HP-65 1333A18516 - [SOLVED] request for advice

[albertofenini]

Hello everyone

we are trying to restore a unit acquired as a non working unit,
when arrived it looked in very good conditions, so we thought it was a easy one ...

The main problem was that "it didn't turn on" .....

When this happens the first test we do is to move the logic board in a different known to be working unit,
and see what happens, this test proved that the logic board was indeed working

The unit had been previously restored as we found the gummy wheel replaced,
although the coupler between the motor and the worm gear appears to be original

Few cables were desoldered and the three led modules had been replaced,
will post more details about that soon

Anode and Cathode drivers appeared to be original although all of that area
had been reworked

We began resoldering the card reader and the motor cables, we removed the
led modules, the anode and the cathode drivers, cleaned everywhere and resoldered nicely
everything back in place

However, when turned on it remained with a totally black display

With the card reader now connected we tried to read a card, and even if the result was
unkonwn because of the non working display, the card went thru smoothless

A small step forward ...

We tested the inductances that are placed above the leds and those are just fine,
we checked the capacitor, the inductance and the resistance that are placed just
below the switches and those are fine as well

We checked the capacitor in the upper corner, and that it's also fine

Eventually we temporarily replaced the original keyboard pcb, which was in very good condition, with
a "correded but working" one and the display turned magically on

We tried then to read the card again and that also worked fine, so we now know that the card reader is at least reading cards properly

We decided then to transplant the led modules, the anode and the cathode drivers
from the "corroded but working" keyboard pub to the original one, but nevertheless
the display remains off ....

We came to the conclusion that it has to be a faulty trace in the keyboard pcb
unless someone can point to a possible different cause

We also checked all the joint that connect the two faces of the pcb and these are all good with no cold or bad connections

Any idea how could we find the broken trace(s) ???

Should we start from the logic board connector or should we investigate more in
the led modules area ?

Keep in mind that all the components have been proven be in a working condition

Thanks for any help !!!
Take care

Pictures can be found here :
HP-65 1333A18516

[teenix]

If you can use your oscilloscope, you can check the signals going to the anode/cathode drivers.

Page 10 of my Classic Notes

If they are ok, then probably one of those IC's is faulty. If that area of the board has been reworked, maybe damage occurred. If you have spares, then swapping one IC at a time might reveal the problem.

cheers

Tony

[albertofenini]

(06-15-2024 01:00 AM)teenix Wrote:  If you can use your oscilloscope, you can check the signals going to the anode/cathode drivers.

Page 10 of my Classic Notes

If they are ok, then probably one of those IC's is faulty. If that area of the board has been reworked, maybe damage occurred. If you have spares, then swapping one IC at a time might reveal the problem.

cheers

Tony

Thanks for the advice Tony, we will certainly try but all the IC have already been replaced
with working ones from a donor, that's why we believe that the culprit has to be in the pcb,
problem is in which trace or in which area and how to find it

Where and what exactly should we measure ?

The reworked area has been cleaned and clearly inspected and looks ok so far ...

Thank you so much !

[teenix]

Pin 8 of the cathode driver is the Reset Pin (1820-1226)
Pin 9 is the STEP IN pin from the anode driver.

Pin 6 of the Anode driver is the STEP OUT pin to cathode driver. (1820-1029)

Both pulses are on page 12 of the Classic Notes.

I'd make sure these pulses are there, and if not check the traces back to the board.

The RCD pin from the CPU board is the left-most interconnect pin to the keyboard.

You might like to check the ABCDE lines to the anode driver as well, The signals image on page 10 of the Classic notes.

Other end of interconnector, right side pins counting left, 1 2 3 6 and 7. Pins 1 2 3 4 and 5 on anode driver.

cheers

Tony

[albertofenini]

(06-15-2024 05:35 AM)teenix Wrote:  Pin 8 of the cathode driver is the Reset Pin (1820-1226)
Pin 9 is the STEP IN pin from the anode driver.

Pin 6 of the Anode driver is the STEP OUT pin to cathode driver. (1820-1029)

Both pulses are on page 12 of the Classic Notes.

I'd make sure these pulses are there, and if not check the traces back to the board.

The RCD pin from the CPU board is the left-most interconnect pin to the keyboard.

You might like to check the ABCDE lines to the anode driver as well, The signals image on page 10 of the Classic notes.

Other end of interconnector, right side pins counting left, 1 2 3 6 and 7. Pins 1 2 3 4 and 5 on anode driver.

cheers

Tony

Hi Tony

before take any signal with the scope we checked the voltages
(please note that the card reader was not connected)

Vbatt is 3,8
Vbsw @ the logic board (pin 12) is 3,8

1820-1226 Cathode Driver
Vbsw @pin 4 : 0 V
Vcc @pin 10 : 1 V
Vbsw @ pin 15 : 0 V

1820-1029 Anode Driver
Vbsw @ pin 11 : 7 V

1810-0147
Vbsw : 3,8 V

1810-0146
Vbsw 3,8 V


We have also found three traces that looked like had been cut
but continuity has been verified (you can find them in the updated gallery)

However withe ICs and LDEs and Inductances mounted
is pretty hard to check more traces

Thanks for the support !!



Updated gallery :
HP-65 1333A18516

[albertofenini]

(06-15-2024 07:25 PM)albertofenini Wrote:  

(06-15-2024 05:35 AM)teenix Wrote:  Pin 8 of the cathode driver is the Reset Pin (1820-1226)
Pin 9 is the STEP IN pin from the anode driver.

Pin 6 of the Anode driver is the STEP OUT pin to cathode driver. (1820-1029)

Both pulses are on page 12 of the Classic Notes.

I'd make sure these pulses are there, and if not check the traces back to the board.

The RCD pin from the CPU board is the left-most interconnect pin to the keyboard.

You might like to check the ABCDE lines to the anode driver as well, The signals image on page 10 of the Classic notes.

Other end of interconnector, right side pins counting left, 1 2 3 6 and 7. Pins 1 2 3 4 and 5 on anode driver.

cheers

Tony

Hi Tony

before take any signal with the scope we checked the voltages
(please note that the card reader was not connected)

Vbatt is 3,8
Vbsw @ the logic board (pin 12) is 3,8

1820-1226 Cathode Driver
Vbsw @pin 4 : 0 V
Vcc @pin 10 : 1 V
Vbsw @ pin 15 : 0 V

1820-1029 Anode Driver
Vbsw @ pin 11 : 7 V

1810-0147
Vbsw : 3,8 V

1810-0146
Vbsw 3,8 V


We have also found three traces that looked like had been cut
but continuity has been verified (you can find them in the updated gallery)

However withe ICs and LDEs and Inductances mounted
is pretty hard to check more traces

Thanks for the support !!



Updated gallery :
HP-65 1333A18516

Tony forget the previous post as the measurements were probably mistaken

we have measured the voltages from both a working HP-65 (S/N 1508A04995) and
the HP-65 under repair S/N 1333A18516

We have measured with the scope all the suggested curves for both units and both
at the logic board interconnects and at the drivers, I'm posting two links to the images

Please, keep in mind that the logic board S/N 1333A18516 connected to the
keyboard with the drivers from the 150804995 works perfectly

Let us know what do you think
thanks again and take care !!

Curves from 1333A18516 can be seen here :CURVES FOR NOT WORKING S/N 1333A18516

Curves from 1508A04995 can be seen here :CURVES FOR THE WORKING S/N 1508A04995

[teenix]

Hi Alberto,

I'm not sure exactly what I am looking at, but obviously all signals are not present on the faulty unit, so I am wondering if the anode driver failed. That doesn't really explain why the signals ABCDE from the hybrid are not there though unless the anode driver is loading them down.

1V on VCC seems low for some reason, I suspect it would be higher on a working machine. I cannot test anything to clarify results, I was forced to move a few days ago and doing anything at the moment is rather difficult.

Does the anode driver get warm in use?

cheers

Tony

[albertofenini]

(06-17-2024 02:06 AM)teenix Wrote:  Hi Alberto,

I'm not sure exactly what I am looking at, but obviously all signals are not present on the faulty unit, so I am wondering if the anode driver failed. That doesn't really explain why the signals ABCDE from the hybrid are not there though unless the anode driver is loading them down.

1V on VCC seems low for some reason, I suspect it would be higher on a working machine. I cannot test anything to clarify results, I was forced to move a few days ago and doing anything at the moment is rather difficult.

Does the anode driver get warm in use?

cheers

Tony

Hi Tony, thanks for your reply,

no IC get warm when the calculator is turned on,
also don't look at the values we posted in the previous messages,
you need to look at the values we posted in the pictures galleries, are those named
Voltages and Tensioni (Italian for Voltages) the Vcc and the Vbsw are all there and correct in both units

But you are right, the faulty unit has not RCD, not Step in, not step out and no signals at DISP0 - DISP4

Our guess is that a shortened trace put everything down, let's hope this has not fried anything as
the Anode and Cathode drivers from the donor were perfectly working before the transplant

Also, we soldered the Anode driver and cathode Driver originally from this board (1333....) back into
the donor board and they where not working, but this doesn't prove anything

Will be away for a week or so, in the next steps we will remove the Anode and Cathode drivers
and we will inspect the traces underneath , also we will check if with the Anode and cathode drivers
removed the signals at the logic board will get any better

We'll keep you posted on this
Take care and thanks for the support

[albertofenini]

Hi Tony, hello everyone

we fixed it !

after your suggestions and the traces we had captured, we were about to investigate
Looking for shorted, broken or cut pcb traces

we made a sketch with the cathode and anode drivers pinout on a piece
of paper, by turning the page we would have been able to see what the pinout
was when the pcb is showing the traces on the opposte side of the leds

Thank to this we immediately found out that we had did something very stupid,
as the RCD connection, probably the easiest to track down, was pointing to the wrong chip

in fact, when replacing the old drivers, we had soldered the donors swapping the
cathode and the anode chips positions

in one or two of the previous pictures it could also have be seen, the 1226 should be
on the right side (looking at the led side) while it was not

once soldered the two in the right position it started to work like a charm

before to do so, we took some pictures of the back and front of the keyboard pcb
which we are posting under a separate post, it may be of use for someone to see the traces usually
hidden under the drivers without having to remove them

We obviously did the same on the donor machine, where also we had swapped the positions,
but that didn't worked out, presumably because those original drivers were really burned

However, and this is the question for the most expert people here, is there any way to tell
which of the two drivers between cathode and anode could be not working in the possible case
that only one is faulty ?

updated pictures can be seen here : HP-65 1333A18516 - [SOLVED] request for advice

[teenix]

(06-30-2024 07:11 PM)albertofenini Wrote:  Hi Tony, hello everyone

we fixed it !

. . . .

However, and this is the question for the most expert people here, is there any way to tell
which of the two drivers between cathode and anode could be not working in the possible case
that only one is faulty ?

Nice work Alberto.

The anode/cathode driver setup is quite complicated and requires accurate timing of the signals or damage to the cathode driver can occur due to mistimed coil discharges destroying the transistors inside. I found this out when developing the software for the 65 CPU replacement and thanks to help from Harald.

Separately, the cathode driver is essentially a ripple counter and needs a pulse on the step pin to select the next digit for the digit cathode which should go LO. A pulse on the RCD should reset it back to the first digit.

The anode driver will require sequences of pulses on the disp0 - disp4 pins.

There is some display driver info in the HP-45 patent document and more in my Classic Notes with scope traces etc.

cheers

Tony