• Please review our updated Terms and Rules here

Pet 4016 Transformer Question

Yes, an excellent breakfast (as usual). I stay here a lot!

If you got my PETTESTER from my Google drive there is a fully-commented copy of the source code on there explaining how to do it...

Use the online assembler, load the source file into the assembler and copy the relevant table for the 40 column CRTC initialisation into the correct place (following my comments), reassemble and download the resulting HEX file. No patching required.

Dave
Ok so I was able to finally figure out where to make the changes in the .a65 file.

But

Being a Mac user, everytime I attempt to export the compiled file to a .BIN, they are all blank (filled with 00's) - I've used versions online and command line programs as well as loaded some into my VMware OS version of Windows 7 and Win XP...

Since it almost 4:30am here.... gonna call it a night seeing as I cannot get the bin file to compile correctly. I will try again later today...

Hopefully once I get the corrected BIN, I will have more information :)
 
I think the highest priority at the moment is what daver said:

"You need to 'divide and conquer'. You could have a fault on the PET main logic board or the monitor. Disconnect the monitor and 'scope the video and horizontal & vertical drive signals on the monitor connector of the PET logic board. The H and V drive signals need to be correct before you connect the monitor."

Use the current version of PETTEST V4. Look for the vertical timing signal of a skinny negative pulse every 60 Hz, and a horizontal signal of 20 KHz on the J7 connector. Let's make sure we can initialize the CRTC.
 
I am not sure what the Mac has to do with anything? I have developed all of this software on a Mac.

Go to asm80.com.

Create a new file within asm80.com called PETTEST.a65 and tell the assembler that it is for a 6502 CPU.

Copy and paste the contents of your modified .a65 file into the new file within asm80.com.

Hit Save.

Hit compile.

Check for any errors.

Download the HEX file from asm80.com to your computer.

If your EPROM programmer can handle HEX files directly, load that into your EPROM programmer (not forgetting to give it an offset of $E000).

Program your EPROM device.

Dave
 
As a first order of business, reassembling PETTEST V4 for the 40 column constants is not really needed. I think it only makes the line spacing look better and may eliminate losing a little of the top of the first line. The program will display well enough as is.

What is needed is to see if the H and V timing signals are at the correct frequency. If not, the CRTC may still not be initialized correctly due to bad kernal ROM, fault on bus, etc or maybe the 6545 is kaput.

If the timing is correct, there may be an issue with the video board.
 
As a first order of business, reassembling PETTEST V4 for the 40 column constants is not really needed. I think it only makes the line spacing look better and may eliminate losing a little of the top of the first line. The program will display well enough as is.

What is needed is to see if the H and V timing signals are at the correct frequency. If not, the CRTC may still not be initialized correctly due to bad kernal ROM, fault on bus, etc or maybe the 6545 is kaput.

If the timing is correct, there may be an issue with the video board.
Ok makes sense ....

I have gone back to square 1 and conducted all the tests and measurements again from the beginning of this thread.

I have a few differences and some that are still the same ....

1) the 6502 doesn't seem to be executing commands, the timing is still good, the reset functions as expected on the chip, and all other pins as measured (see page one for the pins to measure) - pin 7 seems to "not be pulsing"

2) I don't get anything on the video pin.... that just shows voltage and nothing more

3) the vertical sync doesn't exist. There isn't anything on both the output pin at J7 as well as on UB2

4) horz sync shows "something" but doesn't match the diag images from zimmer.net

None of the measurements that were depicted on zimmer.net for the Video / vert / horz don't match ...

Soooooo I'm not sure how far back I need to try to trace to see what the root of the issue(s) are.

All the voltages on the rails are still within spec.
 
On start-up you should get some activity of /SEL F and /SEL E signals. Study your scope and learn how to set it up for 'single trigger' mode. Sync the scope to Channel 1 and connect channel 1 to /SEL F. Connect Channel 2 to /SEL E. Use your PET reset button to start the test. See if you capture pulses on both scope channels.

The SELECT signals can be found on sheet 1 on UE12 (74154) which is a 24 pin chip. /SEL F is on pin UE12-pin17 and /SEL E is on UE12 -pin16.
/SEL F selects the kernal ROM and /SEL E selects the EDIT EPROM that holds the PETTEST V4 code.
-dave_m
 
Ok let me get onto that. The scope does have single trigger options so I can mess around with that :)

Thank you for the suggestions and assistance. It's 1:30am here so I'm off to rest but have a day of travel tomorrow to Utah and then back. So it will be Saturday before I can get you all an update :)
 
but have a day of travel tomorrow to Utah and then back. So it will be Saturday before I can get you all an update :)
No rush, the Commodore PET is only a hobby for all of us. To keep your enthusiasm, it is best to work on it only when you have sufficient open time to enjoy it. Your PET will always be there for you. Well, anyway as long as the better half allows the messy work table.
 
No rush, the Commodore PET is only a hobby for all of us. To keep your enthusiasm, it is best to work on it only when you have sufficient open time to enjoy it. Your PET will always be there for you. Well, anyway as long as the better half allows the messy work table.
Hahaha good thing I'm single then 🤣🤣🤣
I will actually have time tonight and tomorrow for sure to get some more troubleshooting done
 
On start-up you should get some activity of /SEL F and /SEL E signals. Study your scope and learn how to set it up for 'single trigger' mode. Sync the scope to Channel 1 and connect channel 1 to /SEL F. Connect Channel 2 to /SEL E. Use your PET reset button to start the test. See if you capture pulses on both scope channels.

The SELECT signals can be found on sheet 1 on UE12 (74154) which is a 24 pin chip. /SEL F is on pin UE12-pin17 and /SEL E is on UE12 -pin16.
/SEL F selects the kernal ROM and /SEL E selects the EDIT EPROM that holds the PETTEST V4 code.
-dave_m
Good afternoon/evening....

OK sad news... the scope I have is not able to be set to do a "single trigger" - it has a lot of entry level features but that isnt one of them... I have tried all sorts of different settings with zero success....

The Logic probe I have, I have placed on Pin 16 and 17 of UE12.... Pin 16 stays HIGH, while Pin 17 stays LOW....

With the scope I have, I am unable to detect or meansure any pulse or activity that is clean.... in testing the probes, they are calibrated properly and read accordingly during the calib...

So, I can either deduct that the SN74154N is bad (aka UE12) or I just don't have the proper test hardware to get a proper reading....

As for the EPROM, the burner I have is a GQ-4x4.... its able to burn chips, but the software that the company has produced works in a windows enviroment.... everytime I would attempt to run this on my mac using either Parallels or VMWare images (WIN 7 or XP) - the burner software would crash. I was able to duplicate this on a straight XP Windows PC (seperate laptop) - the company replied to my trouble request, and stated the internet needed to be disconnected from the laptop before starting the software....

Now, it works, but in order to get the EPROM image to load using their software, I have to load the image first, then set the "offset' before selecting a chip to burn.... otherwise, their software states the "image is too big to fit on the chip" and then truncates the software (which I figured out the .bin files and .hex files were starting at the address that Dave mentioned above at $E000, but this software for the burner defaults to address 000000 and was deleting anything beyond what it thought the EPROM could store... I have not yet perfected the use of this software yet, but I have a few things to try)

But since the PET doesn't seem to be talking to any of the EPROMs, burning the PETTEST isn't going to get me very far.... I will need to see what else might be going on .... I can obtain another SN74154N chip from a few different suppliers...

so......... here I am LOL

Thank you guys for being patient with me on this one....

Rich
 
I have a Hantek DSO5102P. I've spent a hot minute with the documentation but I must be missing something.
 
Now, it works, but in order to get the EPROM image to load using their software, I have to load the image first, then set the "offset' before selecting a chip to burn....
I don't think a transfer/offset address is needed in the case of a simple PROM programmer when you are trying to program a 2K binary into a 2K EPROM. If you must use the Intel hex format rather than binary, create it using a transfer address of zero.
 
Wow, this is quite a scope.

As daver2 says, look for:
6.4 Example 4: Capturing Single-shot Signal

You will have to keep the PET Reset pushbutton down and then let it go as you hit the 'Single SEQ' button on the scope. And hope the pushbutton doesn't bounce too much...
 
Ok that makes sense.

I got this one on the cheap (Aka Amazon) and then was able to hack it from 100mhz to 200 without any issues. Lots of forums and YouTube videos on it. Yet the documentation isn't always clear to me. Lol

Anyway let me give that a shot to see what I will get.
 
That manual is pretty useless isn’t it! It was a good job they put that example in there...

Assuming you are not seeing any /SEL F activity as normal, then setting a falling edge trigger on /SEL F this should trigger (when the oscilloscope is set to single trigger) on the first /SEL F event after a RESET (if one is going to occur).

Set the oscilloscope to single trigger and RESET the PET.

Once the trigger has occurred, it will capture the signals on the two channels for one ‘sweep’. You will have to adjust the timebase to get a number of pulses onto the display (assuming you are going to get any of course).

If the oscilloscope triggers, you have probably got at least one pulse.

Dave
 
Haven't forgot about ya.... had to leave for a 10 days business trip. Upon my return to the US on May 17 I will get an update to see where my Pet stands with the snap shot images from the Scope.
 
Back
Top