HP-91 Simulator

[teenix]

Hi all,

I have uploaded a simulator for the HP-91 printing desktop scientific calculator to teenix.org

I did not have access to the microcode for this calculator, so I looked through the various classic models and the HP-55 had the closest match for the functions required for the HP-91.

There were a few extra functions needed for the HP-91 so these were added by the Windows program tapping into the HP-55 microcode and getting it to execute the required mathematics. Therefore, the results should be exactly like the HP-91 would have produced.

The HP-55 code only had 1 ROM location modified due to the Woodstock style "Error" displayed on the HP-91, and of course, not all of it is used.

The simulation proved to be a difficult task as the Windows program has to monitor the code flow of the HP-55 so it can intercept it and extract register and key code data for the print routines.

It all seems to work as per the operating manual.

cheers

Tony

[Dieter]

(02-09-2017 05:46 AM)teenix Wrote:  I did not have access to the microcode for this calculator, so I looked through the various classic models and the HP-55 had the closest match for the functions required for the HP-91.

There was a major change in the internal routines, starting in 1976 with the HP67/97 and 91. At that time HP switched to the now common 13-digit internal precision, among other improvements e.g. in the trig routines. This significantly improved accuracy.

Details can be found in the November 1976 issue of HP Journal, p. 16 ("The New Accuracy: Making 2³=8").

(02-09-2017 05:46 AM)teenix Wrote:  Therefore, the results should be exactly like the HP-91 would have produced.

Sure ?-) Try the examples listed in the mentioned article.
What do you get for tan(pi) in radians mode? Is it 0 or 4,1 E–10 ?

Dieter

[teenix]

(02-09-2017 08:22 AM)Dieter Wrote:  There was a major change in the internal routines, starting in 1976 with the HP67/97 and 91. At that time HP switched to the now common 13-digit internal precision, among other improvements e.g. in the trig routines. This significantly improved accuracy.

Details can be found in the November 1976 issue of HP Journal, p. 16 ("The New Accuracy: Making 2³=8").

Dieter

Always something you find out about too late :-)

When I get some spare time I will take a further look.

Thanks for the tip.

cheers

Tony

[fhub]

(02-09-2017 05:46 AM)teenix Wrote:  I have uploaded a simulator for the HP-91 printing desktop scientific calculator to teenix.org

Thanks for your continuously growing collection of HP emulators, Tony!

There seems to be a little problem with the graphics - see attached image.

Franz

[teenix]

(02-09-2017 09:26 AM)fhub Wrote:  Thanks for your continuously growing collection of HP emulators, Tony!

There seems to be a little problem with the graphics - see attached image.

Franz

Thanks, it's been quite a trip considering I only started out to do the HP-45.

I think I fixed the problem. I forgot to turn off auto scaling for the program so different screen resolution may have made it visible.

New upload available.

cheers

Tony

[fhub]

(02-09-2017 10:25 AM)teenix Wrote:  I think I fixed the problem.

Yes, now it's looking well again.

Thanks,
Franz

[SlideRule]

When I attempt to uncompress this ZIP® file my NORTON® deletes the EXE file and posts a note for infection of same with the WS.Reputation.1 virus.
Does anyone else have this or a similar issue?

ps: many of the previous archives did NOT have this issue & it seems like this is a recent phenom.

BEST!
SlideRule

[Joe Horn]

(02-09-2017 02:55 PM)SlideRule Wrote:  When I attempt to uncompress this ZIP® file my NORTON® deletes the EXE file and posts a note for infection of same with the WS.Reputation.1 virus.
Does anyone else have this or a similar issue?

Neither ESET NOD32 nor Malwarebytes find anything bad in the EXE file.

[Joe Horn]

Possible bug detected:

After turning on, pressing [shift] [Sigma Minus] returns -1 as it should, but if instead you press FIX 9 first (before pressing sigma minus), it returns 0 (or doesn't do anything... I can't tell) and if you press X<>Y after that, it errors.

[fhub]

(02-09-2017 03:31 PM)Joe Horn Wrote:  ... but if instead you press FIX 9 first (before pressing sigma minus), it returns 0 (or doesn't do anything... I can't tell) and if you press X<>Y after that, it errors.

No problem here with this key sequence, it correctly displays -1.000000000, and also X<>Y gives no error at all.

Edit: are you sure you have the latest version?
There was already a quick update because of a small display problem.

Franz

[SlideRule]

(02-09-2017 03:15 PM)Joe Horn Wrote:  Neither ESET NOD32 nor Malwarebytes find anything bad in the EXE file.

I'm using version 22 of Norton 360 (see attached screen capture) and get a green √ after an update query.
[attachment=4471]

BEST!
SlideRule

[teenix]

(02-09-2017 02:55 PM)SlideRule Wrote:  When I attempt to uncompress this ZIP® file my NORTON® deletes the EXE file and posts a note for infection of same with the WS.Reputation.1 virus.
Does anyone else have this or a similar issue?

ps: many of the previous archives did NOT have this issue & it seems like this is a recent phenom.

BEST!
SlideRule

Norton has the [Reputation] feature for its own software and states that it can malfunction on some programs that it tests. This has cropped up on odd occasions before when downloading the emulator files. I don't know how to fix this except to submit the file to Norton so they can find out why it triggers the response.

I'm pretty sure the file is clean, so you can get it back from Nortons vault.

cheers

Tony

[teenix]

(02-09-2017 03:31 PM)Joe Horn Wrote:  Possible bug detected:

After turning on, pressing [shift] [Sigma Minus] returns -1 as it should, but if instead you press FIX 9 first (before pressing sigma minus), it returns 0 (or doesn't do anything... I can't tell) and if you press X<>Y after that, it errors.

I found out this problem did occur when the print mode button is in the MAN position.

I made a modification and it doesn't do it here anymore.

New upload available

cheers

Tony

[SlideRule]

(02-09-2017 09:29 PM)teenix Wrote:  This has cropped up on odd occasions before when downloading the emulator files. I don't know how to fix this except to submit the file to Norton so they can find out why it triggers the response. I'm pretty sure the file is clean, so you can get it back from Nortons vault.

Thanks for the reply. This is an on-off-on again phenom so I'm confident your (& other member) validation is sufficient. Great emulators, many thanks!

BEST!
SlideRule

[fhub]

(02-09-2017 09:52 PM)teenix Wrote:  I found out this problem did occur when the print mode button is in the MAN position.
I made a modification and it doesn't do it here anymore.

Talking about bugs - here's one more:
Switching 'Calculator Key Legend' ON/OFF doesn't work on both HP91 and your last HP97, it's always ON independent of the setting.

Franz

[teenix]

(02-09-2017 11:00 PM)fhub Wrote:  Talking about bugs - here's one more:
Switching 'Calculator Key Legend' ON/OFF doesn't work on both HP91 and your last HP97, it's always ON independent of the setting.

Franz

Those bugs are friendly little critters, always hanging around :-)

Now fixed and uploaded for both programs.

I found while developing the HP91 that it was a bit frustrating at times not to be able to look at the printer roll to see the data that scrolled off screen, so if you now double click the printer paper area on HP91 and HP97 a window will open allowing you to scroll through the entire printed roll.

As it is a bit more complicated to do, I will add this feature to HP10 (KISS), HP19C and HP46 when I have time.

cheers

Tony

[Joe Horn]

'Nuther bug: The three LIST commands (e.g. LIST STACK) don't work in MAN mode. "LIST" is printed, then nothing else.

And another one: -1E-10 prints as -0 in any FIX setting.

And -5.6E-78 displays ok in SCI 2 mode, but ENG 2 mode displays it as -5.6E-22.

[rprosperi]

(02-09-2017 02:55 PM)SlideRule Wrote:  ...my NORTON® deletes the EXE file and posts a note for infection of same with the WS.Reputation.1 virus.
Does anyone else have this or a similar issue?

The "reputation" signature hit really means that Norton's database has never seen this before. You can Restore the file that was exiled and then tell Norton to not flag it again; this also tells the Norton mother ship that it has been trusted and after a few of these from various folks, the reputation is restored.

In this case, Tony will probably have a new version soon, so the whole cycle could repeat. So just Restore the file and you'll be OK.

[teenix]

Hi all,

Sorry, I have had to pull the HP-91 from the web page.

I should have looked deeper into its operation before starting and I cannot make it work properly based on HP55 code, there are too many differences that I wasn't aware of - (Thanks Dieter).

I will remodel it based on the later technology, but it will take awhile.

Apologies for those who spent their time to download and try it, and thanks for the bug reports.

cheers

Tony

[teenix]

Hi all,

A new HP-91 simulator has been uploaded to teenix.org

It was completely rewritten and is now based on 20 series processing.

I have tested and tested but I'm sure a bug will appear :-)

cheers

Tony