HP-80 with bugs ... ?

[ThomasF]

Hi,

A while ago I received a HP-80 from Egypt. Frankly I can say that I have seen better preserved devices ...

When opened the device, some bug really fell out of the calculator, there were insects and eggs of different spices, but I managed to clean and restore it - and it springed back to life!

But, there is still some problem with it, maybe someone can give an hint of what might be going on ...

When I turn on the device, the display shows a random pattern, like the following (different each time I turn on the calculator):

Code:

3.3.3.1.6 .01.
0.0.0.00 0.0
0.0.0.0.0 44
- (just a single minus sign)
.3.3.3.4.1.  0
9.3.3.3.1.6 2 00
0.000
000.00.0.- #.#  (the dash is blinking and the # are a fast running number, like a stopwatch)

... and so on ...

Sometimes the last digit in the "mantissa" is "looping", ie. continously changing, but more often it is static.

The strange thing is that any keypress clears this state, i.e pressing CLx shows "0." and the calculator works as normal (or any other key - but the result is more unpredictable).

Sometime (more specific when a "running" digit is shown), the calculator is locked, and I have to turn it off and on again. And when on, it works like a charm until power off ...

So, something happens during startup, but only the display gets corrupt and the calculator works fine after clearing the display (and somtimes less frequently it stucks in a loop).

Any ideas what might cause this? Failing ROM (only at startup)? Bad capacitors which causes slow starting clock? ACT?

Cheers,
Thomas

[teenix]

It sounds like it is not resetting properly if it works properly after a keypress. It could be the reset hardware or a failing ROM. There is not much going on during initialization, registers are set to zero and just some formatting for the display take place. It is hard to tell without analyzing the code flow at start up.

The reset circuit is based on a couple of transistors, resistors and a capacitor, shown in the image. You would have to check the reset pulse using an oscilloscope or test these components to determine if this is the problem.

The C register may be corrupt at start up due to the reset problem and that is why the dodgy display. Pressing CLx would clear the C register and things may work properly, while other keys may try to process a function based on the invalid C data leading to other problems.

If it fails completely, Harald from the Forum may still have replacement CPU boards for the Classics.

cheers

Tony

[ThomasF]

(02-12-2024 10:21 PM)teenix Wrote:  It sounds like it is not resetting properly if it works properly after a keypress.
...
The reset circuit is based on a couple of transistors, resistors and a capacitor, shown in the image.
...
If it fails completely, Harald from the Forum may still have replacement CPU boards for the Classics.

Hi Tony!

Thanks for your insights! Even if I'm missing the image ...

Yes, I actually just bought a board from Harald in the case the old board should fail completely, but I still hope to be able to fix it - that is at least my goal now that I have a backup plan!

And you are right, after power on, the stack is full of NaN's and most operations (like RDN or X<>Y) results in a blinking "0." - guess that is the normal way of showing a NaN when detected by the HP80.
With a NaN in the display, I can also recognize the digits from the NaN when normalizing the number (ie after SHIFT 9 to show all decimals).

So, hopefully it is only something dodgy with the reset circuit, and I will try to find some description of - maybe this is part of you excellent Classic Notes?

Cheers,
Thomas

[teenix]

Sorry, I forgot to add the image.

Attached. I can't find download for the circuit that Tony Duell did for the HP-80 on hpcc.org, the link appears to be broken for some time. I don't have a circuit board overlay as I don't have a HP-80 to create one.

You could probably try briefly either shorting the 2.2uF capacitor or use a test lead from the base of the left transistor in the circuit diagram to the test point ground on the circuit board to initiate a reset.

cheers

Tony

[ThomasF]

(02-13-2024 07:56 AM)teenix Wrote:  Attached. I can't find download for the circuit that Tony Duell did for the HP-80 on hpcc.org, the link appears to be broken for some time. I don't have a circuit board overlay as I don't have a HP-80 to create one.

Thanks!

Yes, this looks more or less identical to the same circuit on the 35 and other classics.
I'll hook it up and try to verify the components and hopefully find what is wrong with my unit ...

Have also been looking for Duells schematics for the 80 but couldn't find it either ... someone have a working link?

Cheers,
Thomas

[AndiGer]

Yesterday I linked another thread in "HP-25 not working" where a working link is embedded :-D

https://hpmuseum.org/forum/thread-13740-...#pid121606

Andi

[teenix]

I tried this link a few times, but I keep getting a "cannot be downloaded securely" message.

Edit: I went direct to hpcc link and it worked - thanks :-)

cheers

Tony

[ThomasF]

(02-13-2024 09:30 AM)AndiGer Wrote:  Yesterday I linked another thread in "HP-25 not working" where a working link is embedded :-D

https://hpmuseum.org/forum/thread-13740-...#pid121606

Andi

Thanks!
Now I have them as well!

Cheers,
Thomas

[ThomasF]

(02-13-2024 07:56 AM)teenix Wrote:  You could probably try briefly either shorting the 2.2uF capacitor or use a test lead from the base of the left transistor in the circuit diagram to the test point ground on the circuit board to initiate a reset.

Hi Tony!

I followed your tip and measured around the reset/init circuit, and I saw that some signals were a bit "bad".
I replaced the 2.2uF cap as a first step, but that made no difference, so I continued.
Finally after some fault tracing (only had a multimeter at hands) I realized that one of the transistors must be bad, most probably the one connected to the Init signal.

Replaced both transistors, and voila! Another Classic brought back from a faulty state!

Also realized that the default state is DSP 2 ("0.00"). Whenever I turned it on before, it always ended up in DSP 0 state ("0."), so even if it seemed to work, probably some internal states was wacko.

I also found out that the picture you sent didn't correspond with the schematics.
The reset/init circuit is the one on the right hand side of the board (not left as you indicated - that was the PSU circuit), the init circuit is indicated on the included picture below.

So now I have to search for another Classic to repair so I can make use of the spare board I bought from Harald

But big thanks for your explanation that sent me in the right direction ..

Cheers,
Thomas

[teenix]

(02-13-2024 02:35 PM)ThomasF Wrote:  Replaced both transistors, and voila! Another Classic brought back from a faulty state!

I also found out that the picture you sent didn't correspond with the schematics.

Greats news :-)

Sorry about that, as I don't have a HP-80, I just assumed looking at component position relative to the CTC chip.

cheers

Tony

[Thomas Klemm]

(02-13-2024 02:35 PM)ThomasF Wrote:  Replaced both transistors, and voila! Another Classic brought back from a faulty state!

Congratulations!
Time to update your signature.

[ThomasF]

(02-13-2024 10:54 PM)Thomas Klemm Wrote:  

(02-13-2024 02:35 PM)ThomasF Wrote:  Replaced both transistors, and voila! Another Classic brought back from a faulty state!

Congratulations!
Time to update your signature.

Haha, thanks for the remainder Thomas!
Updated, but now I can't add any new ones, since the signature is too long ...

Cheers,
Thomas