HP97 The journey begins

[teenix]

Hi all,

I thought this was an interesting snap shot.

Traces are...
1) SYNC pulses. (About 310uS apart) which is the HP-97 instruction (56 bit) word time.
2) IaIs bus - which has the ROM instruction, ROM address, display data etc.
3) Data bus - X register in the stack = 0
4) Flag 2 - connects to PIK, CRC and ARC.

It shows when the HP-97 code asks the PIK chip if a keypress is available. (pik1320 - First Sync pulse)
You can see the pulse on the Flag 2 trace that came from the PIK which is telling the ARC that a key is available from the PIK's 7 key buffer.

The 3rd instruction is a GOTO $079. This is actually stored in ROM as $079 which is the address where the HP-97 main code loop starts and normally jumps back there when a key is not pressed, among other things.

The Sync line stayed LOW for this instruction because the previous instruction was an "IF" type (if s3 = 0). This tells the ARC that $079 is a ROM address to GOTO, not a code word to execute. However, because the Flag 2 pulsed HIGH, (which sets ARC Status Flag s3 = 1) this GOTO instruction is ignored.

The instruction at the last sync pulse is the instruction that turns the display off while the key press is serviced.
The S3 flag stays as is in the ARC until it is reset by a 0 -> s3 instruction which in this case follows the display off.

Code:


Microcode section             Sync pulse
0094    pik1320               1
0095    if s3 = 0             2
0096      then go to $079     3 (masked)
0097    display off           4
0098    0 -> s3

cheers

Tony

[Chr Yoko]

Great ! Signals are pretty clean and sharp too !

[teenix]

(03-04-2021 05:19 PM)Chr Yoko Wrote:  Great ! Signals are pretty clean and sharp too !

This was logic capture on my DSO. The analog traces are much noisier.

cheers

Tony

[teenix]

Hi all,

Sad days trying to design the new 97 board. I seem to be following the proverbial 1 step forward, 2 steps backward and perhaps got a bit jinxed along the way - maybe the HP Gods are trying to tell me something.

Just as I was about to print out the final design after many hours of work, my hard disk failed and looks to have damaged the data on the disk. The PC tries to fire up but eventually after displaying multiple disk errors, fails. Scan disk fails just after starting too if I invoke it at boot time. Lots of old and new files gone. I did backups along the way, but design by trial and error meant I did a new board from scratch over last the few days and is all gone. This means my old Protel software has finally fizzed so I'll need a new PCB design package (probably KiCad) to learn and continue from scratch.

cheers - well, a little bit still

Tony

[albertofenini]

Hi Tony

It may be worth to try this
Get yourself another bootable pc
Take this hard disk out and try to use it as an external unit
It may have a boot record damaged but you may still be able to access data

Hope it works !
Take care !

[teenix]

(03-14-2021 08:27 AM)albertofenini Wrote:  Hi Tony

It may be worth to try this
Get yourself another bootable pc
Take this hard disk out and try to use it as an external unit
It may have a boot record damaged but you may still be able to access data

Hope it works !
Take care !

Thanks Alberto,

I thought of this but it is an old 486 that I still happily use for certain jobs and for those it works just fine. I bought a new Win10 machine the other day, but new PC, new software, more money and some of my old Win 7 creations will need a rewrite :-(

What did Bones say on Star Trek - "Engineers they just love to change things" :-)

I tracked down another hard disk on the web, so maybe I can use it to help recover some stuff. I am not sure though because I think a lot of sectors have failed. Windows does try to start but dies after trying to access certain parts of the disk and this process is repeatable.


cheers

Tony

[BobVA]

(03-14-2021 08:44 AM)teenix Wrote:  ...I tracked down another hard disk on the web, so maybe I can use it to help recover some stuff. I am not sure though because I think a lot of sectors have failed. Windows does try to start but dies after trying to access certain parts of the disk and this process is repeatable.

Perhaps try a software recovery tool like SpinRite before diving into the hardware?

Bob

[AndiGer]

If old PC maybe your disc was formatted FAT not NTFS?
IIRC I could access parts of some failing disc as external disc using an old Win98 machine. Maybe also possible using XP?
Don't remember I did chkdsk or not ...

[teenix]

I think it is finally time to retire my old beast of a PC.

Because it was a lot of work to initially set up, and after some fiddling, I got the 97 board outline shape and component positions from a memory stick that I had an earlier backup on. I was then able to convert this to a Gerber file on my old WinXP machine and finally managed to import this into KiCad as a PCB design on my new Win10 laptop.

Nothing old is compatible with the new stuff so schematics etc will have to be recreated from scratch.

A lot to catch up on and learn, but it's a start.

cheers

Tony

[Eric Rechlin]

Sounds like a good time to setup a virtual machine on your Windows 10 computer and run an instance of Windows 3.1 or 95 or whatever software you had on the old 486 inside that virtual machine.

[Guenter Schink]

(03-15-2021 03:20 PM)Eric Rechlin Wrote:  Sounds like a good time to setup a virtual machine on your Windows 10 computer and run an instance of Windows 3.1 or 95 or whatever software you had on the old 486 inside that virtual machine.

I went through this just recently. Unfortunately my AMD (Ryzen 5 3600) gave me some headache. On VMware I had to go back to VMware 12 in order to get Win98 and Win95 installed.
Meanwhile I've got successfully WIN98SE and WIN95B installed on VMware 12.

On VirtualBox there were a lot of Problems too.
I don't really remember which OS required an earlier version of VirtualBox to get installed. But after everything was installed, It worked also on the most recent version of VirtualBox. WinXP, Win95B, FreeDOs12, OS/2 4.5,

Günter

[[kby]]

As you said in another post,
Bugger!

If you get to the Hail Mary point I can try to image whatever is imagable off of it. Fir you, I’ll put on the old sysadmin hat. What kind of interface is it? In the meantime, I would n]ot power the drive up unless you have some really good reason to; every time up could contribute to the degradation (assuming it had a head crash; worst case). If it’s just an issue with parameters or a the boot sector, that’s easy. Since your machine sounds old you may want to check applicable BIOS parameters for DIDI geometry, in case they inadvertently got corrupted or erased.

Anyway, I’m willing to give it the recovery try. With of course no guarantee you’ll get anything useful.

-kby

(03-14-2021 08:14 AM)teenix Wrote:  Hi all,

Sad days trying to design the new 97 board. I seem to be following the proverbial 1 step forward, 2 steps backward and perhaps got a bit jinxed along the way - maybe the HP Gods are trying to tell me something.

Just as I was about to print out the final design after many hours of work, my hard disk failed and looks to have damaged the data on the disk. The PC tries to fire up but eventually after displaying multiple disk errors, fails. Scan disk fails just after starting too if I invoke it at boot time. Lots of old and new files gone. I did backups along the way, but design by trial and error meant I did a new board from scratch over last the few days and is all gone. This means my old Protel software has finally fizzed so I'll need a new PCB design package (probably KiCad) to learn and continue from scratch.

cheers - well, a little bit still

Tony

[teenix]

(03-16-2021 07:13 AM)[kby] Wrote:  As you said in another post,
Bugger!

If you get to the Hail Mary point I can try to image whatever is imagable off of it. Fir you, I’ll put on the old sysadmin hat. What kind of interface is it? In the meantime, I would n]ot power the drive up unless you have some really good reason to; every time up could contribute to the degradation (assuming it had a head crash; worst case). If it’s just an issue with parameters or a the boot sector, that’s easy. Since your machine sounds old you may want to check applicable BIOS parameters for DIDI geometry, in case they inadvertently got corrupted or erased.

Anyway, I’m willing to give it the recovery try. With of course no guarantee you’ll get anything useful.

-kby

That is a pretty good offer and much appreciated. I tracked down another hard disk here in Oz but with all the effort of reinstalling and trying to find my original install CD and boot disk which also may have failed, I think I will just resign the old beast to a recycler. No doubt something else will fail at some stage.

This KiCad board design software is nowhere near as user friendly as the old Protel - such is life :-)

Thanks again

cheers

Tony

[teenix]

Hi all,

I got the board designed. I ended up playing around with WinXP and mostly got it working although it was cranky and needed restarts. Some of the problems is the indexing of the drop down list boxes and some text among other things. I'm pretty sure the data format of these items changed with later Windows and is not dealing with them too well.

Anyway, it needs another day to tidy it up and do some checking over, but at least it looks the part.

cheers

Tony

[rprosperi]

(03-16-2021 03:55 PM)teenix Wrote:  Hi all,

I got the board designed....

Nice rescue Tony, would have been sad to lose all the work you've done on this cool project. Well done!

[Massimo Gnerucci]

Pheeeeeeeeeeew!

[[kby]]

(03-16-2021 09:03 AM)teenix Wrote:  That is a pretty good offer and much appreciated. I tracked down another hard disk here in Oz but with all the effort of reinstalling and trying to find my original install CD and boot disk which also may have failed, I think I will just resign the old beast to a recycler. No doubt something else will fail at some stage.

This KiCad board design software is nowhere near as user friendly as the old Protel - such is life :-)

Thanks again

cheers

Tony

I wasn’t meaning to get the disk in usable shape. That drive should never be used again for anything you care about. I meant just to scavenge any useful data that you might have on it that could prevent you from having to re-invent the wheel totally. That’s why I’m only suggesting it when your choices are chuck it and lose everything or potentially getting some useful personal files off before chucking it completely. I would not plan on returning the physical disk to you but only getting you a copy of whatever content was recoverable (if any).-kby

[teenix]

(03-16-2021 06:55 PM)[kby] Wrote:  I wasn’t meaning to get the disk in usable shape. That drive should never be used again for anything you care about. I meant just to scavenge any useful data that you might have on it that could prevent you from having to re-invent the wheel totally. That’s why I’m only suggesting it when your choices are chuck it and lose everything or potentially getting some useful personal files off before chucking it completely. I would not plan on returning the physical disk to you but only getting you a copy of whatever content was recoverable (if any).-kby

That's ok, I wasn't expecting the disk back and your idea most likely would have gotten some results. While I was running around in circles trying to sort a solution, I thought another hard disk might be able to be installed and put the failed one in as a slave. Maybe I could have recovered something. Many thanks again

cheers

Tony

[teenix]

Hi all,

Well I bit the bullet, the 97 CPU replacement PCB design has gone to the fabricator :-)

cheers

Tony

[teenix]

Hi all,

The 97 PCB's arrived already - 4 day turnaround.



Wouldn't you know it, my boss called inundated with work and wants me to fly again for 6 weeks or so starting tomorrow

cheers

Tony