I was wondering a few things about how the Prizm works on a lower-level point, and with such knowledge whether or not it would be possible to send music through the 2.5mm link port like previously accomplished with the ti-8x family. Some questions I have:

- what port(s?) correlate to the Prizm's link port?
- does the port follow the standard of sending/receiving normally done with 2.5mm jacks?
- are there any syscalls that could currently aid us in an effort like this?

This is not known.There aren't really any standards when it comes to this kind of ports, but it's pretty close to how things normally work.If there are any, they haven't been documented yet. The ones that exist probably only deal with Casio's linking protocol.

hmm, interesting. However, we would probably learn a lot by dissecting any transfer syscalls (once they are discovered).

Yes, we definitely would. As JosJuice said, there's no "standard" for three-wire, two-channel, 2.5mm serial, but once we had access to the port and could arbitrarily change voltage levels, we could make it speak the TI link protocol, make a TI speak Casio, or make the Prizm talk CALCnet.

interesting I know very little about stuff like this, as I'm no EE. But, does the port itself regulate how many channels of sound you can have, or the actual software? Or neither?

The port has exactly two independent channels, tip and ring. You have a single "core" as it were of CPU on the Prizm (and the TIs, for that matter), so the number of audio channels is more or less up to your software. For mobileTunes, a total of four channels of painstakingly-interleaved sound is possible, for instance, two simultaneous notes per ear.

1) "Ports" as you think of them from the z80 aren't quite the same. There are ports on the Prizm CPU, but it's pretty much unknown what they do or for some, if they're even wired up.

2) There's no standard per se, but Casio has released documentation for the old fx-9860g, which probably has a similar or identical protocol. I've managed to get my TI to talk to the Prizm through the link port as well, so it's definitely possible to send music through.

3) Yes. Understanding OS code isn't exactly easy though. It's the software equivalent of extracting rotten teeth with rusty tongs </Exaggeration>
Link protocols aren't the easiest thing either...

Do keep in mind that the SH3 has hardware support for sound. That is through the SIOF which controls the serial port for I/O operations including sound. It would be good to test this as there is the possibility that Casio may have left it intact with this version of the Super H

I think that's unlikely, since they didn't even bother to use the hardware USB either.

Qwerty, I hope that you're wrong about that, although sadly you're probably not. Qwerty, by making them talk together, you just mean detecting the line-level changes generated by the Casio at the TI, correct?

Yes, as well as have the Casio detect line level changes by the TI.

How did you do that? You accessed the 2.5mm port using asm?

I tend to think not, considering we don't know how to do that yet, as far as I've heard.

Well, my curiosity still exists on the matter on hand, and it piqued today when I saw a new library added to the SDK v0.2 that had three serial-controlling functions. It seemed to only have an open, close, and send unsigned char function, but I'm wondering if this would be enough to generate any sound? I haven't any knowledge of the subject so I'm just curious what degree of control over the port would be needed for sound generation.

Argh, I keep meaning to check if the link cable is a crossover cable. If it is, then stereo sound and CALCnet are definitely out the window, as it will mean that the link port is indeed RX, TX, and GND. If it's a straight-through cable, then we may have some hope yet.

I would like Ashbad to port me MTV Music Generator/Music 2000 to the Casio Prizm. No more need for ePSXe.

I've confirmed that the cable is indeed a crossover cable. I'll post Cemetech news about this later.

That only tells me that the link port has an RX line and an TX line, that the signals are most likely self-clocking, and that the port might support full-duplex communication.

The TI-9x series has link hardware that handles the TI linking protocol, but it also has a mode that gives programs direct access to the signal levels. Don't you think it's possible that the Casio has similar hardware for its link port? Of course, the real challenge is figuring out how to access the functionality, if it exists.

Has anyone tried asking Casio directly about things like that, or are they probably not going to release any information, like TI does? If not, we could try asking them for a set of standard libraries or a list of syscalls on the device.

I believe that z80man has sent at least one email to Casio Japan, and it seemed like they didn't hate third-party programming. I'm not sure if they're going to give us anything right now, though. There might be an SDK release later...