• Please review our updated Terms and Rules here

CBM PET 4032 Restoration

Dide you have a PIA missing (and on order)?

The sounder is driven by BOTH the VIA and one of the PIAs (keyboard I think, but I would have to check).

Swapping back the 16K banks would have made data bit 0 of the replaced RAM stuck '1' then. That device possibly went faulty between powering off and on...

Dave
 
daver2 said:
Dide you have a PIA missing (and on order)?

The sounder is driven by BOTH the VIA and one of the PIAs (keyboard I think, but I would have to check).

Swapping back the 16K banks would have made data bit 0 of the replaced RAM stuck '1' then. That device possibly went faulty between powering off and on...

Dave
Click to expand...
I have another PIA on order. There is currently one missing.

Im not sure what happened with the ram test then, but it seems to be ok now.
 
Who knows...

The correct sequence of events should have been to run PETTESTER with the 'CAS swap on' to see if PETTESTER detects 32K RAM and, if so, whether it tests OK (with the 16K bank we are working with in the upper bank).

Remember, one step at any one time.

But, if everything works OK now...

Dave
 
daver2 said:
Who knows...

The correct sequence of events should have been to run PETTESTER with the 'CAS swap on' to see if PETTESTER detects 32K RAM and, if so, whether it tests OK (with the 16K bank we are working with in the upper bank).

Remember, one step at any one time.

But, if everything works OK now...

Dave
Click to expand...
I will do that as a matter of interest. I havent permanently removed the bodge.
 
daver2 said:
If you are waiting, you could test the IEEE488 port. Did I post the book reference and page number?

Dave
Click to expand...
Yes thanks Dave, you posted that. I was just watching the Darts and remembered it. I will try it out tomorrow.

I think I could use the memory dump feature of Romulator to save it, as my new PIA is 'in the post'.
 
Hugo Holden said:
Can you explain how to put the programs into the 2114 video RAMs, leaving their other functions un-altered and how to run the programs, from BASIC, from that location ? It sounds like a good idea.

Though BASIC can only run of course if the lower 1k of DRAM is working anyway. So if it is (or a substitute RAM is) there & working, there is is no worry running the diagnostic programs block moved to that location in the unused area, so I cannot see the advantage yet, of putting the memory fill and clear programs in the Video RAM, though I might be missing something.
Click to expand...

Hugo - a few months ago on vintage-radio.net, I had a lot of help fixing a 4032. Part of that was a new bit of diagnostics written that only used video RAM to run the program. The github link with souce code, compiled code to burn in a ROM, descriptions and screenshots is here:

github.com

GitHub - JulieMontoya/ToePost: A(nother) PET memory test utillity

A(nother) PET memory test utillity. Contribute to JulieMontoya/ToePost development by creating an account on GitHub.
github.com github.com

The author was very very helpful through this process and I'm sure they will answer any questions you may have.

Hope this helps.

Colin.
 
daver2 said:
If you are waiting, you could test the IEEE488 port. Did I post the book reference and page number?

Dave
Click to expand...
Thinking about it I cant until the PIA turns up. I am using the one for the IEEE to run the keyboard. I think the new one is arriving today.
 
Programming EPROMS is a lot of fun as it turns out eh ? I bought a T48 programmer. From my research, they removed the older eproms from the device list after version 13.02 of XGPro. Turns out this is because the new T48 doesn't support 21v.

If I am correct, I have a 2532 which should work as a Kernal replacement - but I cannot use it in the T48 because I am running a later version of XGPro.
I also have a 27C32, which should work in the T48, but will need an adaptor in the PET (2732 to 2332). This also wont program because 1 byte is full (could be faulty? Who knows), and it requires 21v to program.

Just had a shipping update on my PIA - now sometime in December.

Can anyone recommend a modern replacement Eprom from use in my T48 for the kernal please ?

As a last resort - I've ordered one of these :

www.ebay.co.uk

One ROM - Replacement 8KB 4KB 2KB ROMs for vintage retro computers c64 vic20 pet | eBay UK

Pins are supplied for you to install if you wish – you may choose not to if your system has limited clearance above the ROM. Alternatively, include a note with your order that you would like the header pins installed and we will pre-solder them for you.
www.ebay.co.uk
 
Last edited:
Hugo Holden said:
Can you explain how to put the programs into the 2114 video RAMs, leaving their other functions un-altered and how to run the programs, from BASIC, from that location ? It sounds like a good idea.

Though BASIC can only run of course if the lower 1k of DRAM is working anyway. So if it is (or a substitute RAM is) there & working, there is is no worry running the diagnostic programs block moved to that location in the unused area, so I cannot see the advantage yet, of putting the memory fill and clear programs in the Video RAM, though I might be missing something.
Click to expand...
It would be hard to do form Basic, but with assembler / machine code you can have your code and data anywhere in memory.
 
I have never used that particular programmer.

But there are plenty of threads on here from people who have (I am sure).

Yes, the PET requires the 2532 pinout, so a 2732 will require an adapter on the PET side.

Dave
 
Lol. I just need to stop rushing. Waiting for parts is a killer for me. Im going to use that ebay replacement for now, and then spend sometime researching the eprom. I would like to get it as stock as possible- otherwise I may as well just leave the Romulator in it !

I cant load any software from my PetDisk because of the PIA. I may take it apart and restore the metal chassis for the monitor. The key caps are also quite yellow (graphics chars are still white). I also see if I can do something about that.
 
Just my 5 cents worth on the issue of PET Roms:

The Roms you mainly need are the TMS2532. These have a 25V programming voltage. (The 2716 is handy for the character ROM and the Pettester. But if a 2716 is used for the character ROM it must be disconnected from the /INIT line and that pin tied high)

There are plenty of TMS2532A chips out there, these have 21V programming voltage. Even when a programmer supports these, they are notoriously troublesome to program. People in the video game industry have struggled with these 21V parts. I tried a few myself and got unreliable results. Don't get the A version.

My advice is to keep well away from the 21V parts, only ever buy the TMS2532 not the "A" version and don't go to the 2732 either initially unless you have adapters that are known good. And avoid the 2732A as well.

Many programmers can program a 2732A (which is a 21V part, so you can program the 2532A on an adapter this way) but I would caution not to do it, especially in fault finding scenarios, where you have to rely on the ROMs you make being 100% good or it throws in another variable to confound the process. Obviously if everthing is working fine later you can experiment with different ROMs and programming permutations, but not initially unless you want to dig your own grave and shoot yourself in it.

Also it is better to buy new/unused TMS2532 as they are more likely to be in good order and undamaged and you won't require an eraser.

The original TMS2532 are out there, if you hunt around and are careful what you buy. They have a large die and a central rectangular area on the package. Exactly like these ( I have used this seller and these parts are good)

www.ebay.com

TMS2532JL-45 Texas Instruments 32K EPROM - UV Eraseable | eBay

Find many great new & used options and get the best deals for TMS2532JL-45 Texas Instruments 32K EPROM - UV Eraseable at the best online prices at eBay! Free shipping for many products!
www.ebay.com

These 25Vpp parts program perfectly in the GQ-4x programmer. It gives trouble free programming and supports many other vintage chips. During programming it is wise to have the power adapter connected and not just rely on the USB power alone. It is also in a proper housing and comes with support software on a DVD. It runs in Windows and uses a USB link.
 
Hugo Holden said:
Just my 5 cents worth on the issue of PET Roms:

The Roms you mainly need are the TMS2532. These have a 25V programming voltage. (The 2716 is handy for the character ROM and the Pettester. But if a 2716 is used for the character ROM it must be disconnected from the /INIT line and that pin tied high)

There are plenty of TMS2532A chips out there, these have 21V programming voltage. Even when a programmer supports these, they are notoriously troublesome to program. People in the video game industry have struggled with these 21V parts. I tried a few myself and got unreliable results. Don't get the A version.

My advice is to keep well away from the 21V parts, only ever buy the TMS2532 not the "A" version and don't go to the 2732 either initially unless you have adapters that are known good. And avoid the 2732A as well.

Many programmers can program a 2732A (which is a 21V part, so you can program the 2532A on an adapter this way) but I would caution not to do it, especially in fault finding scenarios, where you have to rely on the ROMs you make being 100% good or it throws in another variable to confound the process. Obviously if everthing is working fine later you can experiment with different ROMs and programming permutations, but not initially unless you want to dig your own grave and shoot yourself in it.

Also it is better to buy new/unused TMS2532 as they are more likely to be in good order and undamaged and you won't require an eraser.

The original TMS2532 are out there, if you hunt around and are careful what you buy. They have a large die and a central rectangular area on the package. Exactly like these ( I have used this seller and these parts are good)

www.ebay.com

TMS2532JL-45 Texas Instruments 32K EPROM - UV Eraseable | eBay

Find many great new & used options and get the best deals for TMS2532JL-45 Texas Instruments 32K EPROM - UV Eraseable at the best online prices at eBay! Free shipping for many products!
www.ebay.com

These 25Vpp parts program perfectly in the GQ-4x programmer. It gives trouble free programming and supports many other vintage chips. During programming it is wise to have the power adapter connected and not just rely on the USB power alone. It is also in a proper housing and comes with support software on a DVD. It runs in Windows and uses a USB link.
Click to expand...
Thanks Hugo. I will have a look at those.
 
Hugo Holden said:
The Roms you mainly need are the TMS2532. These have a 25V programming voltage. (The 2716 is handy for the character ROM and the Pettester. But if a 2716 is used for the character ROM it must be disconnected from the /INIT line and that pin tied high)

There are plenty of TMS2532A chips out there, these have 21V programming voltage. Even when a programmer supports these, they are notoriously troublesome to program. People in the video game industry have struggled with these 21V parts. I tried a few myself and got unreliable results. Don't get the A version.

My advice is to keep well away from the 21V parts, only ever buy the TMS2532 not the "A" version and don't go to the 2732 either initially unless you have adapters that are known good. And avoid the 2732A as well.

Many programmers can program a 2732A (which is a 21V part, so you can program the 2532A on an adapter this way) but I would caution not to do it, especially in fault finding scenarios, where you have to rely on the ROMs you make being 100% good or it throws in another variable to confound the process. Obviously if everthing is working fine later you can experiment with different ROMs and programming permutations, but not initially unless you want to dig your own grave and shoot yourself in it.
Click to expand...
Related tangent:
Does this warning apply to all of these old 21V EPROMs, or is it just that the 2532A are known to cause grief?
I'm asking because I had trouble programming a pair of 2732(A) EPROMs (can't remember if they were A or not) a while ago with my ALL-03 and I had to restart the programming so it would program each 256 byte block twice, as the checksum failed the first time, seemingly due to a too short programming time.

I didn't investigate this further at the time, I.E. I can't remember for example measuring anything using an oscilloscope or whatnot while programming, so I don't know if the programming voltage and the length of the programming pulse is/was correct, but still. (A variable here is that the software is really intended for running in DOS but due to reasons I'm running it under Windows XP with a third party add-on driver thing that opens up a hole allowing the DOS VDM to access the I/O ports of the ALL-03 ISA card, so there could be timing issues with the software. It could also be the case that the software just doesn't work with something as "new" as a P166MMX...)
 
The TMS2716 are weird devices anyhow (triple voltage rails).

The TMS2532A device may also use a weird programming algorithm.

 Dave
 
MiaM said:
Related tangent:
Does this warning apply to all of these old 21V EPROMs, or is it just that the 2532A are known to cause grief?
I'm asking because I had trouble programming a pair of 2732(A) EPROMs (can't remember if they were A or not) a while ago with my ALL-03 and I had to restart the programming so it would program each 256 byte block twice, as the checksum failed the first time, seemingly due to a too short programming time.

I didn't investigate this further at the time, I.E. I can't remember for example measuring anything using an oscilloscope or whatnot while programming, so I don't know if the programming voltage and the length of the programming pulse is/was correct, but still. (A variable here is that the software is really intended for running in DOS but due to reasons I'm running it under Windows XP with a third party add-on driver thing that opens up a hole allowing the DOS VDM to access the I/O ports of the ALL-03 ISA card, so there could be timing issues with the software. It could also be the case that the software just doesn't work with something as "new" as a P166MMX...)
Click to expand...
I am not sure about other 21V programming parts, it is just the 2532A that appears to have given a number of people, including me, some grief. It is possible that 2732A parts might be "easy" I have never tried these in the GQ-4x or my BP1400. But, in the case of the PET, & repairs, initially at least I think it is better to use the TMS2532 parts off the bat, and avoid adapters and avoid 2732's, because many (not all) adapters have pins of such an over-sized geometry that when they are used in an IC socket, they ruin it for a standard IC pin geometry later and make the connections unreliable when a real IC is put in the socket at a later date. It is better to minimize all variables in computer repairs or you can get into a run in circles mad world situation.
 
I have never had any problems with non TMS 2732A parts - but I don't use them in PETs of course.

But I use 'professional' EPROM programmers that I inherited from work when they were throwing them out!

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