Cartridge port adding RW...

[exxos]

There was a brief talk over on AA last year about replacing the cartridge port with a super I/O port. There was of course mixed views about this.

Currently I am thinking the major problem with the port is lack of RW signal. So here is my proposed solution.

Pardon the crude drawing..

CRW.JPG

Basically as there is 3 GND connections on the cartridge, the idea is to remove one of them use with a RW line instead. We would keep the existing 2 GND connections on the top and bottom of the port just remove the inner GND connection.

The problem is that we cannot simply route RW to the cartridge port, because any existing cartridge would short out the RW line, but we can actually solve this easily.

Basically we have a pull-up on the cartridge side. This will always pull-up the pin to 5 Volts. So what happens when a original cartridge is connected, that pin will be of course shorted down to GND. The resistor will limit the current to maybe 1mA or whatever. So there would be no problem to use existing cartridges.

Should a cartridge be used which now uses this RW signal, when the RW CPU goes high, the pull-up resistor will pull up the pin on the cartridge to 5Volts also. When the CPU sets RW low, the diode will pull down the cartridge side to GND.

Diode will isolate the CPU side RW from being shorted out when a older cartridge is used.

[Icky]

Very simple idea.

Makes you wonder why Atari didn’t put a R/W line on the ROM port in the first place.

[stephen_usher]

Very simple idea.

Makes you wonder why Atari didn’t put a R/W line on the ROM port in the first place.

An interrupt line would have been nice too.

[Icky]

Have been thinking of making this mod on the STF Remake and use pin 38 as it seems to be the most logical one to use.

Enhanced Atari ST Cart Port.png

[Smonson]

There was a brief talk over on AA last year about replacing the cartridge port with a super I/O port. There was of course mixed views about this.

Currently I am thinking the major problem with the port is lack of RW signal. So here is my proposed solution.

Pardon the crude drawing..

I might be misunderstanding something here, but that diode is only going to pull it down to about 0.7 volts minimum - if you want to interface that with some TTL ROMs, isn't that cutting it a bit fine?

[exxos]

I might be misunderstanding something here, but that diode is only going to pull it down to about 0.7 volts minimum - if you want to interface that with some TTL ROMs, isn't that cutting it a bit fine?

Yes and no.. It depends on the spec of the diode. Some have drops as little as 0.2v. It also depends on current. A diode may well have 0.7v drop, but at what current ? May well be 1amp. If we assuming 1k resistor on 5v current that's just 5ma. Given the same diode its voltage drop may be just 0.4v.

[Smonson]

Thanks for the explanation! This is why I'm not an analogue guy... :)

[Icky]

I might be misunderstanding something here, but that diode is only going to pull it down to about 0.7 volts minimum - if you want to interface that with some TTL ROMs, isn't that cutting it a bit fine?

Yes and no.. It depends on the spec of the diode. Some have drops as little as 0.2v. It also depends on current. A diode may well have 0.7v drop, but at what current ? May well be 1amp. If we assuming 1k resistor on 5v current that's just 5ma. Given the same diode its voltage drop may be just 0.4v.

Not sure if the calculations off the top of my head are wrong but I was thinking of a 4.7k resistor?

[exxos]

Not sure if the calculations off the top of my head are wrong but I was thinking of a 4.7k resistor?

1K, 4.7K, 10K value not really much important.