HP-67 1705A01234 [erase other's cards, writes and sometime read back its own cards]

[albertofenini]

Hi everyone

This is quite a weird one ..
Any card previously written is erased when read by this HP-67
This happens also with what it should be a write protected corner clipped card
It is however a blank card, recorded and than corner clipped

I have checked what was obvious to me i.e. :
the W/PRGM switch really switches from one state to the other as it should
PGMSW pin at CPU board connector only goes to GND only when W/PRGM switch is moved

I still have to try if the HDS and WPS sequence is correct when a card write protected and one not write protected are inserted

I have written a simple program (math operations) recorded it with this HP-67 and while it is
not read by any other HP67, it is sometimes read back by this unit

This happens since I have slightly adjusted the pressure of the wheel

While I have not been able to record CA and CB, I have successfully recorded RA and RB when reading a card
and WA and WB when writing a card.

As always, any help will be greatly appreciated!
Take care !!!

Pictures here : HP-67 1705A01234

[teenix]

The write protect checking is not activated when reading cards so shouldn't have any effect.

The only way the cards can be erased while reading is if a current is passing though the heads, or the head is magnetized somehow. I can't see any other way. However, you say the cards can be read back sometimes, so maybe not erasing, but not writing properly.

If you released pressure on the card adjustment it may make the card write weaker because the head is not in proper card contact. Try changing it back or increasing the pressure a bit and see what happens.

Make sure WE stays at a LOW level all the time when reading.

cheers

Tony

[albertofenini]

Thanks Tony
I’ll surely check it
Still, it fails me how is it possible that just reading a card can damaged it
so that any other unit, including the one that has originally recorded it, will
not be able to read it again …
I’ll check it better tomorrow and let you know

When you says “make sure WE stays low” you mean it has to be ground connected right?
And when it should go low and how long it should stay low ?

Thanks !!!

[teenix]

(11-21-2021 12:43 AM)albertofenini Wrote:  Thanks Tony
I’ll surely check it
Still, it fails me how is it possible that just reading a card can damaged it
so that any other unit, including the one that has originally recorded it, will
not be able to read it again …
I’ll check it better tomorrow and let you know

When you says “make sure WE stays low” you mean it has to be ground connected right?
And when it should go low and how long it should stay low ?

Thanks !!!

WE stays Logic LO when reading and probably stays LO all the time when idle. If it goes Logic HI, the write circuits become active which is undesirable for any other mode than writing. If WE goes HI when reading, it will affect the stored card data.

cheers

Tony

[albertofenini]

Thanks again Tony
By low you mean connected to ground ?
By hi what do you mean?
Take care and thanks

[teenix]

(11-21-2021 01:07 AM)albertofenini Wrote:  Thanks again Tony
By low you mean connected to ground ?
By hi what do you mean?
Take care and thanks

Logic LO would be close to zero volts.
Logic HI would be somewhere between 4 and 6 volts. I can't remember exactly.

cheers

Tony

[albertofenini]

Hi Tony, Hi everyone

I did some further troubleshooting and this is what I discovered ...

WE stays LO all the time (when reading)

A known working CPU board moved in this chassis with the original card reader doesn't work
A known working CPU board AND a known working card reader moved in this chassis don't work either

This card reader in a known good chassis works with a good known CPU board
This card reader and this CPU board in a known good chassis both work fine
However, this CPU board in any of the other three faulty HP67s I have is not working except for one combination

It must be said that in this chassis the upper left and middle left screw holes (with keys face down)
are worn out, and this probably is preventing a good contact between the card reader switches and
the keyboard PCB, although the switches closes correctly when a card is read

I have reinforced these two holes by inserting a small wire before tightening the screws and it seems to
stay together well, however the calculator doesn't read nor writes, in a way it is worse than yesterday, but I think I may
have isolated the problem with the chassis

Also, after this tests and all this handling now it is not damaging cards where reading which is good

I have updated the photo album with a capture of RA and RB when reading, they look normal to me but
at the end of reading the "Error" message always shows up

One more thing I may try is to change the "screw" that goes thru the card reader wheel, when rotating that in
order to have more or less pressure on the card it does not move the wheel, so it may as well be that the
problem

Thanks for any advice you want to share, take care !!

Updated pictures here : HP-67 1705A01234

[teenix]

I learnt that the card switches can be finnicky if the chassis is distorted or if they are not adjusted properly, probably from bent contact fingers. These problems might upset the timing of the switches and confuses the CRC.

When the keyboard PCB is positioned in the front panel, it should sit flat on the mounts that are moulded into the front panel with no rocking. If it doesn't sit flat when the screws are tightened, the assembly will warp and put twisting forces on the switch contacts.

Sometimes the power wires can get in the way too.

cheers

Tony