Hopefully, an easy question

[John Kuhn]

All,

My PAL-1 computer PCB seems to be working well. It's been a lot of fun to build and program.

When I purchased my PAL-1, I also purchased the 32K RAM expansion module, along with the ROM expansion module and the motherboard. I just got around to building these RAM and ROM expansion modules and the motherboard. I can't seem to run BASIC, Forth, or Woz monitor out of bank 0, and nothing seems to work out of bank 1. Here's what I get when I try running out of Bank 0:

A000 4C G

KIM

A000 4C

When I check a few ROM locations, I get the following:

A000 4C

A001 0F

A002 A0

A003 4C

A004 32

A005 A0

A006 4C

This seems like it might be correct from what I read in the information on the ROM. The switches on the ROM PCB are set to 28C256, and A11, A12, and A13 are set just like the images in the ROM manual (A13 = up, A12 and A11 = down).

I also thought that I would try changing locations in RAM. I loaded FF into $2000, and also into $9FFF and checked the locations. It looks like it saved it in both locations:

2000 00 FF.

2001 00 2000

2000 FF 9FFF

9FFF 00 FF.

9FFF FF

It look at first glance that RAM is storing information. Next step is to write a program to try and program each location and test it.

Any thoughts or ideas on where to look would be appreciated. Could I have a corrupted 28C256?

John 

[Jim McClanahan]

You've done everything I would have suggested. The next step I'd suggest is a RAM test. If you have a short between two address pins or something of that sort, it could look fine just looking at a few random locations.

I doubt the ROM is corrupt. You are seeing what you should with it.

You might also try changing the card location. I remember having a similar issue when I was using a protoboard. Nobody seems to have had trouble with the production boards though.

There is a MOS hex file of the widely used memory test on my GitHub page:

https://github.com/w4jbm/PAL-1-6502-SBC

It gets a bit trickier if that isn't it.

Thanks,

Jim W4JBM 

--
You received this message because you are subscribed to the Google Groups "PAL 6502 computer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pal6502+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/c6b2db23-5d1b-48d8-8dfa-fa48d1098064n%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[GN L]

Is the Memory Selector jumper on PAL-1 open?

> On Jul 6, 2022, at 7:30 AM, John Kuhn <firs...@gmail.com> wrote:
>
> lding these RAM and ROM expansion modules and the motherboard. I can't seem to run BASIC, F

Best,
Liu

[John Kuhn]

Jim,

Thanks for the help. The RAM test listing is very helpful. I'll let you know how it turns out.

Liu,

the memory selector is open.

Thanks,

John

[John Kuhn]

All,

Here is the latest update:

I removed the ROM Expansion board, and just focused on the RAM expansion board. I ran Jim's memory test (starting at $0000) using the onboard keypad/display  (I/O keypad jumper=open, memory selector=open). As part of the test, I checked the memory page starting at $02 and ending at $03, and it worked. I tested again with the memory page starting at $20, and ending at $9F. By examining the contents of memory, all worked as expected. I then tried typing it in using the serial port (I/O keypad jumper=closed, memory selector=open). To my surprise, the memory test ran (I got the KIM prompt), but did not change the contents of memory as expected. 

Next, I tried a simple program (starting at $0002) to load AA into $2000 using the onboard keypad/display (I/O keypad jumper=open, memory selector=open). The program ran fine, and the contents of $2000 changed to AA. The program is

A9 00

8D 00 20

4C 4F 1C

The DEN line is also toggling when the simple program runs, and when I manually change locations at $2000 - $9FFF as expected.

I tried the same program using the serial port. The program runs, I get the KIM prompt, but $2000 does not get loaded with AA. I check the program at locations at $0002- $0009, and the program is properly loaded. When I ran the program, I monitored the DEN line to see if the 6264 is being addressed, and in this case, it is not toggling as expected. When I manually read and write locations from $2000 - $9FFF through the serial port, the memory contents change, and I see DEN toggle as expected. 

Any thoughts? The simple program works fine with the I/O keypad jumper=open, and doesn't work with it closed. DEN toggles as expected for both configurations when manually changing or addressing RAM from $2000 - $9FFF, works when running the program with the onboard keypad/display, but doesn't work when you run the program using the serial port. The only program difference is that when I use the onboard keypad/display, I end the program with 00. When I use the serial port, I end with 4C 4F 1C. 

John

[Jim McClanahan]

When you run the memory test from the terminal, what is the address the monitor shows when you return to it?

My gut feel is a shorted address line and that the memory test isn't making it through all 32K. You have to be sure that when you write to an address, one and only one memory location changes.

Thanks,

Jim W4JBM

--
You received this message because you are subscribed to the Google Groups "PAL 6502 computer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pal6502+u...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/c43211b0-bd0a-4aad-8605-e9b93c250db8n%40googlegroups.com.

[Jim McClanahan]

On other thing you might want to try is removing the expansion board and running the memory board only with a direct connection to the PAL. If it is some kind stray noise causing your problems, that is likely to eliminate it.

Thanks,

Jim W4JBM

[John Kuhn]

All, 

So far, no luck. Here is what I tried today:

• Measured for a short between any address and data lines. This included each pin to every other pin. No shorts.
• Measured for a short between any adjacent pins on the expansion boards connector. No shorts. 
• Measured connections from every pin to every other pin on the same line.This means A4 on the 6502 to A4 on 6530 to A4 on 6264 to all other chips that were connected to A4. I did this for all address lines, data lines, IRQ, K0 to K7, etc. Every line was normal. I measured directly on each pin, to make sure the measurement took into account any socket issues.
• I removed the expansion board and ran the memory board with a direct connection to the PAL via the ribbon connector. No luck.

Jim - If I load the memory test from terminal, and then run at 0002, I get the KIM prompt, and the address the monitor is back at 0002.

Any thoughts?

John

[Jim McClanahan]

The memory test should jump back to the monitor showing the address of the last address tested plus one. The fact that you are starting and ending at $0002 as the monitor address is strange.

I will have to look at the schematic a bit. I don't think the memory card is the problem. I am wondering if something in the single step portion could flake out and be causing a reset when you jump to code outside of ROM. You might try loading some simple code (LDAs and NOPs for example) and see if turning on single step works like expected. (You will have to set the interupt pointers in RAM.)

I want to look at what might force it into single step mode besides the single step switch.

Thanks,

Jim W4JBM 

--
You received this message because you are subscribed to a topic in the Google Groups "PAL 6502 computer" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pal6502/r3fuw1duvdc/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pal6502+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pal6502/e8a2cbf6-49de-467f-a257-c13d74a59ce4n%40googlegroups.com.