• Please review our updated Terms and Rules here

Basic 4.0 upgrade issues

A

acenewman

Experienced Member
Joined
Apr 6, 2020
Messages
117
Location
Farmington MI
I was kind of bored this evening and thought why not upgrade my PET from Basic 2 to Basic 4. So after some digging I came up with the following ROM images as the correct ones needed:

D5 - 901465-19
D6 - 901465-20
D7 - 901465-21
D8 - 901474-29
D9 - 901447-22
F10 - 901447-10

My machine is a 2001N-16 with graphics keyboard and currently runs Basic 2 with no issues. I pulled all of the binaries from Zimmers.net and burned them using my trusty EMP-10. All images verified fine after burning. I am getting the same random looking garbage screen I got previously when my Editor ROM for Basic 2 was bad. Burning a new -24 ROM fixed that issue. First question is does it look like I am using the right set of ROM images for basic 4.0? I reburned and replaced all of the ROMS a second time with no change. Also to rule out some other random failure I put the Basic 2 ROMs back in and it works fine. Assuming these are the correct set of ROMs I am thinking I should start troubleshooting the D5 socket as Basic 2 does not use it.
Perhaps one of the traces are bad??

I appreciate any feedback that can be provided. Thanks, Chris
 
When you say you burnt the ROMs, do you mean you burnt the EPROMs?

If so, what type (manufacturer and part number) are you using? Depending upon the specific PET, you can use a ‘normal’ EPROM of a particular type in the EDIT socket, but the others can be more problematic (and may require adapters). I will check a bit more presently when you tell me exactly what you have usef.

Dave
 
A 2532 EPROM should work OK in a 2001N - but not a 2732.

I agree with most of your ROM IDs apart from one (D9 - $Fxxx). I assume 901447-22 is a typo. and you mean 901465-22? There is a 901447-22 (but it won't work).

Dave
 
Last edited:
The edit rom D8 is usually 2716 (2k) eprom. Does your board have the white sockets as they are notorious for bad connections.

Andy
 
Yeah sorry, should have specified. I used TI 2532 for the 4K ROMS and AMD 2716B for the 2K ROMS. And yes, good catch, D9 is a typo. 901465-22 is what I actually used. I have white sockets only in the system and video ram locations.
 
Last edited:
Should be good then. Yes, we did spot the deliberate mistake!

Can you get your EPROM programmer to display the 16-bit checksum for each ROM - I can compare that with my documentation for my PETTESTER ROM. That should tell us whether they are good or not.

What speed are your EPROMs? Too slow and they will misread (although it is unlikely that you will be able to purchase EPROMs that are too slow for a PET)!

You could always try a copy of my PETTESTER in the EDT ROM socket. This will require an operational kernal ROM (or at least the first few instructions and the reset vector). This will test out your RAM and ROM checksums (other than the EDIT ROM of course).

Andy's suggestion about the dreaded white IC sockets is also a good place to start (but my PETTESTER will give you the checksum so you will know if it is good or not).

Dave
 
Dave,
Pretty sure the EMP-10 will give me a checksum value. Just need to consult the manual for the details. Will set it up after work and see what I get. Think I have at least one spare 2716B so will burn a PETTESTER ROM as well while it's out. Chris
 
Ah yes, RTFM...

My PETTESTER documentation has details of the expected 16-bit checksum for the Version 4 ROMs.

Dave
 
The latest D5 ROM ($B000) part number is 901465-23. The -19 is older but I think should still be OK. I'm not sure what the change was for. Does anyone know?
 
Well the good news is that Basic 4.0 is working fine now. The bad news is the issue was the "chair to keyboard interface" ;)

Started checking the checksum values and noticed right away that two of my basic ROMs had the same checksum. With my programmer (EMP-10) you have to pick the binary you want to program and in a second step you have to tell it to load that file into the buffer. My hunch is that I picked the second file but skipped the step of loading it into the buffer. Thus programing two ROMs with the same contents. Erased and reprogramed the one ROM and all is good now.

In other news all of the Zimmer.net binary checksum values matched up with what is in Dave's PETTESTER documentation. Not that I didn't think they would, but it's always nice when the numbers add up. Just in case it might be useful to someone else, here is the actual checksum values reported . A couple of these are not listed in the PETTESTER documentation.

D5 - 901465-19 checksum 4C5E
D5 - 901465-23 checksum 4168
D6 - 901465-20 checksum 5960
D7 - 901465-21 checksum A425
D8 - 901474-29 checksum 8854
D9 - 901465-22 checksum CF19
F10 - 901447-10 checksum 7B67

With that out of the way, I am going to go start playing with my recently acquired 8050 disk drive. It survived shipping with no physical damage and with a simple head cleaning one of the two drives is already reading and writing disks. Gonna do a deep clean and some lubricating and see if I can get the other drive working. Thanks for all the input, Chris
 
Ah well, you found the issue!

My PETTESTER has no access to the character generator ROM (so it can't compute the checksum) and it sits in the EDIT ROM socket (so I can't compute the checksum of the PET EDIT ROM firmware).

However, I didn't know that there was two variants of the D5 ROM. I will get around to updating the PETTESTER documentation. It would be good to know what the patches are though (as dave_m has already asked).

Dave
 
You must log in or register to reply here.
Back
Top