Hi everyone,
I recently became the owner of an Apple IIc...
Even before I got the computer, I already had a "Checkmate MultiRAM C" memory expansion lying around. I also had the matching RAM chips on hand to fully populate the card.
Since I want to use an SP2SD with the machine, I swapped the original "255" ROM for a "ROM4x" and modified solder jumpers W1 and W2 accordingly.
I wanted to check if the RAM expansion was actually working, so I first asked the AI Grok how to go about it...
The answer and test procedure: First, enter the Monitor using `call-151` and run the following test:
C073:00 ; IIc internal Aux bank (Bank 0)C003C005C00900:1100 ; should show 11 => OK
C073:01C003C005C00900:2200 ; should show 22 => OK
C073:00C003C005C00900 ; should show 11 again (not 22!) => shows 22.
I don't put too much stock in what the AI says, though, so I thought I'd ask here: is this test supposed to work this way, or is there something else I should try?
OK, unfortunately I can't edit my post, so here is the test again, a bit more readable.
C073:00 ; IIc internal Aux bank (Bank 0)
C003
C005
C009
00:11
00 ; should show 11 => OK
C073:01
C003
C005
C009
00:22
00 ; should show 22 => OK
C073:00
C003
C005
C009
00 ; should show 11 again (not 22!) => shows 22.
You could always boot up a copy of locksnmith 6.o and run it's memory test.
It usually finds all the memory expansion available.
It finds and often uses but does not test any auxiliary RAM. So on any standard Apple //c half of the built-in RAM cannot be tested with Locksmith. In addition, RAM could be defective to such an extent that no software will be able to 'boot'. As far as a Checkmate MultiRAM is installed in OP's //c, it follows the RamWorks standard, therefore it also cannot be tested with Locksmith.
It's easy to see the motivation behind those commands, they just won't work -- and it's just not practical to fix them. (Blunder #1 was using "read" instructions for RAMRD, RAMWRT, and ALTZP, which all require "write" instructions. Blunder #2 would have been activating RAMRD from the middle of a Monitor command string, which would prevent the Monitor from parsing the rest of the command. Blunder #3 would have been switching zeropage memory via ALTZP which would cause the Monitor to crash, but also disable input and output so the user won't see how it crashed. Luckily, Blunder #1 prevented Blunder #2 and #3 from causing a crash.)
Try this simple machine-code program that counts the AUX memory banks and lists them on the screen:
AUX.BANK_.COUNT_.png
Here's an annotated disassembly of that program, explaining how it works. This program uses $C071 instead of $C073 as the bank-selection port because it's supported by the widest variety of third-party AUX memory cards, including MultiRAM, RAMWorks, Neptune, Super Expander, Z-RAM, and more.
AUX.BANK_.COUNTER.png
On your Apple //c, bank 00 is the motherboard's AUX memory. If you have RAM installed in your Checkmate MultiRAM and it's working correctly then it should print additional bank numbers after 00. (Note: the numbers might not be consecutive depending on the model of MultiRAM.)
@S.Elliott OK, I tried your machine code program (I think it was correct to type in call-151 and then type the program in ?), but it only shows me 00 for the internal Aux Memory. What could be the problem with the MultiRAM Memory expansion ? I allready checked all 74LS and 74HC Chips with my T48-Programmer... This Chips should be fine. The Part I cant test is the PAL16L8 Chip...
There was a bug in the hex listing, in the first byte: it should be A0 (LDY...), not A9 (LDA...).
The disassembly is the correct version.
OK, I think that did the trick !
After changing A9 to A0 at the beginning of the first line, the program shows me 00 01 02 03 04 05 06 07 08
I think that means the RAM Expansion is working fine ?
If you don't know the answer to your question that means you spent money on something you don't need, and will not be able to use, and have no knowledge about it.
It shows that the bank-decoding hardware for 8 additional memory banks (8 times 64 KB = 512 KB) is working, although the DRAM comprising those banks is not actually tested. Only the byte at $06 in each bank is written and then verified.
The MultiRAM C came with a disk containing a memory test program, in addition to two RAM Disk drivers and an AppleWorks patcher.
This might be the floppy: checkmate_multiram.zip