PET 8096-SK boots to machine code monitor

Welcome back to our Commodore Discussion forum. If you have a question, comment, problem, or just want to chat about Commodore, post it here for free.

Have some fun and talk to others interested in the Old and New Commodore world.

Moderators: wiskow, Trazan

Post Reply
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

PET 8096-SK boots to machine code monitor

Post by spaceship »

Hi,

I just got myself one of these machines and it boots to what appears to be the machine code monitor:

Image

The text in the image (for search engines) is:

Code: Select all

b#
	pc	irq	sr	ac	xr	yr	sp
.;	c5b9	e455	31	a4	16	20	f6
.
I've heard that this could be symptomatic of a corrupt BASIC ROM.

Anybody have any ideas?

I can take photos of the motherboard if needed.

Many thanks,

Greg
User avatar
LoneWolf33
Posts: 92
Joined: Tue May 14, 2013 9:42 am
Are you a real person?: No... I am a spambot, delete my account!
Location: Budapest, Hungary

Re: PET 8096-SK boots to machine code monitor

Post by LoneWolf33 »

Hello, and welcome to the forum !

Congratulations for the machine, I think it's a rare piece.
For a start, when the boot sequence ends up in the ML monitor,
you can try to enter the following command:
m c5b0,c5b9
mem_dump.jpg
mem_dump.jpg (11.05 KiB) Viewed 18036 times
Compare the actual numbers with the ones in the screenshot.
If difference is found that suggests ROM related issue. If there's no difference at all
then the execution of the STY $6D command was failed (@ addresses $c5b8-$c5b9) for some reason.
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

Re: PET 8096-SK boots to machine code monitor

Post by spaceship »

Hi,

Thanks!,

I've always wanted a PET of some description and this is the one that I ended up with. Very pleased.

Here's the result of the command:

Code: Select all

.m c5b0,c5b9

.: c5b0  a0  20  c4  40  80  00  80  20
.: c5b8  00  35  01  50  00  40  a0  7f
It might be worth noting that one of the option ROM slots (UD12) is occupied too, not sure what with.

Cheers,

Greg
User avatar
LoneWolf33
Posts: 92
Joined: Tue May 14, 2013 9:42 am
Are you a real person?: No... I am a spambot, delete my account!
Location: Budapest, Hungary

Re: PET 8096-SK boots to machine code monitor

Post by LoneWolf33 »

Well, if the machine is equipped with Basic 4.0, the address $c5b0 should be the
entry point of a string processing routine.
The data you've provided can't be interpreted as a working ML routine, it contains
invalid instruction codes, and break instructions, preventing the machine from boot.
Even if I'm wrong in regard of the Basic version, this piece of code won't run.

This is how it looks when translating the numbers to Assembler:
Untitled.jpg
Untitled.jpg (24.18 KiB) Viewed 18014 times
I've marked the critical parts with arrows.

So the program execution is not possible from $c5b0, due to the errors.
My guess is that the machine crashes during displaying the power up message,
which needs this string processing subroutine.

I think your first hunch was right, the Basic ROM contents are seem to be corrupted.
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

Re: PET 8096-SK boots to machine code monitor

Post by spaceship »

That's great, thanks for your help. I found the corresponding ROM on ebay quite cheap.

It's soldered in so I ordered some sockets too in case I have to remove it more than once.

The next problem may be more difficult to solve though: the CRT only stays on for 2-3 minutes before fading/squashing down to a single horizontal line.

I'll get the ROM sorted first though and open a new thread for the CRT.

Thanks again!

Greg
User avatar
rbernardo
Posts: 3131
Joined: Fri Jun 09, 2006 8:00 pm
Are you a real person?: No... I am a spambot, delete my account!
Location: Visalia, California, USA
Contact:

Re: PET 8096-SK boots to machine code monitor

Post by rbernardo »

LoneWolf33 wrote:...the Basic ROM contents are seem to be corrupted.
Though Ray Carlsen has no experience with the 8096-SK, this is what he says for the C128, "In the 128, whatever keeps it from booting to the 128 opening screen will automatically dump it into the monitor if the rest of the computer is working. I would check the Kernal and BASIC ROMs." So, you should also check the Kernal ROM in your computer, too.

Back in California,
________________
Robert Bernardo
Fresno Commodore User Group
http://www.dickestel.com/fcug.htm
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

Re: PET 8096-SK boots to machine code monitor

Post by spaceship »

Thanks for your reply; I've found a replacement Kernal ROM online too, I'll try this if the BASIC one doesn't work alone.

Cheers,

Greg
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

Re: PET 8096-SK boots to machine code monitor

Post by spaceship »

Hi,

Made a little bit of progress with this. I installed both the BASIC and Kernal ROMs, but I still get the ML monitor when I boot it up, albeit with different values:

Image

Code: Select all

b#

	pc	irq	sr	ac	xr	yr	sp
.;	c6b9	e455	30	a4	16	20	f6
.
I have however realised that there are 2 other ROMS on the motherboard that pertain to BASIC (901465-20 and 901465-21), the one I have already replaced was the 901465-23, and the Kernal was 904465-22 (info from http://www.zimmers.net/anonftp/pub/cbm/ ... index.html).

Can't find a source for 20 and 21 online so far so if anyone knows where I might find those I'd be very grateful.

Cheers,

Greg

EDIT: Some shots of the motherboard for those interested: https://imgur.com/a/h3H00
User avatar
rbernardo
Posts: 3131
Joined: Fri Jun 09, 2006 8:00 pm
Are you a real person?: No... I am a spambot, delete my account!
Location: Visalia, California, USA
Contact:

Re: PET 8096-SK boots to machine code monitor

Post by rbernardo »

spaceship wrote:Can't find a source for 20 and 21 online so far so if anyone knows where I might find those I'd be very grateful.
Though I don't have a 8086-SK, I know that Rob Clarke in Switzerland has a few of those computers... and he's a pretty good repairman/tech and might be able to help you out on those ROMs. Send me a private message at rbernardo(at)iglou.com and I'll give you Rob's e-mail address.

Back in California,
________________
Robert Bernardo
Fresno Commodore User Group
http://www.dickestel.com/fcug.htm
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

Re: PET 8096-SK boots to machine code monitor

Post by spaceship »

Hi,

Thanks for your reply, I ended up finding someone locally so I got the chips already.

Installed them, system booted to basic 4.0 prompt! Only showed 32k of memory though, then I had to turn it off (as mentioned earlier in this post, I still haven't fixed the CRT so it switches itself off after a few minutes and I have to cycle the power).

Turned it back on, no nice chirping noise, and back to the ML monitor, this time with:

Code: Select all

c#
   pc   irq   sr   ac   xr   yr   sp
.;   0fff   e455   ff   00   ff   00   01
.
what does the c# refer to? It was b# before.

Cheers,

Greg

EDIT: On closer inspection, the first number in the above snippet seems to change, anywhere from 02ff to 82ff.
Also, I have unplugged the memory expansion board and put the CPU on the main board to narrow things down.
User avatar
LoneWolf33
Posts: 92
Joined: Tue May 14, 2013 9:42 am
Are you a real person?: No... I am a spambot, delete my account!
Location: Budapest, Hungary

Re: PET 8096-SK boots to machine code monitor

Post by LoneWolf33 »

As I know b* is displayed when program execution is aborted by a "brk" instruction (software interrupt).

RAM space between $0400 and $8000 is the Basic program area. If garbage is present at $0400-$0401
(e.g. due to a faulty RAM chip, or due to noise on the data bus) while the Basic is initialized during power-up,
then the interpreter starts (or at least it tries to start) to execute the garbage present in the RAM.
(without issuing the RUN command)

That might explains the random PC values.
spaceship
Posts: 8
Joined: Tue Aug 01, 2017 11:55 am
Are you a real person?: No... I am a spambot, delete my account!

Re: PET 8096-SK boots to machine code monitor

Post by spaceship »

Thanks, I will look into replacing the RAM then I guess.

Do you happen to know at what point in the boot process the chirping noise is initialised? Because now it doesn't chirp so I guess this could be a good diagnostic marker.
User avatar
LoneWolf33
Posts: 92
Joined: Tue May 14, 2013 9:42 am
Are you a real person?: No... I am a spambot, delete my account!
Location: Budapest, Hungary

Re: PET 8096-SK boots to machine code monitor

Post by LoneWolf33 »

As far as I can tell there's a subroutine call at address $E5B7 during boot.

The chirp routine itself is located at $E6A7, generates only one chirp,
so it's called multiple times during a successful boot.

Once Basic was started normally you can call the chirping routine manually by typing the sys59047 command.
User avatar
LoneWolf33
Posts: 92
Joined: Tue May 14, 2013 9:42 am
Are you a real person?: No... I am a spambot, delete my account!
Location: Budapest, Hungary

Re: PET 8096-SK boots to machine code monitor

Post by LoneWolf33 »

If it's not already done, I'd recommend to check all power supply voltages on the mainboard (+5V, +12V, -5V).
The 4116 RAM chips are extremely sensitive to the missing negative supply, it destroys the chip in no time.

Dead filter capacitors can cause fluctuating or too low voltage, causing abnormal operation of the supplied circuitry.
Post Reply