Hello all, As mentioned in the title, I would be interested to know if somebody has a broken Sega Katana, as I want the spare parts, specially the CPU. The reason behind it is because I'm building my own Dreamcast devkit and I need the CPU to have JTAG access to it. Please let me know if there is anybody interested. Sh0X.
I'm slightly confused here - the CPU in a retail DC has a JTAG interface (well, they called it "H-UDI", but electrically it is precisely an IEEE 1149 boundary scan interface), a breakpoint module on the chip and an external break line, all nicely bought out to resistors on the PCB. I don't think Hitachi ever made a public release of the debugger protocol, though...
No sorry, The Retail Dreamcast / Naomi / Naomi2 ... got a CPU which is the HD6417091 - R - RA - T. This CPU has a dummy H-UDI interface. The Katana DevKit has the HD6417091ES9 which is also called EVA (evaluation). This one is the only one that allows H-UDI interface. I have couple commercial JTAGs and they don't work, so yes, I'm pretty certain it does not work.
Well, I can remember getting the JTAG working well enough to drive signals onto the AD bus and read the boot ROM using a retail DC, so there is definitely at least some functionality there - Hitachi (and it was Hitachi at the time, this was before the Renesas era) were perfectly happy to give me the BSDL files for the part, but were not willing to divulge any information about how their debug interface worked. And if that "ESx" notation is on the line below the part number, then it means "Engineering Sample".
That's amaizing, definitely Hitachi was much more friendly than Renesas. I managed to the the CPU hardware manual for the variant SH7091. You are correct, the JTAG efectively returns something, but the CPU id is always 0x0, and the internal SDSR used for handshake between JTAG and CPU always returns 0x12345678. So please, if anybody has a broken Katana, do not hesitate to post here .
