HP-97 weird behaviour

[AndiGer]

Hi all,

I have a HP-97 here that does not calculate correctly, but I don't can get to it what's wrong.

Some keys work ok as X-square or square root but some don't.
Positive number + cos gives 9.999999999 99
Negative number + cos gives -9.999999999 99
Any number + tan results in .00 and a minus sign in the exponent.
Strange: 10 + 1/x gives 0.10, 100 + 1/x gives 0.01 (any other exponents of 10, also negative ones give correct result). But 50 or 500 or 5.000 + 1/x gives 5.00 (or more 0 zeroes as DSP is set to more digits).
657 + 1/x gives 6.57, after about 1/10 second the displays flashes shortly to all 0 including exponent (as I can see - looks like a program is running but only very short) and then 6,57 again.
Pressing ->R after power on results in 0.000000000 exponent 0d. Weird.

Some function key pressed set the internal program steps counter up by 1! The first step is incremented by 2.
E.g.: Calc off, Prgm on shows step 000. Prgm off, e to the x, Prgm on shows 002 51. Prgm off, e to the x, Prgm on shows 003 51. Incremented to 004 51 after same procedure.

In Prgm mode nearly all steps are registered as 51 (i.e. R/S), but some not.
tan sets the program counter back to 000 as does RTN.
sin, 1/x 1, 2, 7 or +-key result in Error in the display.
R/S pressed results in a shortly flashing display and step counter incremented by 2.
e to the x results in a shortly flashing display and also step counter incremented by 2.
Pressing x<>y followed by 7 results in a blank display.

I changed the complete keyboard including PCB with a known good one. Behaviour is the same.
I also checked connections from the chips to the golden fork connections and test points resp. All are ok, no interruptions there.

I have the service manual but does not help for that case I guess. Some advice from the experts here?

BTW: I can create program cards on my other HP-97 if necessary. The card reader of the confused machine is also repaired, so it can read programs. Don't know by now it runs the read programs correctly. Don't think so.

TIA
Andi

[Paul Berger (Canada)]

That unfortunately sounds like a failed ROM. I had a 67 that displayed the similar symptoms and I isolated it down to one of two ROMs.

[AndiGer]

The HP-97 has 6 ROMs, numbered 0 thru 3, 5, 6.
Source for replacement is unfortunaltely only another HP-97 if I'm right :-(

[PANAMATIK]

(05-02-2016 07:35 PM)AndiGer Wrote:  The HP-97 has 6 ROMs, numbered 0 thru 3, 5, 6.
Source for replacement is unfortunaltely only another HP-97 if I'm right :-(

You are right. But this might not be true forever. Im planning to develop ROM replacements for HP97, but unfortunately will not have time before next year. And there should be demand for at least 10 repairs

Bernhard

[Massimo Gnerucci]

Bernhard, you've officialy become The Replacer.

:)

[PANAMATIK]

(05-02-2016 08:31 PM)Massimo Gnerucci Wrote:  Bernhard, you've officialy become The Replacer.

To replace the complete HP-97 ROM chip set, only one ROM replacement chip would be necessary, which contains the complete 6k firmware. You don't have to bother with asking which ROM is defective, just one chip replaces them all and it could even contain RAM and replaces defective RAM chips. Of course an ACT for HP-97 would do the same and would be the better solution, but this needs a new concept including exact timing for card reader and printer, it is much more difficult than a new RAM/ROM chip.

Bernhard