|
|
A design from TEAL (Tokyo Electronic Application Laboratory), OEM'd for Riccar Sewing Machine Co. of Japan. Riccar doesn't seem to have given this calculator a model designation, this may have been the only foray by Riccar into calculators. From the design and implementation it appears to have very shortly preceded or be contemporary with the Royal Digital I, another OEM product from TEAL.
This model features zero-blanking, unusual in Nixie and pre-microprocessor calculators, implemented at the expense of an additional IC and assorted components.
The unit shown here is in the possession of Jef Ongena of Belgium. It had been in his family since new, but stopped working in 2020. Jef set out to repair it and found the power supply had failed. Repairing the power supply however, merely uncovered another problem. Repairing that uncovered a third problem, and so it went for six major faults, including four bad JMOS ICs. An over-voltage situation from the power-supply fault had likely taken out the ICs.
After considerable effort, communication across two continents and a 'remote' reverse-engineering via photographs, Jef successfully restored the unit to full function.
Replacements for two of the failed ICs were scavenged from orphan boards from an unrelated defunct calculator. This was fortunate as JMOS ICs are very difficult to obtain for repairs, but no such luck was forthcoming for the other two, which were of more complex functionality. In the alternative, substitutes were designed and constructed from 4000-series CMOS ICs. See the JMOS page for some commentary on this approach for repairing JMOS-based calculators, and the schematic for details.
Notes: |
|
10401 | |
1971 (PCB edge connector labeled 71.5.10) | |
Correspondance with Jef Ongena / 2021 Jan. | |
Unit owned by J.O. Physical repairs by J.O. | |
Fully functional (2021 Summer). |
2020 Fall | |
Fault #1.
Nixies present a non-sensical, 'garbled' display. | |
Power supply failure. | |
Transistors and capacitors in power supply replaced. An over-voltage from the power supply may explain the subsequently-discovered failure of four ICs. |
2021 Winter | |
Fault #2.
LSD nixie is overly bright and showing multiple numerals. | |
Symptoms are indicative of the nixie being stuck on, i.e. not being multiplexed. ΦD1~/UK5.5 is stuck LOW. ΦD ring counter still operates however, so UK5 M5812 is otherwise OK. | |
Outboard circuit using 4013 designed and constructed to regenerate missing ΦD1. |
2021 Spring | |
Fault #3.
No display on power-up. | |
Low resistance between –24V rail and GND on main board.
With board connected to ~ –2V (drawing ~ 1A), IR camera shows UF3 heating up.
UF3 µPD105 IC failed, shorted between the power pins (~ 80 ohms). | |
Substitute for µPD105 designed and constructed from two CMOS 4557s, one 4013 and assorted components.
Later, a smaller version is created using one 4517 and a 4013. |
2021 Spring | |
Fault #4.
Nixies display 7&9 numerals lit simultaneously. | |
Bitstream through N is all-1s, 7&9 is from partial decoding of 1111 in UJ1 µPD116C.
All-1s also sends machine into error state.
Inverter UD6.6-8 failed. MOSFET gate at UD6.6 has non-linear conductivity to GND (greater Vg produces greater conductivity). | |
Substitute using bipolar designed and tested.
Eventually resolved by replacing UD6 µPD13C with a µPD13 scavenged from a surplus board. |
2021 Spring | |
Fault #5.
Decimal point is appearing one digit lower in the display than it should be. | |
DPD counter is getting an extra shift pulse, resulting in the DP being turned on one digit early.
ΦD13 is not being suppressed when it should be.
Inverter UD2.1-13 failed with a leaky drain-to-gate. | |
UF3 µPD105 substitutes adapted to take over inverter function, using a bipolar or clock enable input.
Eventually resolved by replacing UD2 µPD13C with a µPD13 scavenged from a surplus board. |
2021 Spring | |
Fault #6.
Addition and secondary operations function, but subtraction, multiplication and division fail. For subtraction, a negative result produces ERROR. Multiplication produces an odd display of zeroes but with leading-zero blanking partially effective. Division produces ERROR. | |
Subtraction fault suggests issue with Ten's Carry control of transition to ST22 versus ST15. | |
In pursuing ST15/22 question, S,M,D function was restored after some work around MO2 and MO3.
Precise failure not known.
All operations now work. |
2021 Summer | |
Prototyped IC substitutes finalised into soldered boards and fitted into case interior.
Unit fully restored and complete. |
TEAL Riccar
Calculators | Integrated Circuits | Displays | Simulations EEC |
bhilpert |