Another HP-67 repair query

[coitboy2000]

Hi all

I am trying to repair an HP-67 I purchased as not working. I have searched the forum and compiled all the info and and schematics etc but thought I would ask the brains trust a few initial questions based on my probing of the calculator so far. It suffered from a battery leak and the terminals essentially dissolved. There wasnt much evidence the leak spread and I have given it a good clean. My observations initially are

- when connecting the battery terminals to by bench powers supply I notice the red led dot illluminates when I supply less than 3.6v. It draws about 90ma current.
- increasing the supply to 4v max the red light dissapears but the display is blank. The current draw increases to 110ma
- Initial probing indicates voltage is transferring to the cpu board but only 4.5v or so max.
- the sense amp chip gets pretty hot to touch. Normal ?

Any suggestions on what to start probing first would be greatly appreciated. I guess it may be a dead sense amp or power supply but I am unsure how to establish this for sure.

Cheers

Richard

[teenix]

(10-14-2020 05:40 AM)coitboy2000 Wrote:  Hi all

I am trying to repair an HP-67 I purchased as not working.

- when connecting the battery terminals to by bench powers supply I notice the red led dot illluminates when I supply less than 3.6v. It draws about 90ma current.
- increasing the supply to 4v max the red light dissapears but the display is blank. The current draw increases to 110ma
- Initial probing indicates voltage is transferring to the cpu board but only 4.5v or so max.
- the sense amp chip gets pretty hot to touch. Normal ?

Cheers

Richard

The low battery indicator appears to be working.

The sense amp definitely should not get hot. Try carefully unplugging the sense board from the CPU board and try the power again.

cheers

Tony

[coitboy2000]

(10-14-2020 06:13 AM)teenix Wrote:  

(10-14-2020 05:40 AM)coitboy2000 Wrote:  Hi all

I am trying to repair an HP-67 I purchased as not working.

- when connecting the battery terminals to by bench powers supply I notice the red led dot illluminates when I supply less than 3.6v. It draws about 90ma current.
- increasing the supply to 4v max the red light dissapears but the display is blank. The current draw increases to 110ma
- Initial probing indicates voltage is transferring to the cpu board but only 4.5v or so max.
- the sense amp chip gets pretty hot to touch. Normal ?

Cheers

Richard

The low battery indicator appears to be working.

The sense amp definitely should not get hot. Try carefully unplugging the sense board from the CPU board and try the power again.

cheers

Tony

Ok, I will try that and report back. Thanks.

[coitboy2000]

(10-14-2020 07:08 AM)coitboy2000 Wrote:  

(10-14-2020 06:13 AM)teenix Wrote:  The low battery indicator appears to be working.

The sense amp definitely should not get hot. Try carefully unplugging the sense board from the CPU board and try the power again.

cheers

Tony

Ok, I will try that and report back. Thanks.

I decided to give it another good clean, in particular the processor board board. It is behaving as before but the sense amplifier does not seem to get as hot.

There is an adjustable potentiometer (10k) on the card reader board that currently reads around 1.4k, at least in circuit anyway. What should this be set to ? The 3x capacitors on the card reader board still seem ok.

Regards

Richard

[teenix]

(10-14-2020 10:11 AM)coitboy2000 Wrote:  

(10-14-2020 07:08 AM)coitboy2000 Wrote:  Ok, I will try that and report back. Thanks.

I decided to give it another good clean, in particular the processor board board. It is behaving as before but the sense amplifier does not seem to get as hot.

There is an adjustable potentiometer (10k) on the card reader board that currently reads around 1.4k, at least in circuit anyway. What should this be set to ? The 3x capacitors on the card reader board still seem ok.

Regards

Richard

If the sense chip is still getting warm, it suggests a problem remains. If it was allowed to get "hot" then it may be damaged, but hopefully not.

Switch mode power supply output should be around -12V and +6.25V.

The resistor sets the motor speed. It can be any value within the pots range to set the correct speed, although I doubt it needs to be full scale one way or another. I wouldn't touch it until you know the card reader is working.

Setting the correct speed practically speaking would be to use an oscilloscope to measure the data stream bits sent to the CPU board and they should be around 1mS Hi to Lo or Lo to Hi, or you can see the entire data stream fits on the card just after being sensed in and before sensed out. See Page 7, Fig 8 of HP Journal 1976-11.

cheers

Tony

[coitboy2000]

(10-14-2020 10:40 AM)teenix Wrote:  

(10-14-2020 10:11 AM)coitboy2000 Wrote:  I decided to give it another good clean, in particular the processor board board. It is behaving as before but the sense amplifier does not seem to get as hot.

There is an adjustable potentiometer (10k) on the card reader board that currently reads around 1.4k, at least in circuit anyway. What should this be set to ? The 3x capacitors on the card reader board still seem ok.

Regards

Richard

If the sense chip is still getting warm, it suggests a problem remains. If it was allowed to get "hot" then it may be damaged, but hopefully not.

Switch mode power supply output should be around -12V and +6.25V.

The resistor sets the motor speed. It can be any value within the pots range to set the correct speed, although I doubt it needs to be full scale one way or another. I wouldn't touch it until you know the card reader is working.

Setting the correct speed practically speaking would be to use an oscilloscope to measure the data stream bits sent to the CPU board and they should be around 1mS Hi to Lo or Lo to Hi, or you can see the entire data stream fits on the card just after being sensed in and before sensed out. See Page 7, Fig 8 of HP Journal 1976-11.

cheers

Tony

Thanks Tony

With just 4V supply should I be seeing -12V somewhere on the cpu board ? The highest voltage I see is -11.33V on pin 2 of 1820-1596 and 6.3V on the majority of pins on 1820-1751 ?

[teenix]

(10-14-2020 11:24 AM)coitboy2000 Wrote:  Thanks Tony

With just 4V supply should I be seeing -12V somewhere on the cpu board ? The highest voltage I see is -11.33V on pin 2 of 1820-1596 and 6.3V on the majority of pins on 1820-1751 ?

1820-1751 = CRC chip
pin 28 is GND
pin 15 is Vgg = -12V
pin 2 is Vss = +6.25V

The circuit diagram that I have has the 1820-2530 ACT chip which I believe the 1820-1596 replaced. If they are the same then -11.33V would seem ok for pin 2, and about +6.25V for pin 1. Pin 12 is GND.

cheers

Tony

[coitboy2000]

(10-14-2020 11:42 AM)teenix Wrote:  

(10-14-2020 11:24 AM)coitboy2000 Wrote:  Thanks Tony

With just 4V supply should I be seeing -12V somewhere on the cpu board ? The highest voltage I see is -11.33V on pin 2 of 1820-1596 and 6.3V on the majority of pins on 1820-1751 ?

1820-1751 = CRC chip
pin 28 is GND
pin 15 is Vgg = -12V
pin 2 is Vss = +6.25V

The circuit diagram that I have has the 1820-2530 ACT chip which I believe the 1820-1596 replaced. If they are the same then -11.33V would seem ok for pin 2, and about +6.25V for pin 1. Pin 12 is GND.

cheers

Tony

OK, thanks I will keep probing. Any thoughts on why the display is still blank Is it likely any of the components may be the reason ? i.e are capacitors, resistors, transistors ever the reason ? If only there was a whole suite of published base voltages throughout the circuit for beginner electronics people like me. Its hard to hunt down faults.

I appreciate your replies.

[teenix]

(10-14-2020 11:58 AM)coitboy2000 Wrote:  OK, thanks I will keep probing. Any thoughts on why the display is still blank Is it likely any of the components may be the reason ? i.e are capacitors, resistors, transistors ever the reason ? If only there was a whole suite of published base voltages throughout the circuit for beginner electronics people like me. Its hard to hunt down faults.

I appreciate your replies.

It could be multiple things unfortunately. ROM ACT etc. It would be hard to tell without checking with a scope on how the data busses and display drivers are operating.

I'd still try unplugging the card reader board from the CPU board and make sure it is not shorting anything and try the power again.

cheers

Tony

[coitboy2000]

(10-14-2020 12:06 PM)teenix Wrote:  

(10-14-2020 11:58 AM)coitboy2000 Wrote:  OK, thanks I will keep probing. Any thoughts on why the display is still blank Is it likely any of the components may be the reason ? i.e are capacitors, resistors, transistors ever the reason ? If only there was a whole suite of published base voltages throughout the circuit for beginner electronics people like me. Its hard to hunt down faults.

I appreciate your replies.

It could be multiple things unfortunately. ROM ACT etc. It would be hard to tell without checking with a scope on how the data busses and display drivers are operating.

I'd still try unplugging the card reader board from the CPU board and make sure it is not shorting anything and try the power again.

cheers

Tony

Unplugging the card reader had not affect. Is there any way to check some of the ICs ? I have a 4 channel scope.

Richard

[teenix]

(10-14-2020 09:33 PM)coitboy2000 Wrote:  Unplugging the card reader had not affect. Is there any way to check some of the ICs ? I have a 4 channel scope.

Richard

I've put some info here for classics but it may prove useful near end of document. It won't be too hard to check display drivers if you look at the circuit diagram.

http://www.teenix.org/ClassicNotes.pdf

HP-67 circuit

http://www.hpcc.org/cdroms/schematics5.0...c/hp67.pdf

cheers

Tony

[coitboy2000]

(10-15-2020 12:48 AM)teenix Wrote:  

(10-14-2020 09:33 PM)coitboy2000 Wrote:  Unplugging the card reader had not affect. Is there any way to check some of the ICs ? I have a 4 channel scope.

Richard

I've put some info here for classics but it may prove useful near end of document. It won't be too hard to check display drivers if you look at the circuit diagram.

http://www.teenix.org/ClassicNotes.pdf

HP-67 circuit

http://www.hpcc.org/cdroms/schematics5.0...c/hp67.pdf

cheers

Tony

Wow, thanks Tony. It would be nice to save this calculator

Richard

[coitboy2000]

(10-15-2020 01:26 AM)coitboy2000 Wrote:  

(10-15-2020 12:48 AM)teenix Wrote:  I've put some info here for classics but it may prove useful near end of document. It won't be too hard to check display drivers if you look at the circuit diagram.

http://www.teenix.org/ClassicNotes.pdf

HP-67 circuit

http://www.hpcc.org/cdroms/schematics5.0...c/hp67.pdf

cheers

Tony

Wow, thanks Tony. It would be nice to save this calculator

Richard

I am out of depth on this I think. I have captured some scope images for the cathode, anode & led pins when the calulator is powered up, If I twist the the keyboard near the connector with the cpu board I get some random elements light up on some of the leds'.

https://www.dropbox.com/sh/3kvxiilqx1jd3...0_nLa?dl=0

[teenix]

(10-18-2020 10:10 AM)coitboy2000 Wrote:  I am out of depth on this I think. I have captured some scope images for the cathode, anode & led pins when the calulator is powered up, If I twist the the keyboard near the connector with the cpu board I get some random elements light up on some of the leds'.

I would expect fairly uniform waveforms being displayed on the scope. Some seem to be changing amplitude a bit which may be a symptom of the pin connections you mention.

Figure 4.7 on Pg 4-12 of the HP-97 service manual shows proper STB and RCD waveforms.

http://www.decadecounter.com/vta/pdf/HP-...Manual.pdf

Can you notice changes in the partially working display at all if any number keys are pressed? If so then maybe the CPU components are working.

Maybe the connector pins and/or holes are corroded or not quite making proper contact. Deoxit spray to clean things up??

If you suspect some are a loose fit in the holes, then I did manage to successfully widen some a bit. A word of caution though, these pins are brittle so it is not really a good idea to try and bend them, they can snap quite easily.


With the CPU board removed, and observing static handling procedures, I gently inserted a metal scriber (which has a long tapered shaft and a pointed end) into the gap between the two prongs just enough to expand them "a little bit". Once the scriber is withdrawn, the prongs spring back towards each other, but they may have opened out a bit which provides some more tension for better contact in the hole. This method fixed a similar problem, but if the pins have lost most of their springiness by being inserted into the circuit board holes over the years, then this fix probably won't last long.


cheers

Tony

[coitboy2000]

(10-18-2020 11:08 AM)teenix Wrote:  

(10-18-2020 10:10 AM)coitboy2000 Wrote:  I am out of depth on this I think. I have captured some scope images for the cathode, anode & led pins when the calulator is powered up, If I twist the the keyboard near the connector with the cpu board I get some random elements light up on some of the leds'.

I would expect fairly uniform waveforms being displayed on the scope. Some seem to be changing amplitude a bit which may be a symptom of the pin connections you mention.

Figure 4.7 on Pg 4-12 of the HP-97 service manual shows proper STB and RCD waveforms.

http://www.decadecounter.com/vta/pdf/HP-...Manual.pdf

Can you notice changes in the partially working display at all if any number keys are pressed? If so then maybe the CPU components are working.

Maybe the connector pins and/or holes are corroded or not quite making proper contact. Deoxit spray to clean things up??

If you suspect some are a loose fit in the holes, then I did manage to successfully widen some a bit. A word of caution though, these pins are brittle so it is not really a good idea to try and bend them, they can snap quite easily.


With the CPU board removed, and observing static handling procedures, I gently inserted a metal scriber (which has a long tapered shaft and a pointed end) into the gap between the two prongs just enough to expand them "a little bit". Once the scriber is withdrawn, the prongs spring back towards each other, but they may have opened out a bit which provides some more tension for better contact in the hole. This method fixed a similar problem, but if the pins have lost most of their springiness by being inserted into the circuit board holes over the years, then this fix probably won't last long.


cheers

Tony

Still trying to get the calculator going.

I did find that the transistors that supply the display current through pins 3 & 4 of J1 did not appear to be working correctly. My calculator actually had an aluminium capsule that contained both transistors.

I am a little confused about the type of transistor to replace it with as this schematic:

https://www.hpmuseum.org/forum/thread-83...+schematic

shows them as NPN 2N3904 but the schematic by Tony Duell (I think) has them shown as PNP, at least the symbol anyway ?

[teenix]

(10-23-2020 11:22 PM)coitboy2000 Wrote:  Still trying to get the calculator going.

I did find that the transistors that supply the display current through pins 3 & 4 of J1 did not appear to be working correctly. My calculator actually had an aluminium capsule that contained both transistors.

I am a little confused about the type of transistor to replace it with as this schematic:

https://www.hpmuseum.org/forum/thread-83...+schematic

shows them as NPN 2N3904 but the schematic by Tony Duell (I think) has them shown as PNP, at least the symbol anyway ?

As far as I know, they should be PNP. My HP-67 replacement CPU board uses PNP as well.

Cheers

Tony

[coitboy2000]

(10-24-2020 04:14 AM)teenix Wrote:  

(10-23-2020 11:22 PM)coitboy2000 Wrote:  Still trying to get the calculator going.

I did find that the transistors that supply the display current through pins 3 & 4 of J1 did not appear to be working correctly. My calculator actually had an aluminium capsule that contained both transistors.

I am a little confused about the type of transistor to replace it with as this schematic:

https://www.hpmuseum.org/forum/thread-83...+schematic

shows them as NPN 2N3904 but the schematic by Tony Duell (I think) has them shown as PNP, at least the symbol anyway ?

As far as I know, they should be PNP. My HP-67 replacement CPU board uses PNP as well.

Cheers

Tony

Thanks

Probably what the 2n2222a made no difference. Will a 2N3906 be ok?

Cheers

Richard

[teenix]

(10-24-2020 11:26 AM)coitboy2000 Wrote:  Probably what the 2n2222a made no difference. Will a 2N3906 be ok?

Cheers

Richard

You can only try, the device I used has similar characteristics.

cheers

Tony

[coitboy2000]

(10-24-2020 12:28 PM)teenix Wrote:  

(10-24-2020 11:26 AM)coitboy2000 Wrote:  Probably what the 2n2222a made no difference. Will a 2N3906 be ok?

Cheers

Richard

You can only try, the device I used has similar characteristics.

cheers

Tony

What is the compatibility like between the HP-45 and HP-67 ? I have a HP-45 with a working display. Anode, Cathode and LED's ? Different part numbers I notice. I have re-soldered pretty much everything on the calculator and I still have no display. The RCD signal seems fine. Is there a way to check for shorts on the keyboard ?

I may purchase a new board to try in the meantime

Regards

Richard

[teenix]

(10-25-2020 10:14 PM)coitboy2000 Wrote:  What is the compatibility like between the HP-45 and HP-67 ? I have a HP-45 with a working display. Anode, Cathode and LED's ? Different part numbers I notice. I have re-soldered pretty much everything on the calculator and I still have no display. The RCD signal seems fine. Is there a way to check for shorts on the keyboard ?

I may purchase a new board to try in the meantime

Regards

Richard

Not compatible. The 45 display setup is quite different to the 67. It could also be ROM 0 which controls the display anodes.

With the CPU board removed, the rows should all be open circuit. The columns (not all) are connected to the cathode driver so may not be open circuit.

cheers

Tony