HP-65 card reader - purpose of WE pin

[albertofenini]

Dear all,
we are in the process of restoring an HP-65 thanks to the contribution of one of the members
that has provided a logic board and a cathode and anode driver.

So far so good, we have also began to use the TAMIYA matt finish spray ad the results are very
encouraging as the unit really recover that "like new" finish appearance.

However we have trouble, as often happens, with the card reader.
Waveforms RA and RB are good and the unit reads all of the cards, including those written by other HP65
Waveforms WA and WB are also good and clean, but the unit does not record.

We will perform soon the switch test, checking that the WPS switch closes correctly when the
card is inserted.

However, what is the purpose of the WE pin in the card reader PCB ?
Has it something to do with Write Enable (??) and is a possible cause of malfunctioning ?
How can we check that ?

It also must be said that the logic board is new, the card reader PCB is also new
(I mean is none of the spares we had before that were also not working) but the keyboard
plastic assembly, including the card reader switches on the back and the keyboard PCB are a spare part we had before, and
that we were not be able to use it until now because it was lacking the anode and cathode driver
that we transplanted now from the donated one (that was too corroded to be used)

Any idea or hint ?
thanks for help very much !!

[teenix]

(05-02-2019 04:03 PM)albertofenini Wrote:  Dear all,

We will perform soon the switch test, checking that the WPS switch closes correctly when the
card is inserted.

However, what is the purpose of the WE pin in the card reader PCB ?
Has it something to do with Write Enable (??) and is a possible cause of malfunctioning ?
How can we check that ?

If the WPS switch is faulty (open circuit), you should get "Error" displayed when trying to write a card, because the CRC will interpret that as a write protected card. For a short circuit, the card should write ok, even for write protected cards.

In normal operation, if WPS is HI when HDS goes LO, the card is write protected.

The WE pin is controlled by the CRC chip. It goes Logic HI when the card chip is to write a card and Logic LO when it is to read a card.

cheers

Tony

[albertofenini]

Hi Tony
This is an HP65 so you don’t get any error message
With the switch in W/PRGM the card goes thru
The waves WA and WB look clean and ok and the display is steady like everything was ok
But if you try to read the card just written you get a blinking display
I’ll check the switches and report here

We have another unit that has a noisy WB and doesn’t write as described in a
different thread but this is the first time that with good waves the card is not written

We have of course tried a bunch if new cards and not just one

Any suggestion is welcome ! Thx in advance

[teenix]

(05-02-2019 11:12 PM)albertofenini Wrote:  Hi Tony
This is an HP65 so you don’t get any error message
With the switch in W/PRGM the card goes thru
The waves WA and WB look clean and ok and the display is steady like everything was ok
But if you try to read the card just written you get a blinking display
I’ll check the switches and report here

We have another unit that has a noisy WB and doesn’t write as described in a
different thread but this is the first time that with good waves the card is not written

We have of course tried a bunch if new cards and not just one

Any suggestion is welcome ! Thx in advance

The switches in the HP65 look like they do the same functions as the HP67/97. The reader chip has a different serial number, but look essentially the same and I would imagine function the same although they may have been improved with later models. The HP65's main difference is in how the data is transferred to and from the calculator's memory. That process is directly controlled by the CRC where as the 67 is by software control.

If the wave forms are looking ok, then assuming the PSU, CPU, motor speed, caps, connections and cards are ok and the RW head is clean, then that just leaves the sense chip and RW head itself as the probable weak link. They could be faulty due to age and may only ever work "some" of the time or not at all.

The data going into the sense chip on the WA WB lines won't necessarily reflect the data coming out of the RW head onto the card surface.

The head, through years of use, could have a flattening of the usually rounded surface which the card rubs over during record/playback. The gap could probably be taken up by the cam adjustment, but if the head is really worn, this adjustment might not suffice. See attached. This calculator has read errors and I can clearly see data bits missing on the RA RB pins. The card reads ok on a different 67.

The writing probably has enough energy from the worn heads to change the magnetic state of the card surface though a mildly worn gap, but would be less effective.

Reading that data back I guess would be problematic in these circumstances as a worn or a good head tries to read the weakly recorded data back.

The actual head itself might be "weak" especially if it has become magnetized.

I don't think the 65 detects card errors other than write protect and the RUN/PRGM switch in the wrong position as there is no checksum saved to the card.



Edit: My bad - I don't even think the RUN/PRGM switch is tested either, wouldn't make sense :-)

cheers

Tony

[albertofenini]

Hi Tony
We will check the switches over the weekend and see
We will report results of the test here
Thank you very much for sharing your ideas

[teenix]

(05-03-2019 05:57 AM)albertofenini Wrote:  Hi Tony
We will check the switches over the weekend and see
We will report results of the test here
Thank you very much for sharing your ideas

No problems Alberto,

I think lately I share some of your frustration, or perhaps enlightenment into the card reader mysteries :-)

cheers

Tony

[albertofenini]

Quick question...
Would it be worth measure the resistance of the heads ?
Just to make sure the magnetic head is ok ?
And if yes does anyone knows which values should we expect and between which cables should it taken the values ?
Looking at the schematics I would measure between the yellow and red and between the orange and blu as these appears to me the two inductances
Or can we assume that if the cards are read well than the head is ok and the problem is elsewhere?
Thx for help !!

[albertofenini]

Guys we fixed it !
We tested the switches and the WPS switch was not closing at all ...
We had to work a little has we raised it too much and then it was always closed
However we got the right position and now it read and write cards like new
To be honest we think that the clutch should also be reinforced so will take care of that sooner or later
Will post some pictures tomorrow !!!
We want to thanks everyone for the technical and moral support and encouragement !!!

[teenix]

(05-03-2019 09:47 PM)albertofenini Wrote:  Guys we fixed it !
We tested the switches and the WPS switch was not closing at all ...
We had to work a little has we raised it too much and then it was always closed
However we got the right position and now it read and write cards like new

From the 97 service manual, the procedure is to turn the screw until the contact just closes, then back the screw off 1/4 turn.

Even though I didn't physically test my previous hypothesis, your repair proved what I mentioned about the WPS being open circuit causes a write error. In your case the CRC would have thought the card was write protected, aborted the write and gave the correct error indication.

It is interesting that the data is still fed to the card reader as it passes through. I also noticed on a 67, that during a read, the WA WB pins have activity as well, but it doesn't exactly match the read data. I haven't checked, but maybe as card data is being shifted in to the holding buffers in the CRC, it is also being shifted out on the WA WB pins but delayed because of the time it takes to initially fill the buffer. As the sense chip is in read mode, the WA WB pins would be ignored.

Well done Alberto

cheers

Tony

[albertofenini]

Hi Tony
thank you very much for your support, we haven't been able to post pictures will do it soon,
however, in HP97 is easier because the screws that tighten or loosen the switches are accessible
with the card reader in place, while in the HP65 you need to dismantle everything, tune the screws,
reassemble the unit and check if it is ok, a little tricky but can be done.
We are now facing a new challenge with a unit (the last one by the way) that as a WB channel noisy
and does not record cards.
Any idea is welcome, take care !