progress with Radiosat 6010 / 22dc462
I finally got hold of a RC026 CD changer, at least for a few weeks. I've been hacking around a little, and have snooped some data.
Protocol is indeed async dual rail. When the HU is on, D2B+ and - float up to ca. 2.5V (2.48V for those who like it precise). A voltage difference of 200mV seems to code a 0, and no V difference = 1. I find this quite strange, as one would expect a robust comm's protocol to be based on current loop signaling. Esp. since cars generally enjoy their share of em noise from the spark plugs. I put 10Ohm in the way, and found current levels to be .2mA quiescent and about 2-4mA when signalling.
Protocol-wise there is one long start bit, then 13 address bits with a slow-ish period, then (strangely) the subsequent comms go about 1.5 times faster. Next, 14 bits sender address, then a spacer 5 bits.
Subsequent data/ commands are in 10 bit packages, 8 for saying things, 1 as terminator, 1 odd parity bit.
At least for the Radiosat 6010 the CD player doesn't seem to transmit track lengths, so one can get away with emulating a 1-TR-99minute (or so) CD.
Anyways, more to follow soon. I have some more data for analysis.
VICNE - Do you want to include these in the Sourceforge project? May I email you to the email address quoted in the SF project, or is that a black hole?
Anyone else - I'll analyse my way through some more data myself, but any help and opinions appreciated!
VDO CDC simulator C souce code
I have made a VDO CDC simulator using the amtega8L chip.And here is the sch and soucecode. I have test this simulator in two modle of VDO HU.
The source code need WINAVR for complie.