Casio fx-8500G weird behavior with FA-80 interface

[Dave Britten]

This is a long-shot, but I'm curious if anyone here has ever seen this. I've got an fx-8500G and a pair of FA-80 cassette/printer interfaces. When I connect the calculator to the interface and power it up, it suddenly acts like it's an fx-7000G with slightly more RAM - none of the fx-8000/8500 features seem to be present. Pressing MODE 0 to enter the text editor just types a 0 on the "home" screen, and MODE Ans to recall the input buffer just types Ans. Also, most of the RAM is reportedly missing, with only around 1500 bytes available according to the program write mode screen. And of course, none of the cassette I/O commands are recognized.

When I take the calculator off the interface and switch it on, everything is back to normal. The editor, all the RAM, and the fx-8000/8500 commands and features are back.

I also have an fx-8000G which works fine with these interfaces (as far as cassette I/O is concerned - I don't have a parallel printer to test with). Anybody ever run into this before? I opened up the 8500 and reflowed the solder on the I/O connector, and replaced the memory-backup super capacitor, on the off chance that it had enough current leakage that some part of the calculator wasn't getting enough power when the I/O port was connected.

Note that I do not have an oscilloscope or logic analyzer to get really fancy with troubleshooting.

[blackjetrock]

(05-02-2022 07:28 PM)Dave Britten Wrote:  This is a long-shot, but I'm curious if anyone here has ever seen this. I've got an fx-8500G and a pair of FA-80 cassette/printer interfaces. When I connect the calculator to the interface and power it up, it suddenly acts like it's an fx-7000G with slightly more RAM - none of the fx-8000/8500 features seem to be present. Pressing MODE 0 to enter the text editor just types a 0 on the "home" screen, and MODE Ans to recall the input buffer just types Ans. Also, most of the RAM is reportedly missing, with only around 1500 bytes available according to the program write mode screen. And of course, none of the cassette I/O commands are recognized.

When I take the calculator off the interface and switch it on, everything is back to normal. The editor, all the RAM, and the fx-8000/8500 commands and features are back.

I also have an fx-8000G which works fine with these interfaces (as far as cassette I/O is concerned - I don't have a parallel printer to test with). Anybody ever run into this before? I opened up the 8500 and reflowed the solder on the I/O connector, and replaced the memory-backup super capacitor, on the off chance that it had enough current leakage that some part of the calculator wasn't getting enough power when the I/O port was connected.

Note that I do not have an oscilloscope or logic analyzer to get really fancy with troubleshooting.

Hmm, odd. Presumably this happens on both FA-80s?

The circuit for the FA-80 is here:

http://www.pisi.com.pl/piotr433/fa80.png

I don't see anything obvious that would cause this behaviour. You could try disconnecting some of the control signals on the connector and see if that affects the behaviour. Are all pins and sockets on the I/O connector on the calculator OK?

Andrew

[Dave Britten]

(05-03-2022 11:01 AM)blackjetrock Wrote:  Hmm, odd. Presumably this happens on both FA-80s?

The circuit for the FA-80 is here:

http://www.pisi.com.pl/piotr433/fa80.png

I don't see anything obvious that would cause this behaviour. You could try disconnecting some of the control signals on the connector and see if that affects the behaviour. Are all pins and sockets on the I/O connector on the calculator OK?

Andrew

Yeah, exact same behavior with both FA-80s, which work fine with the fx-8000G. I haven't done any extensive continuity testing and tracing to see if there are any dead-ends on the I/O connector pins, but I did reflow all the solder where the port connects to the board (which didn't change anything). It's super weird, it's like a good chunk of the RAM and ROM are just not there anymore, and as soon as you disconnect the interface, the calculator acts normal again.

[Dave Britten]

I finally got ahold of another fx-8500G, and this one is working fine with the FA-80. So there's definitely something specific to the other one preventing it from working properly with the interface. Still no clue what, though! But at least I have one that's fully working now.

[blackjetrock]

(05-06-2023 09:25 PM)Dave Britten Wrote:  I finally got ahold of another fx-8500G, and this one is working fine with the FA-80. So there's definitely something specific to the other one preventing it from working properly with the interface. Still no clue what, though! But at least I have one that's fully working now.

If you fancy a project, then on this page:

http://pisi.com.pl/piotr433/fx8000xe.htm

are details of how to dump the ROM of the FX8000P and also the FX8500P. You could dump the FX8500P ROM and see if this is a feature that is in the code somewhere. The ROM for the FX8000P is also on that site so it's maybe not such a huge job as I suspect that there is some overlap between the FX8500P and the FX8000P, one probably based on the other.

Andrew