Imperial Desktop Electronic Calculator

[David Carter]

Have an issue trying to get her to work.

There is a zener diode which tests faulty but with no schematics I can't identify value

The only writing on it is W150 any ideas?

[osa...@gmail.com]

I'll ask on a few Facebook groups....

[osa...@gmail.com]

Do you have the model number of the calculator?

[Manfred von Gunten]

A Zener diode with the designation W150 (often in connection with 1.3W to 1.5W power) is a

Silicon power Zener diode specifically designed to stabilize or limit voltages to a nominal value of 150 volts. These diodes are used in power supplies, surge protection circuits and clipping circuits.

Important technical data and characteristics:

 

    Zener voltage (

    ): 150 V

    Power consumption (

    ): Typical 1.3 W, 1.5 W or 2 W, depending on the exact design

    Design: Often axially wired, e.g. DO-41 or DO-204AM

--
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 visit https://groups.google.com/d/msgid/oldcalculatorforum/3c8df86c-bf02-4a41-a21f-bc6df10bb4d5n%40googlegroups.com.

[David Carter]

The rear label with makers name on had deliberately been covered with black paint to obscure it !

Managed to get enough off to read - Imerperial 1260

The PCB has E54914 GTC-A

Serial Number is 52537

Made by Litton Imperial Typewriter Company

[David Carter]

It's voltage rating of the Zener i'm worried about.

The Main board with the zener has 3 AC voltages passed from the separate transformer board.

That being as stated on the PCB 20V, 3.5v and 30v

The Zener is attached to pin 1 which is the 20v 

[Pieter La Grange]

That sounds like AI generated garbage.

David, zeners are 99% of the time used in vintage calculators to regulate voltages.

You will have to reverse engineer the PCB around the zener to figure out what it is used for and what its value should be.

Is the calculator working, is part of it working, what voltages can you measure? How do you know the zener is faulty?

[Patrick Souty]

Hello,
try the following test:
- replace the dead Zener by a 5.6V Zener (=Vz1) and check that the DC/DC converter oscillator is running (check at the transformer output with an oscilloscope) and then measure the 20V output DC value = V1

- if V1 < 10V, replace the Zener by a 10V Zener (=VZ2) and measure the  20V output DC value = V2

- if V1 >= 10V, replace the Zener by a 3.3V Zener (=Vz2) and measure the  20V output DC value = V2

- then the correct Zener value should be:  Vz = Vz1 +(20 - V1) / (V2-V1) * (Vz2 - Vz1)
 

[Patrick Souty]

This is an exemple of a DC/DC power supply of a vintage Royal Digital V calculator using a VFD display and probably similar to your calculator.
This schematic is extracted from http://madrona.ca/e/eec/calcs/RoyalDigitalV.html. On this excellent site, you can find a lot of schematics of vintage calculators.
Reverse engineering of the PCB will help you to identify key parts of the schematic and find potential failed parts.
However, as it is difficult to determine the transformer ratio, reverse engineering will not permit to determine the Zener value.
That's why test and try procedure is necessary.

[osa...@gmail.com]

Wow! That's a very useful site, I had no idea about it. 

[David Carter]

Tested zener with these results

Simple diode test get 0.055 one direction and 0.056 the other

Basically turned into resistor with a measured value of 55.2 ohms!

[Pieter La Grange]

That sounds normal. I think that was a common thermal failure mode for a zener diode, turning into a short.

[David Carter]

I've created a diagram of the PCB top with Traces from underside in red

Thanks for any help

[Patrick Souty]

That's an impressive job. Which SW are you using to draw it ?
Is it a single side PCB ?

[David Carter]

Just Affinity Photo/Design

Yes just single sided nicely hand drawn traces!

The black lines are top surface link wires

[David Carter]

I can't figure out how all the voltage regulation occurs or how they even rectify to DC.

I'm more used to 80's computers!

Bit concerned if this has suffered in some way a catastrophic event to toast the Zener and cooked the IC

Or do you think the Zener just took one for the team??

[Patrick Souty]

I will draw the schematic with Kicad and try to understand the power supply design.
The processor seems very similar to the Rockell 15473PA of the Berkey 4030 that I have renovated.

[Douglas Miller]

I find it very odd that they are feeding A.C. directly into a zener like that, without any rectification. Are you certain that it is really A.C. coming into the PCB? Or is there rectification happening on the power board? Have you observed the power with an oscilloscope? Could you be mistaking half-wave D.C. for (a lower voltage) A.C. ?

--
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 visit https://groups.google.com/d/msgid/oldcalculatorforum/7fe4dc28-cac2-47ac-88dd-b51586853b46n%40googlegroups.com.

[Patrick Souty]

AC signals are fed through D3 and D4 which acts as rectifier diodes and filtered by C4 and C6. So that seems normal.
However aren't there any wires between some AC output pins on the power board side ?
I have started the drawing of the schematic and will release soon a preliminary version.

[David Carter]

It's definitely just 3 AC voltages coming from a triple tapped transformer nothing on the power board apart from fuse switch and transformer.

I'm at a loss

Some photos Attached

The Zener at D5 is still removed

[David Carter]

Hi 

Decided to bite the bullet and give you guys a chance by removing the display!

Better photo of top of the board

[Douglas Miller]

It's not the "conventional" use of a zener, but I think I see what's going on. D5+C5 are acting as a voltage drop, and that 560 ohm resistor presents the voltage being delivered to the circuit. That's the only explanation I can think of for the + side of C6, D5, and C5 "going nowhere". Still, this looks wrong to me. But you often see this sort of thing in circuits of this vintage, and they require rather detailed and in-depth analysis to truly figure them out. But, if it is a voltage drop then the rating of the zener is NOT the voltage delivered to the circuit, but rather 20 - that.

To view this discussion visit https://groups.google.com/d/msgid/oldcalculatorforum/a5cc9707-1309-451a-a2a2-6be43b230dd2n%40googlegroups.com.

[Pieter La Grange]

Well done tracing the board.

I drew the basic schematic on a piece of paper and couldn't understand something until I saw the missing resistor that connects the D4_C6 node to C5_D5. What is its value?

You received this message because you are subscribed to a topic in the Google Groups "OldCalculatorForum" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/oldcalculatorforum/V5ijD7gDWpQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to oldcalculatorfo...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/oldcalculatorforum/a5cc9707-1309-451a-a2a2-6be43b230dd2n%40googlegroups.com.

[David Carter]

Hi 

That will be R11 @ 560R

[Pieter La Grange]

I will have another look later, but from what I can see D5 clamps the DC voltage to a value that is supplied as is to the IC.

There are 3 voltage rails if we ignore the filament circuit.

AC5 is the one rail going to the IC. That rail is connected to the positive side of C3 and C4. AC1 is the other rail also going to the IC. AC1 is clamped to the C3-C4 positive rail with the zener.

So unless I am wrong that broken zener is equal to your IC supply voltage.

To view this discussion visit https://groups.google.com/d/msgid/oldcalculatorforum/5aa561b1-77b5-4cef-a571-ce761186e688n%40googlegroups.com.

[Patrick Souty]

This is an extract of the schematic drawn with Kicad. The Zener D5 seems to be used to generate the regulated supply voltage for processor M1.
The Zener value is probably 20V or 25V but it's difficult to be sure of the exact value.
On my Berkey 4030 calculator with a processor Rockwell 15473PA processor  which seems similar to your 15350A , it was running correctly with an initial supply voltage of 21V.

On the Keystone 2050 with the same 15473PA processor, the voltage is 25V. For these NMOS LSI devices, I think that the supply voltage is not very critical.

I suggest then than you replace the dead Zener with a 20V Zener and test if M1 is running.

If it doesn't run or with erratic behavior, replace the Zener with a 25V value.

I will search on the net for other possibly known calculators using the 15350A and may be we could find additional information on the supply voltage.

[Patrick Souty]

The 15350A seems to be used in Digitron 1214 and Olympia CD402 (source calcuseum.com).

[David Carter]

Incredible work everyone. I will get on with fitting a new zener tomorrow.

Everything crossed that the failure of D5 hasn't toasted anything else, especially the main chip!

It'll be great if this is all it takes to get her back working!

[Patrick Souty]

After review of other schematics using similar Rockwell chips, some of them are using 16V voltages, so I suggest to use a 16V Zener 1st and then 20V and so on.
It would be sad to burn the chip by accident.

[Patrick Souty]

After analysis of the picture of the PCB bottom, the larger width of the wires connected to pin 6 and 8 of M1, confirms that they are the supply voltage for the chip.

[David Carter]

It would be sad to burn the chip now.

What do you think may have caused the zener to go bad apart from old age and more importantly what and where may damage have occured as a result?

[Patrick Souty]

It is difficult to answer.
If the failure results from a voltage surge on the main input, it could have burned the Zener and maybe the M1 chip.
If the failure is coming from the aging of the Zener, then I think the only induced stress was on D4 diode but you can check easily if D4 is still alive.
Warning, I have just identified an error in my previous schematic.
Wait for an update and simulation of the behavior to be sure of the conclusions.

[David Carter]

D4 tested as a perfectly healthy regular diode.

I remember I was concerned when I first looked at this machine there was a small solder blob or similar between the negative leg of C5 and the trace right next to it leading to pin 6 of the power connector 30v AC

I am sure someone had replaced C5 previously.

The blob fell away when I examined it. this is a possible cause.

What are your thoughts on what that might have caused?
Taking a very close look at the board you can see a slight mark!

[Patrick Souty]

Failures of tantalum and electrolytic capacitors are common in vintage electronics.
The failure of the original capacitor is probably not the cause the Zener failure but the failure cause could be the same in case of a voltage surge.
Or maybe, the guy who has replaced the cap, has over heated the PCB and burned the Zener that was close to C5.

[Patrick Souty]

Could you check on the connector coming from the transformer if there is any connection inside the transformer between 1 and 3, 1 and 5 and 3 and 5.
Because as it is, I cannot understand how it works without a DC path between 20V AC coil and 30V AC coil.

[Pieter La Grange]

The zener most likely failed due to overheating and not age. Check R11 to make sure it is also still in good condition.

I have tested and repaired many vintage calculators from the 1960s and 1970s and have yet to find a single faulty electrolytic capacitor. I have found many broken transistors, diodes, zeners, rectifiers, potentiometers and reed switches.

To view this discussion visit https://groups.google.com/d/msgid/oldcalculatorforum/dd1b18ee-b604-443c-a31a-0c80077d489an%40googlegroups.com.

[Pieter La Grange]

Patrick, there will be no connection between the different transformer outputs. From what I can see it will work fine as is.

What do you mean there is no DC path between the 20V and 30V coils?

[Patrick Souty]

Peter, I mean the Vss node of my schematic (D5 anode and C5-) is not connected to the 20V coil or D4 . That means that no voltage can develop across D5 and C5. This is confirmed by LTSpice simulation. There is a DC voltage across R11 and C6 only.
So there must be something wrong in my schematic, excepted if there is a hidden connection inside the transformer. I have checked the drawing from David and found it OK.

Same as you, I have never found a failed electrolytic capacitor, but many dead tantalum. However a lot of people reports dead electrolytic caps in vintage electronic. I assume it depends on the voltage margin used in the design.

Its a good idea to check R11 though I think the failure mode of resistors is rather open circuit that short circuit.

[Pieter La Grange]

In your schematic R11 (560) is not connected correctly. When you mentioned an error in your schematic I thought you were talking about that.

To view this discussion visit https://groups.google.com/d/msgid/oldcalculatorforum/f0416a56-5822-4937-b2d6-3616f43cd1ban%40googlegroups.com.

[Patrick Souty]

Here is the full schematic including R11 correction.
Warning: it may still contain errors and the pinout definition of M1 is uncertain.

[David Carter]

Transformer Output Winding Results:-

1-2 20v = 25R6

3-4 3.5v = 1R1

5-6 30v = 53R4

No other interconnects found between windings

[David Carter]

Great news and so much thanks to everyone who put in so much work!

She lives again.

A 16v Zener worked.

Need to do some work on the case cleaning wise and recreate a label.

Then we have a fully working calculator from I guess 1974 according to the M1 code!

Brilliant.

Not much info on the web for these Imperial Calculators.

Next projects include:-

Sanyo ICC-82D and Sharp ELSI MATE EL-1195 very small printing calc!

[Pieter La Grange]

It is a good feeling to see a calculator come back to life!

Let us know when you start with the ICC-82D. I have already repaired a few of them. The schematic is similar to the ICC-0081 which can be found on the internet.

To view this discussion visit https://groups.google.com/d/msgid/oldcalculatorforum/74608975-1e05-44d5-8ed6-c533d3431525n%40googlegroups.com.

[David Carter]

I have made a start with the ICC-82D a lovely piece of kit.

Started with making some new 3D printed battery boxes.

Some in clear and some in white. I like the idea of clear at the moment.

[David Carter]

Replacement label done

[Patrick Souty]

I was wrong when I wrote that my Berkey 4030 with a Rockwell chip 15473PA was using a 24V supply voltage. Actually, the supply voltage of the chip is 16V (Vdd = 24V, - Vss = 8V).
So I have reviewed again some schematics using Rockwell chips from the same period.
A5901CA (from Microlith 113G): 15V

A1241PA (from Lloyd’s Accumatic 55 and Keystone 395): 15V

15471PA (from Digimatic D8): 16V

15330PC (from Xonex ELSI III): 15V
Then a 16V or 15V Zener was indeed the right value for the 15473PA chip inside the Imperial Destop calculator.