Canon Canola 163

[Ramiro]

Hello Everyone, 

I acquired this nixie tube calculator this year but its not fully functional. 

The memory function and some of the other functions are none functional. 

I have gone in and replaced the capacitors but no results. 

I am waiting to purchase an oscilloscope to further diagnose this device. 

I was wondering if anyone had one of these calculators willing to help guide me in diagnosing and repairing as I do want to fully restore it. 

Also, I was wondering if anyone else had one of these calculators?

[Rick Bensene]

Greetings, Ramiro,

I don't have a Canon 163 specifically, but I do have Monroe's version of this machine, the Monroe 990

(http://oldcalculatormuseum.com/monroe9m90.html) .   Monroe was an OEM customer of Canon.  Canon would manufacture calculators to specification for their OEM customers such as Monroe, based on existing Canon calculator designs.   The Monroe 990 has some subtle differences from the Canon 163, but for the most part, they are cosmetic things, like cabinetry and color scheme differences.   Internally, the electronics of the Monroe 990 are virtually identical to those of the Canon 163.

You didn't go into much detail in terms of what is not functioning properly on the calculator.   Does the machine allow you to enter digits into the display?

The photo shows the display with all zeroes, which could be an indication that keyboard entry is not working.   If you can enter number, does addition and subtraction work properly? 

Addition and subtraction are operations are used as "subroutines" in the multiply, divide, and square root functions, as well as in the add and subtract to/from memory operations.   If the basic addition and subtraction operations are not working properly, then it's highly likely that multiplication, division, square root, and memory add/subtract operations will not function properly.   

If numbers properly enter into the machine, and addition and subtraction seem to work properly, any problems with multiplication and division/square root (square root uses parts of the divide function to do its job) are likely due to issues with the state machine logic that performs these functions.     

If you can't enter numbers into the machine properly, or numbers enter incorrectly or change with no reason (e.g., flicker from one digit to another, e.g., like changing from 2 to 3, then back to 2 in a single digit), then the problem likely lies in the delay line.  

These calculators use a magnetostrictive delay line to store their working registers, including the entry/display register, hidden registers for storing the first number of a math operation while the second number is being entered, as well as the memory register(s).  The bits are stored by sending torque pulses through a long coil of special wire inside the delay line package.  The pulses take a specific amount of time to go from one end of the wire to the other.   The wire is long enough, and has enough delay that all of the bits necessary to represent all of the digits in the registers of the calculator can be resident in the wire as torque pulses at the same time.  When the calculator is idle, the bits in the delay line are just recirculated...whatever bit pops out the end of the wire gets pushed back into the beginning of the wire.  While this idle recirculation is going on, logic picks off the bits that represent the entry/display register, accumulates four of them at a time (four bits represent each digit), then decodes the binary code for the digit into a 1-of-10 signal that drives the appropriate Nixie tube for that digit.   Thus, the display is continually refreshed, a digit at a time, with the content of the entry/display register.  This happens fast enough that the human eye can't perceive the refresh of the Nixie tubes.   If numbers can be entered into the machine and they remain stable in the display, then the delay line is probably working fairly well.   

If numbers enter  and then flicker or disappear, then it's very likely that the delay line is out of whack.   There is very sensitive amplifier circuitry that is part of the delay line that amplifies the tiny signal that comes out the end of the delay line, and component degradation due to the passage of time can cause this circuitry to become intermittent or stop working altogether.  It is possible, with care, to monitor what is going into the delay line and coming out with an oscilloscope, preferably a digital oscilloscope with a fairly deep acquisition memory so that comparisons can be done of what goes into the delay line, and what comes out after the delay period.  Delay lines are somewhat delicate electromechanical devices, and if the calculator has suffered any kind of strong G-force shock, it can cause damage to the delay line components inside the delay line enclosure.  This kind of damage can be very difficult, if not impossible, to repair.   Replacement delay lines using this type of technology are no longer available, as the technology has long-since been obsolete.

The DTL-family integrated circuits used in this series of calculators are generally quite robust and generally stand the test of time pretty well.  

To my knowledge, there is no service documentation available for this series of Canon and OEM customer calculators.   Certainly if such were available, it would make troubleshooting a whole lot easier, as with schematics it is possible to identify possible problem points and then locate where they are on the circuit boards, and probe the signals to see if they are doing what they are supposed to do, and much more easily identify any failures.   Repair of these machines basically comes down to having an understanding of the general architecture of operation of the machines, and meticulously tracing circuit paths between ICs and components to probe with an oscilloscope to try to identify any signals that don't seem to fit with the way things should be operating, then trace the fault(s) to their source.   

Findiing replacement ICs can be a chore, but they are out there.  Just avoid buying from Chinese ebay sellers that offer these ICs, as, sadly, while the IC's might have the right numbers printed on them, typically either they are just empy shells that will do nothing, or worse, they may be some other kind of IC that could potentially cause damage to other ICs in the calculator if it is put in the machine. Many, but not all, Chinese/Hong Kong/Taiwan IC sellers on ebay are legitimate, but the few that sell these "fake" ICs make it very hard to pick the right source.  Some of these bogus chip sellers have taken to having someone in North America to  "represent" them, making it appear that the chips are coming from somewhere in the US, but actually, they are the bogus chips from China.    So, just be careful if you have to search out replacement chips for the calculator.   

While I've babbled about a lot of things, the first most important thing to check out is the power supply.   Make sure that the supply voltage to the logic (generally 4.5 to 5V DC) is stable and clean (e.g., <2mV ripple, and no other types of noise), and that grounds are solid and ground has minimal noise on it.  Check the grounding points in the calculator to make sure they are all solidly connected.  

Also check the other voltages, including (be careful) the Nixie high-voltage supply...generally around 190V or so.   There are probably a few other voltages developed by the power supply for the drivers for the delay line input transducer and the output amplifier.  These voltages should also be stable, and free of noise. If anything with the power supply is amiss, all bets are off for being able to do any kind of reasonable problem diagnosis, as power supply issues can and will manifest as nothing working at all (blank display, no response to keyboard), or erratic behavior that can result in all kinds of strange malfunctions.   

If the logic power supply is over 6.5 volts or so, then it is possible that damage to the ICs could have occurred, as they are not intended to operate at supply voltages above 5V.   If the power supply is putting out more than 6.5 volts on the IC logic supply, then there could be numerous ICs that have been damaged, which would make the repair of the machine extremely difficult.   

Hopefully the power supply was checked out /before/ you powered up the machine and began testing it.   Generally, I disconnect the power supply from the rest of the calculator, and test the voltages with dummy loads put on them (power resistors of appropriate resistance and power rating) to verify that they are putting out the correct voltages with minimal ripple.  If lots of ripple is found on the power supply voltages, chances are some of the electrolytic filter capacitors in the power supply have become malformed over time, and are no longer properly acting as AC-blocking filters to soak up  the switching spikes of the diodes in the rectifier bridge (full or half-wave) and deliver a smooth DC voltage.   Sometimes, a diode acting in a rectifier circuit will fail, resulting in part of the DC component being missing, meaning the resulting DC voltage is lower than it should be, and can also cause excessive ripple on the output voltage.   

Of course, a thorough visual inspection of the power supply components is always a good idea before applying power.  Look for electrolytic capacitors that have bulging cases, fluid leakage around the leads, "fuzzy" stuff, usually whitish or yellow-tinted almost crystalline looking growths near the leads of the capacitor.   If a capacitor is leaking, the electrolyte can cause damage to circuit board traces it may come in contact with, corroding them such that sometimes the  circuit trace is broken.   Any capacitors exhibiting any kind of visual anomaly should be replaced with a replacement device that is the same capacitance and working voltage rating. If you can't find a capacitor with the same working voltage rating, it's possible to substitute a capacitor with the same amount of capacitance, but a higher working-voltage rating.  Never install a capacitor with a lower working voltage rating than the original, as that can be a recipe for disaster.   

Also, many calculator power supplies have very inexpensive potentiometers (variable resistors) to act as voltage adjustments to set the output voltage to the correct level.  Many times those potentiometers have exposed guts such that atmospheric contaminants can get on the resistive track and the contact that moves across the resistive track, causing intermittent or incorrect behavior.  

Cleaning these potentiometers with a shot of a good contact cleaning solution will generally make them much happier.  After years of sitting around, it's amazing how much gunk can build up on parts like this just from exposure to the atmosphere.  I recommend a product called De-Ox-It, which works fantastic for cleaning contacts in calculators.   

Another thing to check are the fingers on the edge connectors on the boards that plug into the backplane.  Atmospheric contaminants and oxidation can occur on the fingers, leading to poor contact, which can cause malfunctions.   Use a good contact cleaner to clean all of the edge connector fingers on both sides of the circuit boards, and also give a shot of it along all of the contacts in the edge connector sockets that the boards plug into..

Sorry for rambling on for so long, and kind of in random order, but this is a pretty good synopsis of the things to check and problem areas to look for before powering up the calculator and giving it a test workout.   

Many times, the problem lies with corrosion or gunk collecting on circuit boards and connectors, as well as power supply components.  Cleaning these up in a lot of cases will restore a malfunctioning machine to operation assuming all else is good.   Getting rid of any problems in the power supply and interconnects puts things in a state where you can trust that any problem you investigate in the logic is caused by something in the logic rather than some weird power supply or interconnect issue.

I hope that this is helpful in some way, both to you, as well as others who read this who might be thinking about digging into repairing their  malfunctioning vintage electronic calculator.  Many of the things discussed here will apply to a wide range of electronic calculators from the mid to late 1960s and early 1970's.

Sincere regards,

Rick Bensene

The Old Calculator Museum

http://oldcalculatormuseum.com

Beavercreek, Oregon

--
You received this message because you are subscribed to the Google Groups "OldCalculatorForum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oldcalculatorfo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/oldcalculatorforum/13c1cc0e-f6aa-48e9-a57f-b018f03e2a65n%40googlegroups.com.

[J Ongena]

Dear Ramiro and Rick, A schematic is available for the Monroe 925, which has resemblances to the Canon 163. See : http://madrona.ca/e/eec/calctd/Monroe925.pdf

Perhaps it can be of help !

Jef Ongena

To view this discussion on the web visit https://groups.google.com/d/msgid/oldcalculatorforum/CAMbJjuTVAAs5oa5cBmsXDriNKnVCbRwt%3DHbAPeFM0YL%2BhQiLSA%40mail.gmail.com.

[bruceflamm]

Hi Rick

This is Bruce in Southern California.

Although I did not ask the question I want to thank you for answering it.  It was kind of you to provide such a thorough explanation of the general troubleshooting routine.  I’m still hoping to make it up to Oregon one of these days and if you ever plan to visit California  please let me know.

 Bruce

Sent from the all new AOL app for iOS

To view this discussion on the web visit https://groups.google.com/d/msgid/oldcalculatorforum/CAMbJjuTVAAs5oa5cBmsXDriNKnVCbRwt%3DHbAPeFM0YL%2BhQiLSA%40mail.gmail.com.