printing calculators.
This is an early example of an electronic printing calculator.
There is no display besides the printer and it doesn't print keypress digits until the entire number is entered.
The printer uses a form of electrographic printing requiring special paper.
Rotating styli scan across the paper delivering high-voltage pulses to burn pixels
on the paper.
A MOS character generator ROM like those used in CRT computer terminals of the day
is used to generate the character pixel patterns.
Variations of electrographic printing go back decades earlier,
for example Western Union Desk-Fax facsimile machines from the 1940s.
| Notes: |
- Unit will seem dead (no motor activity) until paper is manually fed down into the printer.
- Upon power-up, it may be necessary to press both the C and CM keys to clear OVF.
- To remove the printer cover, unscrew the ground wire, remove the curved metal spring bar and plastic tear-off bar,
then pull up on the round aluminum cover to clear two internal ball-bearing spring catches.
- This model and other Canon models of the same technology and period, such as the Canon 163,
can be a nightmare to maintain and keep functional.
The soldering at the thru-hole stubs in lieu of plated-thru holes develop cracks, resulting in faulty connections.
Handling of the boards in the process of finding and repairing a fault may temporarily 'fix' the fault, or produce another faulty connection.
|
Internal view.
The cylindrical cover on the print mechanism has also been removed here.
|
Board 3, with the PMOS shift register memories.
|
Board 3 solder side.
|
Internal top view.
|
Board 4, containing the control state machine.
|
Board 4 solder side.
|
A sample of the print output.
Note the thousands marker - a dot above and between the digits.
The last calculation is 111111*111111 producing 12345654321.
The 4 in the MSD is some fault.
|
Board 6 with the character generator ROM.
|
Board 6 solder side.
|
The print mechanism is removable as a modular unit.
|
Cylindrical cover removed to expose the rotating disc which carries the four print styli.
The disc rotates at 720 RPM, printing 4 scan lines in one rotation.
|
Front view of the print mechanism.
|
Japanese date code on the inside of the case.
The 46 is the Showa year of the emperor and converts to 1971.
|
- Unit Log -
| Serial Number: | 200464 |
| Year of Manufacture: | 1971 (Integrated circuits stamped with 6927, 7031, 7126, etc.) |
| Date of Receipt: | 08 Dec 2000 |
| Source: | Brownlee Office Equipment. |
| State upon Receipt: | Some neon lamps light up. No printer motor activity. |
| Current State: | 05 Apr 2002 - Stuck in print mode. No known source for the special electro-sensitive paper. |
| Date: | 02 Apr 2002 |
| Procedure: | Cleaned. |
| Date: | 02 Apr 2002 |
| Procedure: | Portion of reverse engineering done:
power supply, printer, clock, memory, character generator. |
| Date: | 05 Apr 2002 |
| Procedure: | Filmy gunk on PCB edge connectors cleaned off.
Upon reassembly unit seems to function properly, as far as it is possible to test it in the absence of the special paper required.
Digit lamps and motor function as would be expected and a scope indicates the pixel stream is present at the printer. |
| Date: | 05 May 2002 |
| Procedure: | Reverse engineering complete. Unit is back to being stuck in print more (CC2 presumably). |
| Date: | 27 Dec 2004 |
| Procedure: | Half-role of aluminized paper received.
Unit mostly functional, addition and subtraction work.
Multiplication results in overflow lamp coming on,
although the correct result will accumulate into the accumulator.
MSD seems to print as '4' when it should be '0'. |
| Date: | 2021 Feb [x] |
| Symptoms: | Upon power-up, unit is stuck in print mode (printer runs continuously as soon as paper is inserted).
Digit lamps increment as printer rotates and will increment if the printer styli disc is rotated manually.
This goes away after a few minutes warm-up, but also returns intermittently. |
| Analysis: | When unit is in continuous-print fault, CC2 observed to be high.
R measurements around U4106 show poor conductivity around feed-thru stub for Vcc to pin 14.
Also bad at stub for Vcc to U4112. |
| Procedure: | Resoldered. |
| Date: | 2021 Feb [y] |
| Symptoms: | Printer will operate with paper and pushing the paper-feed button.
Clear resets digits lamps to first digit, pressing any or most keys then increments the digit lamps and results in overflow lock-up.
May clear up to function after some warm-up, but erratic.
Digits 0-3 print as 0, 1-7 correct, 8-11 as 0, 12-14 correct.
Zero-blanking may be working now.
Should look at 4-bit of digit-phase comparator. |
| Procedure: | Boards 1-4 checked for continuity through feed thru stubs for IC Vcc & GND pins.
OK except ~10 ohms at U1106.14 & U3208.7.
Resoldered. |
| Date: | 2021 Mar 03 |
| Symptoms: | Paper feed works but no print cycle on Clear.
Over flow lamp always on except C or CE held down.
CM results in increment of digit lamps.
Eventually, overflow lamp will stay off after C/CE. |
| Procedure: | CM produces CC2 pulse of 0.2mS.
When overflow will stay out this pulse is 0.4mS : one state cycle.
CC6 also pulses in sync with CC2.
Bad connection through stub for FCC4 to U4105.13.
Resoldered.
Printer now operates and prints on C/CE/CM, but cycle ends up with overflow lamp lit (and operation lockout).
Working from observed half-state-cycle period of CC2, speculate that digit timing counter is faulty, double clocking state cycle FF.
Bad connection through stub for D8 to U1206.2.
Resoldered. |
| Date: | 2021 Mar 15 |
| Symptoms: | Backplane pin N5C2 bent while inserting board extender due to inadequate chamfer on extender edge. |
| Procedure: | Connector N5A/C removed from backplane, pin C2 replaced with unneeded pin from A8. |
| Date: | 2021 Jul 01 |
| Procedure: | Extender boards completed.
Simulation was also completed earlier. |
| Date: | 2021 Jul 02 |
| Symptoms: | Upon power-up, overflow is lit and will not clear. |
| Procedure: | B register data is 1 but will 0 when C is depressed.
Phase-1 clock is only a few volts amplitude, not pulling down to -22.
Bad stub on board 1 between V-22 and 560R pull-down for phase-1 driver base.
Resoldered. |
| Date: | 2021 Jul 03 |
| Symptoms: | 12345678912345 # prints as 12000078910000 #. |
| Procedure: | Problem of 2021 Feb [y].
Earlier work on board 1 did not fix this problem, handling must have just temporarily improved a bad solder joint.
Bad stub feeding FTD4 to U1107.2.
Resoldered. |
| Date: | 2021 Jul 03 |
| Symptoms: | Flaky behaviour, entering numerals may immediately results in OVF, or entering numerals is possible but operation key results in OVF.
Varies with position of DP switch. |
| Procedure: | FPC2 sometimes does not toggle, FPC1 sometimes toggles but reverts quickly to U7107.8=HIGH.
FPC2 FF (U1107.8-13) connections resoldered.
Intermittent fault sources not specifically found.
FPC1 not dealt with. |
| Date: | 2021 Jul 04 |
| Symptoms: | Contrast tweaked higher. |
| Date: | 2021 Jul 04 |
| Symptoms: | A number can be entered, but an operation key may result in OVF.
Reassembled, perhaps re-attempt in future.
Should construct an electronic substitute for the printer, to allow display and testing without using paper. |
