Repurposing a Super Expander (or 3K RAM)

[chysn]

There's something I'd like to do, but I'm not sure whether it's easy, slightly-involved, or plain impossible.

I have a VICMon cartridge and a Super Expander cartridge.

My end goal is to have a VICMon cartridge with 3K of extra RAM.

I'd like to remove the Super Expander ROM from its board, add a socket, and then replace it with the VICMon ROM. But I know that there'd be at least one extra step, because Super Expander lives in block 5, and VICMon runs from block 3. So I'd need to tell the board (possibly with the jumper points on the board) to locate the ROM at block 3.

Has anyone here done anything like this?

My initial, perhaps naïve, guess is that I should cut the jumper labeled "5" at pin 13 and put a dab of solder across the "3" jumper at pin 12.

[Mike]

Hi chysn,

I'd like to remove the Super Expander ROM from its board, add a socket, and then replace it with the VICMon ROM. […] Has anyone here done anything like this?

The thread "Modding the 3K RAM or Super Expander cartridge" might be of peculiar interest to you.

[...] My initial, perhaps naïve, guess is that I should cut the jumper labeled "5" at pin 13 and put a dab of solder across the "3" jumper at pin 12. [...]

I'd suggest a slight change of plans (see photos in aforementioned thread):

1. There's a second place for a 24-pin (E)PROM available on the PCB of Super Expander - BTW, +3K RAM and SE share the same PCB!
2. The pair of jumpers labeled "5" and "2" go to the upper position and provide either /BLK5 or /BLK2 to /CS of that (E)PROM. "5" is default.
3. The pair of jumpers labeled "3" and "1" go to the lower position and provide either /BLK3 or /BLK1 to /CS. "1" is default.

You can actually extract the PROM of VICMON, put it into the lower place of SE, add a 100 nF stabilising capacitor, open the "1" jumper and close the "3" jumper. Then you have Super Expander and VICMON combined in one cartridge! If you just want to use VICMON on its own, you can temporarily disable the BASIC extension of Super Expander with SYS64818 until the next reset or power cycle.

One thing you should reconsider is the use of a socket: you'll either need a (slightly more expensive) low-profile socket, or solder the VICMON PROM directly on the PCB, as there's not much room above it. With a normal height socket, the PROM will likely 'collide' with the cartridge shell.

Another thing to be wary of - VICMON changes some bytes in the zeropage normally allocated to the BASIC interpreter, even when the so-called "virtual zeropage" is used.

Greetings,

Michael

[eslapion]

I'd like to remove the Super Expander ROM from its board, add a socket, and then replace it with the VICMon ROM. But I know that there'd be at least one extra step, because Super Expander lives in block 5, and VICMon runs from block 3. So I'd need to tell the board (possibly with the jumper points on the board) to locate the ROM at block 3.

Has anyone here done anything like this?

My initial, perhaps naïve, guess is that I should cut the jumper labeled "5" at pin 13 and put a dab of solder across the "3" jumper at pin 12.

I did a lot more than this with the Ultimate Expander I was offering between 2006 and 2010.

Actually, Super Expander uses the lower half of BLK5 ($A000-$AFFF) and there is a version of VICMON which resides in the upper half of BLK5 ($B000-$BFFF) so you can easily have both SE and VICMON at the same time in the same cart.

See: http://www.zimmers.net/anonftp/pub/cbm/ ... index.html

If you choose to have the standard BLK3 version of VICMON, know VICMON uses only the lower half of BLK3 and you could combine it with Programmer's Aid which uses the top half of BLK3. With 2 (EP)ROMs, you could have SE, VICMON and PA all in a single cart.

[Mike]

I did a lot more than this with the Ultimate Expander I was offering between 2006 and 2010. [...]

Of course, when you put more resources at the job - like burning a custom EPROM or even making a new PCB - you can produce a fancy expansion cartridge much to your liking, with whatever combination of tools (in ROM) and extra RAM you want.

However, I think the magic words in the OP were: "Repurposing a Super Expander" with VICMON. As I pointed out, it's possible to combine the original PROMs of SE and VICMON on the original PCB of SE with minimal changes, namely opening and closing two jumpers and adding another stabilizing capacitor.

[MCes]

Hacking a PCB game.... BLK 3 and 5 (no RAM onboard: only ROM access was possible with that PCB...)
march 2015:

[chysn]

I'd suggest a slight change of plans (see photos in aforementioned thread):

1. There's a second place for a 24-pin (E)PROM available on the PCB of Super Expander - BTW, +3K RAM and SE share the same PCB!
2. The pair of jumpers labeled "5" and "2" go to the upper position and provide either /BLK5 or /BLK2 to /CS of that (E)PROM. "5" is default.
3. The pair of jumpers labeled "3" and "1" go to the lower position and provide either /BLK3 or /BLK1 to /CS. "1" is default.

You can actually extract the PROM of VICMON, put it into the lower place of SE, add a 100 nF stabilising capacitor, open the "1" jumper and close the "3" jumper. Then you have Super Expander and VICMON combined in one cartridge! If you just want to use VICMON on its own, you can temporarily disable the BASIC extension of Super Expander with SYS64818 until the next reset or power cycle.

Ah! Great information! It might be fun to have a VICMon/Super Expander. But apparently, I could also simply add VICMon to my 3K RAM cartridge. The downside with that is that my 3K RAM cartridge is kind of a sentimental keepsake, and I'd rather risk the life of the Super Expander.

One thing you should reconsider is the use of a socket: you'll either need a (slightly more expensive) low-profile socket, or solder the VICMON PROM directly on the PCB, as there's not much room above it. With a normal height socket, the PROM will likely 'collide' with the cartridge shell.

Fair enough, I wondered about that.

Thank you so much!

[chysn]

Another question...

Can I replace the Super Expander's PROM with 8K RAM and keep it in block 5, and have 8K at $A000 that will be isolated from BASIC? Or does that PCB only support ROM?

And if it does support RAM, will I run afoul of any auto-start at $A000, which will now be pointed at a bunch of undefined code in RAM?

[eslapion]

Another question...

Can I replace the Super Expander's PROM with 8K RAM and keep it in block 5, and have 8K at $A000 that will be isolated from BASIC? Or does that PCB only support ROM?

And if it does support RAM, will I run afoul of any auto-start at $A000, which will now be pointed at a bunch of undefined code in RAM?

The ROM footprint only supports ROM but the cartridge itself can easily be modified to do what you suggest by comparing the pinout of a 2332/2364 ROM with that of a standard 8k x8 static RAM IC and adding a few wires. The one critical connection missing from the ROM footprints is the R/W line which is connected to all 6 SRAM chips found in the SE cart.

You could even go for a 32k x 8 SRAM IC and turn your SE cart into a 35k expansion. Using the Super Expander with more than 3k is really nice.

[chysn]

The ROM footprint only supports ROM but the cartridge itself can easily be modified to do what you suggest by comparing the pinout of a 2332/2364 ROM with that of a standard 8k x8 static RAM IC and adding a few wires. The one critical connection missing from the ROM footprints is the R/W line which is connected to all 6 SRAM chips found in the SE cart.

Thanks! I think it's best that I start with VICMon+3K as an entry-level project, and put the +8K idea on the "maybe someday" back burner.

[Mike]

Can I replace the Super Expander's PROM with 8K RAM and keep it in block 5, and have 8K at $A000 that will be isolated from BASIC? Or does that PCB only support ROM?

As eslapion wrote, that's entirely possible with the Super Expander PCB by thoughtful cuts and rewiring in the lines of what I posted in the other thread I linked to above.

And if it does support RAM, will I run afoul of any auto-start at $A000, which will now be pointed at a bunch of undefined code in RAM?

The KERNAL also checks for the A0CBM signature, which must be present at $A004 to indicate the validity of cartridge reset and NMI vector at $A000 and $A002, respectively.

It is very unlikely, that a SRAM will show that signature by itself on a fresh power-up - in many cases, they present an alternating $00/$FF pattern - however there are documented cases, where trickle power provided by the floppy drive (over the serial bus) kept up the contents of a soft-loaded cartridge image, which then restarted upon a power cycle. Eslapion knows more details about this.

I think it's best that I start with VICMon+3K as an entry-level project, [...]

Great! A follow-up post with a photo of the modified PCB will be much appreciated.

[eslapion]

It is very unlikely, that a SRAM will show that signature by itself on a fresh power-up - in many cases, they present an alternating $00/$FF pattern - however there are documented cases, where trickle power provided by the floppy drive (over the serial bus) kept up the contents of a soft-loaded cartridge image, which then restarted upon a power cycle. Eslapion knows more details about this.

Today, looking on Digikey, all small size static RAM chips you can find (8kx8 or 32kx8) that are still being made include a micropower mode which will keep the content intact with only a trickle amount of power.

If you load the image of a ROM cartridge which includes the A0CBM signature into RAM visible at BLK5 and you turn off your VIC-20 with a 1541 disk drive still attached and turned on, the minuscule amount of power required to prevent the reset line from going low will be enough to keep this image in RAM intact. When you turn on your VIC-20, the program in BLK5 will start again.

If you have a 32k or more RAM cart, the whole content will be intact.

[MCes]

Mmmm....

What do you think about a "RAM autostart KEY killer" circuit?
It's a theoretical schematic...

The first valid access to BLK5 is done by kernal for reading the autostart key:
this circuit transform the first acces from read to write and the 8 resistors store "$FF" inside the RAM corrupting the key,
at the same time CPU read $FF also skipping the autostart....

The RCD cell on 74LS74 "R" clear the ff at power-up, the "D" of ff prevent event as: during the HW RESET the address is pointing inside BLK5.

[eslapion]

What do you think about a "RAM autostart KEY killer" circuit?
It's a theoretical schematic...

The idea is good. The small reset RC filter requires the addition of Schmitt trigger to square up the signal.

You could also have a similar RC filter with the output inverted and use an OR gate to BLK5. SO /CS = /RCF||/BLK5. This way all access to BLK5 is disabled for about the 1st second the VIC-20 is on and you kept all data in BLK5 intact without auto starting it and you have a simpler circuit.

[chysn]

I think it's best that I start with VICMon+3K as an entry-level project, [...]

Great! A follow-up post with a photo of the modified PCB will be much appreciated.

I certainly will post photos. I'm hoping I'll have time to start and finish the first leg of the project on Sunday. I'm just waiting on some 104 caps. Might as well start with new caps.

I'm calling an audible here; I've decided to use my 3K RAM PCB as the VICMon host instead of the Super Expander PCB. Why? Because there's like a 100% chance that I'll want to continue with the +8K project in the bank 5 slot. If I do it this way, I won't have to waste a Super Expander PROM down the road.

[Mike]

I've decided to use my 3K RAM PCB as the VICMon host [...] I'll want to continue with the +8K project in the bank 5 slot. If I do it this way, I won't have to waste a Super Expander PROM down the road.

... and you can soft-load a ROM dump of SE into the RAM in BLK5 indeed.

BTW, what (mass-)storage solution do you use? (Still) a CBM floppy drive or rather a SD2IEC?