Hello and asking for help with HP 67 power unit

[Tapti]

Hello alltogether,
I'm new to this forum. The reason to join you is that I'm a proud owner of one HP-32e and meanwhile two HP 67. One of the latter having a problem most probable with the power unit. I'm physicist by profession, so electronic circuit are not completely meaningless to me. But that's the issue with pysicists: They believe they know everything. Actually they know something of the most, but nothing very well ;-)
The problem with my 67 is the following: Overall it is in an exceptional good shape and it would be a pitty not to get him working. When the maschine is switched on (independently from the actual power supply) the reading is quite normal and it is possible to perform some calculation. After say 3 to 5 min the display starts showing nonsense (first only wrong numbers), then it starts decaying and in the end the only thing visible is the decimal point. After switching it off again and letting him alone for some time, the game starts from the very beginning. So my suspicion is that some part within the power unit (switching ps for creating the different needed voltages) is old or broken resp. Did somebody of you already observed a behaviour like that ? These are only some parts to exchange (2 transistors, some diodes etc), but I don't know the actual spare part numbers of these. Otherwise I would just rebuild the psu.
Thanks for your thoughts,
Stefan

[teenix]

Welcome to the Forum :-)

Does it work normally with the power supply attached, it might be a faulty battery? If you have another working 67, try the battery from that unit.

It could be the power supply or (hopefully not) a chip is failing.

If you have the back removed, and when operating, do you feel anything getting warm or hot? That might narrow down the problem. Look for corrosion on the circuit board.

cheers

Tony

[Tapti]

Hello Tony,
thank for your fast reply. For the test I used a stabilized lab power supply connected to the battery terminals, not the original power supply. In the meanwhile I do own one. But actually I got a problem with the german parcel service (DHL). I sent the unit to somebody, who adrvertised being a specialist for the repair of HP calculators. Unfortunetly at the time when the parcel arrived at him he obviously wasn't at home. So the parcel was delivered to the local post office. The guy refused to go there an fetch the parcel. So for the moment I have a running application at DHL searching for the fate of the parcel. Hopefully I hear something from them in the next few days. I will check your proposals as soon as I got my HP. back.
Greetings
Stefan

[Tapti]

Hello Tony and all others,
my calculator found the way back to me without any further damage. Yesterday I tested it with the original wall plug. Unfortunatly the same thing is happening after some minutes, begining with faulty calculations and ending up with a dead display. What I can say in addition, it doesn't seem to be a temperature problem or anything else hinting towards an electric short. My impression is rather it might be connected to some charge accumulation somewhere, because the cure is just to switch it of for some more minutes. And so letting the charge to decay. The point is, I already took the piece apart for repairing the card reader but I never put hands on the electronic componenets as on the power supply or even on the logic electronics. I'm asking whether there may be someone under you guys, who perhaps could try to find the problem and evenrually repaire it. Perhaps somebody from Germany ?
Thanks a lot,
Stefan

[teenix]

Has the unit been operated from the original power supply and without batteries. If so, this may cause damage to the CPU board as mentioned on page 312 of the owners manual.

I could have taken a look at the CPU but am way down here in the land of Oz.

cheers

Tony

[Tapti]

Tony,
I made some further experimenting today. In the meanwhile the symptoms have changed to some extend !?! When I connect either the wall plug or an external voltage source to the battery terminals it still takes some 3-5 minutes until the display starts to show erratic values. BUT: It will not anymore fade away, as it was during my first testing. Perhaps what I saw were different problems, one of them maybe healing up with time (reformating electrolytic ?) and the other one beeing a problem of the logic board, I suspect. Reason for that is that it is possible to produce some 'Error'-Message. So I think it has nothing to do with the LED driver chips, or am I wrong ?
As I already told you I have another running HP67 at home. Tomorrow I will exchange the two logic boards. This might help to localize the faulty area, I hope.
Greetings from Bonn, Germany
Stefan
PS: I know the problem of running the maschine from the wall plug only. So I never did. Question of coarse is, whether somebody else did that already. BTW I seriously doubt that doing so will hurt the maschine as long as one keeps the card reader off the game.

[ThomasF]

(11-23-2021 05:28 PM)Tapti Wrote:  BTW I seriously doubt that doing so will hurt the maschine as long as one keeps the card reader off the game.

Hi Stefan!

I know otherwise! I got a fully functional 67 some 2 years ago. I was running it with a battery and a wall-charger connected to do some timing test. When I checked the device some hour later it was gone ... All symptoms pointed to a burned ACT ...

After some fault searching I blame the battery. It was a homemade one (not mine ) with bad connection - so it was only running on the charger.

I bought a new 67-board from Tony - with all bells and jingles - so now I have a new 67-extended with PC-connection etc!
Panamatik in Germany also have similar 67 replacement boards.

Hope it all turns out well for your 67!

Best regards,
Thomas

[Tapti]

Hi Thomas,
thanks for telling me your experience. Which are the symptoms pointing towards a burned ACT ? Perhaps this is also the case with my 67.
Greetings
Stefan

[teenix]

(11-23-2021 05:28 PM)Tapti Wrote:  PS: I know the problem of running the maschine from the wall plug only. So I never did. Question of coarse is, whether somebody else did that already. BTW I seriously doubt that doing so will hurt the maschine as long as one keeps the card reader off the game.

Its more likely that the ACT chip will go faulty. This is the main CPU of the calculator, so if it fails, the rest won't work. If it partially fails, the calculator may work for a bit, then fail, then work for a bit etc etc.

cheers

Tony

[ThomasF]

(11-23-2021 06:16 PM)Tapti Wrote:  Hi Thomas,
thanks for telling me your experience. Which are the symptoms pointing towards a burned ACT ? Perhaps this is also the case with my 67.
Greetings
Stefan

Hi Stefan,

In my case, only some single digits lighted up in the display (quite bright), typical to the right of the display, no response from the keyboard at all, not always the same character.
E.g. like this ( were '.' is a empty digit):

Code:

..........00...

Mostly the same result at every power up, not always the same digits, sometime no digit at all, but it never worked again (even for short times after power on) until I replaced the board.

So not really the same problem as in your case (if not the problem is temperature related). If lucky it could be a bad solder, so maybe a visual inspection of the board could be a good thing to do.

Cheers,
Thomas

[teenix]

(11-24-2021 08:39 AM)ThomasF Wrote:  In my case, only some single digits lighted up in the display (quite bright), typical to the right of the display, no response from the keyboard at all, not always the same character.
Cheers,
Thomas

If that happens, the processor is probably frozen and the multiplexed LEDs are not being updated properly. In this case they could be getting the full available current flowing through the lit segments which is not good for them. Bets not to turn it on until repaired.

cheers

Tony

[Tapti]

This morning, I did like announced and exchanged the logic boards for each other. The outcome is, that now the maschine is running without problems and the other (previously running) now has got the problems. Ok, that means the logic board is defective. What can I do ? I inspected the board as good as I can for bad solder joints or any corrosion, but I found nothing so far. My feeling is that one of the chips is not working properly over time. Is there any chance to exchange them ? Are there spare parts available ? Concerning soldering the chips I might try asking one of my collegues from the university. Another possibility would be the logic boards of panamatik and / or teenix as far as I know (Thank you Bas !). Just from what I learned from the web pages, I would prefer the teenix option, because their board cooperates with the card reader, right ? Has anybody some experience with these modern boards ? Do I have another option for a repair ?
Thank you guys,
Stefan

[Tapti]

Ok, what I forgot is maybe first measuring the different supply voltages. There might be the chance of a defective non-integrated part like a transistor or so. What do you think ?

[teenix]

There is always the chance of a defective passive component like a capacitor or resistor especially since they are approaching 50 years old.

There are 3 capacitors in the power supply which may cause problems if they are failing, but are cheap to replace.

If a chip has gone, your only option is a replacement board or replacement part if one is available.

The only real way to explore the problems is with an oscilloscope to see what signals are circulating around the board.

cheers

Tony

[ThomasF]

(11-24-2021 12:24 PM)Tapti Wrote:  Has anybody some experience with these modern boards ? Do I have another option for a repair ?
Thank you guys,
Stefan

Hi Stefan,

I replaced the faulty board with one from Teenix, and I'm very pleased with it.
Since then, I think he has improved the board, so it now even have continues memory, i.e. you now will have a "HP67CX" (apart from every thing else - card reader support, flash storage for programs, connectivity etc)! I can recommend getting a board from him!

Panamatik 67 board also works fine (I have one of these as well), but it will only extend with flash storage (i.e. no card reader support, connectivity etc).

Cheers,
Thomas

[GreyUser]

The OP has a 67, not a 20 series Woodstock. There is no risk of ACT damage when operating without a battery. While there is a small risk, it is not the ACT that is at risk. This discussion is well documented in the old forum. It is not relevant to the OP’s situation so I will stick to the problem at hand.

The flashing display, while it could be an ACT failure, it could just as likely be mains ripple caused by a defective 400uf capacitor in the power supply, a common failure in the 82002 charger.

You can test the calculator with a bench power supply, connecting to the calculator’s battery terminals. 4 to 4.5 volts dc at 500 milliamps is required. The battery charger must be disconnected from the calculator when performing this test.

[Kostas Kritsilas]

Please refer to psoting #3; this has already been tried.

[Tapti]

Dear Experts,
thanks a lot for your help so far. Today I made measurements of the two supply voltages Vss and Vgg, which should be around 6.25V and -12V, respectively. With the damaged board I measured 6.2V and -10.1V. While Vss should be fine, Vgg is some 20% to low. I'm not shure that this somewhat lower voltage may be a problem. But when I compare the values with those taken from my second (good) board, giving 6.35V and -12.1 Volts, there is in fact a significant difference. When I'll get some time tomorrow, I will repeat the measurements with an oszilloscope to look for the amount of ripple on these voltages. For the moment I'm still hoping that repairing the power converter might solve the problem. On the other hand of course it is also possible that particularly Vgg has to deliver to much current. Perhaps do to some short within the electronics being supplied with Vgg.
I'll keep you informed,
Stefan

[AndiGer]

Hi Stefan,
See the parallel thread on HP-97 tantalum caps ...
Corresponding to HP-97 C6 in the HP-67 could be the brown one above the torroid. 47µF. It is the same colour as in the (newer) HP-97. Also found some very colourful 47 µF caps. orange - yellow - pink - green from top to bottom.

Alberto has posted links to some pictures here

[Tapti]

Hello Alltogether,
@Andi: Many thanks for the hint concerning the tanalum capacitors. In fact at least these two have to be checked.
This morning I tried to identify the capacitors on the power converter part of the logic board. Comparing the actual wiring with the schematics hand drawn by J. Laporte

http://home.citycable.ch/pierrefleur/Jac...20unit.htm

(Fig. 7) it's pretty clear which one is which. But I didn't manage to decode the color scheme on the tantalums. They seem not to match with the values from Laporte, at least if I'm taking the actual color code one finds in the web for tantalums.
Perhaps one has to use an older code (american code) ? Somebody of you know ? I really would like determine the capacitor values by my own either. Just to be shure, there is no mistake in the linked drawing.
Thanks and a happy weekend !
Stefan