HP-67 - 1612A03018 - Card Reader Advice Requested (was HP-67 - strange behavior)

[albertofenini]

Hello to everyone,
we have just received an HP67, we posted in the Classified section a WTB request for a logic board
but with no success, so we purchased a unit from TAS.
It was corroded, but pictured ON so we went for it.
As a matter of fact, the logic board was working, so we began our restoration procedure, both aesthetically and on the circuits.
We cleaned everything with a warm mixture of 50% water and 50% vinegar, and when we put all together it was working pretty well.
We have also rebuild the card reader but it was giving a "Error" message with any card.
However, that's the reason why we are writing it stopped to work, quite suddenly.
Or better, when we move the switch from OFF to ON, the display flash for a second a 0.00 or a string of 0.0000000000 then goes off.
You can see a short clip here



It must be also said that there is still some oxidation, very little on the board.

Also, just before to behave so we had this problem : inserting a card in the card reader
didn't turn on the motor, as opposite to when just reassembled that everything
was working correctly (gave Error message but the cards were going thru nicely)
So we checked the switches but nothing, now as a precaution the card reader has been inhibited by putting some tape
between the card reader switches and the keyboard plate, but even with the card reader pcb removed from the logic pcb nothing change

Has it ever happened to someone before ?
What could we check ?
Thanks in advance for any advice you would like to share with us

P.S. we have already check the obvious, switches, cleaned the holes in the logic PCB, battery was fully charged, and it behave equally with the card reader disconnected.

[teenix]

Hi Alberto,

Seeing that [0.00] was displayed proves that the CPU worked properly, albeit briefly. Does [000] appear if in W/PRGM mode?

Swapping the CPU board if you have a spare one, would verify the rest of the calculator operation.

Because it was corroded and cleaning seemed to have stopped it working, as a start I would look closely at the circuit board traces, then the battery/board wiring terminations, and the brass link at the charge socket.

All that corrosion probably came from the pcb traces and washing the board may have degraded one (or some) a little more than the circuit can handle. Maybe some "gunk" ended up (or dislodged) under a chip, or a weakened component connection has now become a dry solder joint.

Probably a rarity, but board components can be damaged by static electricity during handling.

cheers

Tony

[albertofenini]

Hi Tony
thanks for your reply, sorry it took so long to get back but we were out for vacation,
we started to work again on this unit.
We spotted two points on the pcb that were not clean and we cleaned them up,
in one case removing the circular orange capacitor and between to ICs (those with 8 pins)
What happens now is the following :



the display briefly shows a 8.00 and then it shuts off, the position of the switch W/PRGM - RUN
it does not affect the behavior (to be continued since only one video can be attached per message) ....

[albertofenini]

or, and this is even stranger , if we press ANY key prior to turn it on and we keep it pressed we get :



a steady 0.00 but with the key pressed the rest of the keyboard is non functional,
if we release the key than everything shuts off again.

Thanks again to everyone for helping us out !

[jonese]

Shot in the dark...

On the Woodstock's (21/25) wasn't there a ground circuit path carried over the daughter keyboard board between both connectors, such that you couldn't power it on without it fully assembled?? Is it possible that when you press a key while powering up that you somehow complete the circuit?

[albertofenini]

Hi Jonese
It’s true an HP67 belongs to the Woodstock family more than to the Classic family,
however I can’t figure out what the difference could be between powering up the unit with a key pressed or without ...
And it makes this with any key ...
We’ll check it out ground connections that’s a good idea, thx !!!

[teenix]

(09-16-2019 06:54 PM)albertofenini Wrote:  or, and this is even stranger , if we press ANY key prior to turn it on and we keep it pressed we get :
a steady 0.00 but with the key pressed the rest of the keyboard is non functional,
if we release the key than everything shuts off again.

Thanks again to everyone for helping us out !

The microcode executes exactly the same up until the point where the keyboard is tested, whether a key is pressed or not.

If a key is down at this point, the code will stay in a small loop until the key is released, then it will sit in the main loop waiting for a key press, card insertion, or PGM/RUN switch change.

If no key is pressed at the mentioned point, the main loop executes.

During the main loop, I tried to force the code to branch to other code sections but it did not result in a blank display.

Hard to say but sounds like a ROM issue which may cause a code failure after the mentioned point. The ROM addresses for the main loop are below 256, so the first ROM may have a failure.

cheers

Tony

[albertofenini]

Thanks Tony,
that's a good hint, now tell me if'm wrong:
Can I say that from an electrical standpoint the PSU works fine ?
If PSU was not to be working I shouldn't be able to get the display lit up, right ?
Since the unit was working before we cleaned it with water and vinegar, can I say that
because of that degraded around the ROM chip ?
Which chip is ROM0 and what can we measure to check what's wrong ?
Would you suggest to remove it and re-solder it ?
Thanks for help Alberto

[albertofenini]

Tony, when you say "the first ROM" do you mean the 1818-0268 so called ROM0
or the four little chips 1818-0231, 1818-0232, 1818-0550, 1818-0551 ?
Thanks for help, Alberto

[albertofenini]

Tony,
we re-worked all the logic board,
de-soldered all of the ICs, found little oxydation left from the previous cleaning procedure,
removed that as well, but once we have re-soldered all of the components the behaviour
remains exactly the same.
We have few working units, how can we compare a working unit with this - step by step -
to find out what's wrong with this logic board ?
Remember that when we receid the unit it was working (basic operation)
Thanks again !!!


   

[teenix]

(09-17-2019 07:29 PM)albertofenini Wrote:  Tony, when you say "the first ROM" do you mean the 1818-0268 so called ROM0
or the four little chips 1818-0231, 1818-0232, 1818-0550, 1818-0551 ?
Thanks for help, Alberto

Hi Alberto,

I'm not sure which ROM is which on the PCB, except for the larger ROM 0/display decoder.

Initially, the code executes through ROM 0, ROM 3, ROM 1 and ROM 0. Then it waits for a key press in ROM 0. This code must be functional or the [0.00] won't be displayed when you hold a key down on power on.

If you press a key, say [5], different code executes through ROM 0 then ROM 3 and back to ROM 0.

As the ROMs appear to be selecting properly, it may be bad data in a ROM address in these ROMs.

1 page in ROM has 256 words of ROM code.
ROM 0 has pages 0 - 3 ($0000 - $03FF)
ROM 1 has pages 4 - 7 ($0400 - $07FF)
ROM 2 has pages 8 - 11 ($0800 - $0BFF)
ROM 3 has pages 12 - 15 ($0C00 - $0FFF)
ROM 4 has pages 20 - 23 ($1400 - $17FF)

cheers

Tony

[albertofenini]

ciao Tony,
How can we identify the fault then ?
Would you suggest to replace all of them ?
Would an HP97 be a possible donor ?
Last, sorry for all these questions, what could we compare between a working unit and this one in
order to find out what could be possible wrongs (voltages, waves ?)
Thanks again to everyone !

[teenix]

(09-18-2019 09:34 PM)albertofenini Wrote:  ciao Tony,
How can we identify the fault then ?
Would you suggest to replace all of them ?
Would an HP97 be a possible donor ?
Last, sorry for all these questions, what could we compare between a working unit and this one in
order to find out what could be possible wrongs (voltages, waves ?)
Thanks again to everyone !

It could be a difficult one, I wish I could pull a rabbit out of the hat :-)

If it is a ROM problem it would try replacing ROM 0 or ROM 3. My first try would be ROM 3 (smaller - easier to replace).

However, it may not be a ROM issue as such. It could be a problem with a component (cap, resistor, diode etc) that "changed" after being cleaned and has upset the works.

Firstly, I would go back to things you have tried in the past and run your CRO over the board and see that clock signals appear to work ok, and power supply voltages are good. It might be something simple like a poorly operating capacitor.

The HP-97 ROMs are a different kettle of fish. They have the printer code embedded as well as other changes, different keyboard etc.

cheers

Tony

[albertofenini]

Thank you very much!
What’s a CRO ?
Do you know where we can find some good schematics about the 67 ?
Looks like there is a loro less documentation compared to what’s available for the HP65
We’ll keep posting our results !

[teenix]

(09-18-2019 10:57 PM)albertofenini Wrote:  Thank you very much!
What’s a CRO ?
Do you know where we can find some good schematics about the 67 ?
Looks like there is a loro less documentation compared to what’s available for the HP65
We’ll keep posting our results !

Sorry, a habit - Cathode Ray Oscilloscope - or these days just an oscilloscope. I thought you had a pico-scope which will help. Some of the wave forms shown in the 97 service document may serve as a guide.

My web site has some schematics that I put together in pdf format...

http://www.teenix.org/classSchm.zip

or try...

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

cheers

Tony

[albertofenini]

Hi everyone,
couple of updates on this post ...

We have received a working CPU board (thank you very much !!!) and thanks to this we have been able to test the card reader.

We have been able to assemble three different card readers:
two HP-65 style (no variable resistor to adjust the speed) and one HP-67 style.

One strange thing common to all of three is this :
if the motor is powered directly (attaching a battery) it runs to a good speed.
but when the engine is engaged from the logic board because a card has been inserted it runs to
probably at one third of that speed.
Any idea why this is happening ?

We also have tried a different keyboard logic, we have had some incompatibilities between card readers
and keyboard pcb with the HP65 so we thought this was the case.

However, on this other keyboard PCB we had broken the transistor that sits between the anode and cathode drivers,
it is marked M624 3-393, does anyone knows a good equivalent part ?

We have recovered the transistor from the original keyboard, but this does not allow us to test the
old keyboard again.

Of the two first card readers, one has good RA and RB, and also good DATA but keeps giving Error.
The other one, seems to be physically blocking the card.

The third one to our surprise works.
We have assembled it litteraly from the worst of the spare parts we had,
and it runs so slow that the card barely passes thru, however we tested it with the SD-04A card which
is a two sides card from the standard pac, and is about dates, we tested the DOW program and it works.

Tomorrow we will post videos and pictures.

On the board side : we took some measurements on the broken CPU board.
This is the one that has a suspected ROM0 not working.
We have found good Phi1 and Phi2 on the ACT, the CRC and the ROM0.
However, the side of the ROM0 that handles the segments of the display is dead.

Also, when compared with a working board, Idisp and Ibias are different.
The array of resistors looks good, and is about 24K Ohm for each resistor.
What should try ?
We will open a separate thred for this one

Thanks to everyone !!!

[teenix]

Running from the battery does not provide any speed regulation, it will run as fast as the power available lets it.

The card reader has speed regulation built in so as to compensate for fluctuating battery voltages during normal operation.

This is controlled by the 4K64 resistor. Have you tried testing the value of this part?

Putting a 10K variable resistor may help. I am not sure if the speed input to card reader chip can handle directly to ground or battery so may be wise to keep the variable resistor set around mid way.

Also, I have heard replacing the 6uF power filter capacitor with a tantalum one can enhance the motor operation. The small cap on the motor terminals might also be a problem. I think it also is part of the control circuitry for speed feedback, or maybe just for filtering.

The 67/97 card read/write can deal with slow or fast cards up to a point. As long as the data fits on the card within its limits it should work. The 65 is different as the data transfers have to be sync'd to the CPU but may still work with different card speeds, but again, up to a point.

cheers

Tony

cheers

Tony

[teenix]

(10-19-2019 11:24 PM)albertofenini Wrote:  On the board side : we took some measurements on the broken CPU board.
This is the one that has a suspected ROM0 not working.
We have found good Phi1 and Phi2 on the ACT, the CRC and the ROM0.
However, the side of the ROM0 that handles the segments of the display is dead.

Can you see LED data with the CPU board out? Maybe the anode driver is loading it incorrectly.

cheers

Tony

[albertofenini]

Hi everyone, we have split the thread in two separate threads.

This thread will be about the restoration of the chassis 1612A03018 that originally had
a bad CPU board.

We substituted the board with a working board received from a donor, and we have tested
three types of card reader.

Out of the three card readers only one was found "compatible" and was the one HP-67 style
The other two, were HP65 style as can be seen in the pictures.

We have played a little with the variable resistor but it is very sensitive and at one point it was not
reading nor writing anymore, so we stopped playing ...

It also must be noticed that is the only card reader we have rebuilt we square o-ring, and even
in the past HP-67 restorations we have done it was the only way to make it work,
as both round o-ring or single piece of rubber didn't work out well.

We have also replaced the three capacitors on the card reader pcb and the one on the motor.

Now we are stuck at this point :

Reads well one sided cards
Write (well) one sided cards

Read well two sided cards
Write (apparently well) two sided card (we have been successfully only once)
but when we read back a double sided card, it goes well on the first side,
asks for the second side but is sounds "slippery" in the meaning that accelerate toward
the end of the reading of the recorded second side and gives an Error message.

What could it be to make the unit behave like this ?

Pictures and clips are here : HP-67 - 1612A03018

P.S. it's only us or also your bench looks like this at the end of a working session ?

(10-20-2019 03:17 AM)teenix Wrote:  Running from the battery does not provide any speed regulation, it will run as fast as the power available lets it.

The card reader has speed regulation built in so as to compensate for fluctuating battery voltages during normal operation.

This is controlled by the 4K64 resistor. Have you tried testing the value of this part?

Putting a 10K variable resistor may help. I am not sure if the speed input to card reader chip can handle directly to ground or battery so may be wise to keep the variable resistor set around mid way.

Also, I have heard replacing the 6uF power filter capacitor with a tantalum one can enhance the motor operation. The small cap on the motor terminals might also be a problem. I think it also is part of the control circuitry for speed feedback, or maybe just for filtering.

The 67/97 card read/write can deal with slow or fast cards up to a point. As long as the data fits on the card within its limits it should work. The 65 is different as the data transfers have to be sync'd to the CPU but may still work with different card speeds, but again, up to a point.

cheers

Tony

cheers

Tony

[albertofenini]

We began some testing on the three card readers we have.

First we tried an HP65 style card reader, but we substituted the fixed resistor with a trimmer.
However, this is a noisy card reader, you will see that by just turning on the machine the channels
RA and RB were very disturbed and reading a card from the Standard PAC was not working.
However, from a mechanical point of view this card reader works fine also with blank cards
and with cards from an HP65 standard pac