HP35 problem with a calculator, maybe due to a faulty charger

[aurelio]

Sometimes I review my calculators, checking how they work and then recovering after the parade in their boxes
Of course I also do this with battery packs, charging them every time and checking their efficiency. Today I realized that it's better, maybe mandatory, to check first the functionality of chargers, because who knows, even at rest, they can get damaged.
It happened with one of these and unfortunately it also damaged (it would seem) the unfortunate calculator. I noticed this because while charging it showed strange behavior on the display. I immediately checked the power supply and unfortunately provides about 17V on both sides, instead of the voltage due.
The calculator turns on (with a charge battery pack), but shows erratic behavior by showing meaningless numbers on the display. any idea what might have been damaged?.
A good reason now to do the deep cleaning that we discussed in another thread days ago, and to see the interiors of my HP35 (moreover already repaired by HP years ago, as I see from an internal label), bad for the external label which deserves now the "spaghetti procedure" invented by Alberto, but first of all I must repair it

[Maximilian Hohmann]

Hello1

(03-20-2021 01:45 PM)aurelio Wrote:  The calculator turns on (with a charge battery pack), but shows erratic behavior by showing meaningless numbers on the display. any idea what might have been damaged

Before trying anything else I would connect a laboratory power supply set at something between 3.6 and 4 volts to the battery contacts directly (carefully watching the polarity). If it still behaves strangely you can start troubleshooting. With old battery packs you never know: It could be contact problems or a dying battery cell.

Good luck
Max

NB: And your story is yet another reminder to never run these old calculators from a wall charger, even the models which have some degree of overvoltage protection like the HP-35. I put all my chargers in a big box long time ago (with a little label on each of them showing the serial number of the calculator it belongs to) and run the calculators either from a laboratory power supply or externally charged batteries.

[aurelio]

(03-20-2021 02:11 PM)Maximilian Hohmann Wrote:  Hello1

(03-20-2021 01:45 PM)aurelio Wrote:  The calculator turns on (with a charge battery pack), but shows erratic behavior by showing meaningless numbers on the display. any idea what might have been damaged

Before trying anything else I would connect a laboratory power supply set at something between 3.6 and 4 volts to the battery contacts directly (carefully watching the polarity). If it still behaves strangely you can start troubleshooting. With old battery packs you never know: It could be contact problems or a dying battery cell.

Good luck
Max

Thank-you Max, actually I fed it with different battery pack aleady tested (I mean they work with other calculators) and the result is the same. I read on the forum that "this machine will work just fine without a battery pack installed. The power supply for the classic series machines delivers a filtered and regulated voltage to the electronics and a separate voltage to charge the battery pack". well, feeding the calculator without the battery pack inside using different already tested wall charger the behaviour I get is the same (erratic).

I'm scared that the wrong initial voltage (17V circa) in place of the 4,5 could have damaged the circuit, far away still to understand what, specifically

[Maximilian Hohmann]

Hello!

(03-20-2021 02:32 PM)aurelio Wrote:  ... actually I fed it with different battery pack aleady tested (I mean they work with other calculators)

That doesn't mean too much. I have tried it with an HP-35 and an HP-67 that happen to lie next to my computer. The same half-discharged battery pack works well with the 67 but fails to power the 35.

(03-20-2021 02:32 PM)aurelio Wrote:  I read on the forum that "this machine will work just fine without a battery pack installed. The power supply for the classic series machines delivers a filtered and regulated voltage to the electronics and a separate voltage to charge the battery pack".

Which is true only as long as the power supply does exactly what it is supposed to do. But just like the calculator, it contains 50 year old electronic componets which can fail any moment in a catastrophic way. Catastrophic from the point of view of the calculator that is.

(03-20-2021 02:32 PM)aurelio Wrote:  II'm scared that the wrong initial voltage (17V circa) in place of the 4,5 could have damaged the circuit, far away still to understand what, specifically

There are 17V across two of the terminals of a functioning charger (I just measured one of mine) so the 17V for itself is rather normal. Just try yours from a regulated power supply and see what you get. And if it is broken, at least you killed one of the least expensive classic HP calculators, unless it happens to be a "red dot".

Regards
Max

[aurelio]

(03-20-2021 02:58 PM)Maximilian Hohmann Wrote:  There are 17V across two of the terminals of a functioning charger (I just measured one of mine) so the 17V for itself is rather normal. Just try yours from a regulated power supply and see what you get. And if it is broken, at least you killed one of the least expensive classic HP calculators, unless it happens to be a "red dot".

Regards
Max

Thank-you Max,
Just consider that even the least expensive is for me dear.
I love my calculators all, indifferently

[teenix]

Maybe just pure chance the calculator is playing up this time.

With luck, it may be just a loose or dirty connector between the keyboard and CPU board upsetting the data sent to the anode driver or the CPU clock signal.

cheers

Tony

[aurelio]

(03-21-2021 01:44 AM)teenix Wrote:  Maybe just pure chance the calculator is playing up this time.

With luck, it may be just a loose or dirty connector between the keyboard and CPU board upsetting the data sent to the anode driver or the CPU clock signal.

cheers

Tony

Hi Tony, I've cleaned with isopropilic alcohol the connection between keyboard and CPU board, but the result does not change. I've to say, just to share, that the response on the calculator changes regularly, depending on the key pressed, for example if I key the 1/x or SQR or Ex I have each time different a recurrent results for each key, but if I press any of the numeric key the calculator turns off. I checked the circuit with a magnifier and the components and the solderings look externally all good

   

[teenix]

It does look clean

If you have another 35, maybe you can confirm the CPU board works or doesn't by swapping for a known good one.

Unfortunately, there is a small risk that swapping may damage a good CPU board. If you have a multimeter you can check the power supply rails first to see that they are all good.

The test points are labelled here from Tony Duell...

https://www.hpmuseum.org/cgi-sys/cgiwrap...i?read=385

cheers

Tony

[ThomasF]

Sad story, I hope you get it right Aurelio, fingers crossed!

Tony, as chargers becomes more and more unreliable, do you know the best way to charge battery packs outside the calculator?
Yes, there are HP-docks for that, but I don't have any, but I'm thinking of building one for each type (mainly for Classic and Woodstock). Do you know of any schematics that shows how the charger is connected to the battery, I guess that there is some kind of protection between the charger and the battery?

Best regards,
Thomas

[teenix]

(03-21-2021 09:37 AM)ThomasF Wrote:  Sad story, I hope you get it right Aurelio, fingers crossed!

Tony, as chargers becomes more and more unreliable, do you know the best way to charge battery packs outside the calculator?
Yes, there are HP-docks for that, but I don't have any, but I'm thinking of building one for each type (mainly for Classic and Woodstock). Do you know of any schematics that shows how the charger is connected to the battery, I guess that there is some kind of protection between the charger and the battery?

Best regards,
Thomas

A lot of HP circuits courtesy of Tony Duell are here....

http://www.hpcc.org/cdroms/schematics5.0/index.html

cheers

Tony

[aurelio]

(03-21-2021 09:37 AM)ThomasF Wrote:  Sad story, I hope you get it right Aurelio, fingers crossed!
........
Best regards,
Thomas

Thank-you Thomas, I hope too

[aurelio]

(03-21-2021 08:20 AM)teenix Wrote:  It does look clean

If you have another 35, maybe you can confirm the CPU board works or doesn't by swapping for a known good one.

Unfortunately, there is a small risk that swapping may damage a good CPU board. If you have a multimeter you can check the power supply rails first to see that they are all good.

The test points are labelled here from Tony Duell...

https://www.hpmuseum.org/cgi-sys/cgiwrap...i?read=385

cheers

Tony

Thank you Tony, CPU boards swapping was the first thing that came to my mind, but I also thought about the risk, however minimal, of being then with two 35 opened and, even worse, both not working...
Thanks for the advice, I measured the voltage and read, with my 8.5 V multimeter, where it should be from scheme 8.2, but I do not know the accuracy of my instrument and the tolerance allowed.
I measured on test points what I could measure with the multimeter i.e. Vss = 6.49 (expected value 6 V) and Vgg = -12.94 (expected value -12 V), what do you think of these results?

[teenix]

(03-21-2021 10:19 AM)aurelio Wrote:  Thank you Tony, CPU boards swapping was the first thing that came to my mind, but I also thought about the risk, however minimal, of being then with two 35 opened and, even worse, both not working...
Thanks for the advice, I measured the voltage and read, with my 8.5 V multimeter, where it should be from scheme 8.2, but I do not know the accuracy of my instrument and the tolerance allowed.
I measured on test points what I could measure with the multimeter i.e. Vss = 6.49 (expected value 6 V) and Vgg = -12.94 (expected value -12 V), what do you think of these results?

I think they are pretty good.

cheers

Tony

[CY-CL]

(03-21-2021 09:37 AM)ThomasF Wrote:  Sad story, I hope you get it right Aurelio, fingers crossed!

Tony, as chargers becomes more and more unreliable, do you know the best way to charge battery packs outside the calculator?
Yes, there are HP-docks for that, but I don't have any, but I'm thinking of building one for each type (mainly for Classic and Woodstock). Do you know of any schematics that shows how the charger is connected to the battery, I guess that there is some kind of protection between the charger and the battery?

Best regards,
Thomas

This is the best charging USB cable...

eBay Charging Cable

[ThomasF]

(03-21-2021 12:33 PM)CY-CL Wrote:  This is the best charging USB cable...

eBay Charging Cable

Thanks!

I actually have one of his cables (without the circuit) and it works great, but still I managed to burn a HP67.
I don't blame the cable, my problem was probably bad connection between the battery and the calculator.
Since then I don't really dare to charge or use a cable connected to them, but try to charge them externally, or keep the calculator off while charging.

(The 67 is up and running again after updating the main circuit board )

Best regards,
Thomas

[Maximilian Hohmann]

Hello!

(03-21-2021 09:37 AM)ThomasF Wrote:  ... do you know the best way to charge battery packs outside the calculator?

Since many years I have been using a laboratory/desktop power supply for that purpose. You need one which lets you adjust voltage and current. Set the current at 1/10 of the battery capacity for NiCd batteries and anything between 1/10 and 1/3 for NiMH. This also benefits the longevity of the batteries compared to the unregulated charging of the original chargers.

Regards
Max

[mfleming]

(03-21-2021 09:37 AM)ThomasF Wrote:  as chargers becomes more and more unreliable, do you know the best way to charge battery packs outside the calculator?

The simplest approach is to use a battery pack where the batteries can be removed for external charging. You can modify an existing pack by hand or buy a suitable pack from the previously mentioned supplier. I'd do what Maximilian suggests and toss all chargers in a box

[aurelio]

(03-21-2021 11:07 AM)teenix Wrote:  

(03-21-2021 10:19 AM)aurelio Wrote:  Thank you Tony, CPU boards swapping was the first thing that came to my mind, but I also thought about the risk, however minimal, of being then with two 35 opened and, even worse, both not working...
Thanks for the advice, I measured the voltage and read, with my 8.5 V multimeter, where it should be from scheme 8.2, but I do not know the accuracy of my instrument and the tolerance allowed.
I measured on test points what I could measure with the multimeter i.e. Vss = 6.49 (expected value 6 V) and Vgg = -12.94 (expected value -12 V), what do you think of these results?

I think they are pretty good.

cheers

Tony

Well, if the voltage on test point is the one expected it means maybe that the PSU works properly, so what would be the next step in the troubleshooting?
What does it mean that if I use any of the numeric keyes, the calculator turns off, while clicking on the functions the display shows numbers that vary, from time to time, with different criteria?

[teenix]

(03-21-2021 04:43 PM)aurelio Wrote:  Well, if the voltage on test point is the one expected it means maybe that the PSU works properly, so what would be the next step in the troubleshooting?
What does it mean that if I use any of the numeric keyes, the calculator turns off, while clicking on the functions the display shows numbers that vary, from time to time, with different criteria?

I would try swapping a known good CPU. If you feel this is too risky, then Harald may still have replacement Classic CPU boards which you can try instead.

--------

It may be a faulty ROM.

When the keys are scanned, a key code is generated and this value is used as an address offset in ROM so that the processor can execute the code for that key press. Very efficient.

If the keycode is invalid, the Program Counter could be set to an incorrect ROM location causing the code to run in error.

If one or more ROMs have failed then code execution will fail also.

When a key is pressed, the display is turned off before processing begins. It is possible the code is still running but in a dodgy loop that makes the calculator look dead.

cheers

Tony

[aurelio]

(03-21-2021 06:52 PM)teenix Wrote:  

(03-21-2021 04:43 PM)aurelio Wrote:  Well, if the voltage on test point is the one expected it means maybe that the PSU works properly, so what would be the next step in the troubleshooting?
What does it mean that if I use any of the numeric keyes, the calculator turns off, while clicking on the functions the display shows numbers that vary, from time to time, with different criteria?

I would try swapping a known good CPU. If you feel this is too risky, then Harald may still have replacement Classic CPU boards which you can try instead.

--------

It may be a faulty ROM.

When the keys are scanned, a key code is generated and this value is used as an address offset in ROM so that the processor can execute the code for that key press. Very efficient.

If the keycode is invalid, the Program Counter could be set to an incorrect ROM location causing the code to run in error.

If one or more ROMs have failed then code execution will fail also.

When a key is pressed, the display is turned off before processing begins. It is possible the code is still running but in a dodgy loop that makes the calculator look dead.

cheers

Tony

Thank-you Tony to help me in understanding., I've searched, while, for HP schematics and hardware on the web, reading expecially the Jaques's Laporte site... so it could be as you think a faulty ROM or relared circuit (I read about the reset on power-up). Well to know that it's possible to get replacement CPU boards for the classic, I'have just spare boards for the 41s (after convertion in CL
I've searched on the auction site and found only a few calculators for sale (in to be repaired conditions) .. not cheap and all from overseas with also import fee costs