ADAT sync port pinout
=====================

The ADAT sync pinout supposedly is described in US Patent #5541781.
However, the pinout given in that patent is not actually accurate 
for real-life ADAT devices.

The actual pinout is:

Sync out port:

1   chassis ground
2/6 MIDI Sync Rx +/-
3/7 word clock -/+ (3.3VD)
4/8 time code -/+ (3.3VD)
5/9 MIDI Sync Tx -/+  (3.3VD)

Sync in port:

1   = N/c
2/6 = MIDI Sync out
3/7 = Word clock in
4/8 = Time code in
5/9 = MIDI Sync in 

Pin 5 and 9 on the Sync Out port carry some flavour of Midi
Machine control and run at the regular MIDI rate of 31250 bps.
Rehearse buttons, Play, Record, Stop and Locate commands are
communicated over these pins. The time codes for the locate
commands seem to be scrambled. One thing notable about the
data on pin 5 and 9 is that the HD24 presents itself as
device 0E 08 (Alesis/ADAT) whereas when looking at the .syx
OS file, the HD24 device ID is presented as 0E 20 (Which must
mean Alesis/HD24). No doubt the ADAT device ID is used to
maintain backwards compatibility with existing ADAT systems.

Pin 4 and 8 contain a timecode signal. This signal is clocked
at the wordclock as present on pin 3 and 7 and consists of
a start bit, 32 time code bits (most significant bit first)
and a stop gap. During playback, the last six bits of the time
code remain constant, which shows that the time code increases
in steps of 64 samples at a time. Steps need to be this big
because each bit in the time code also corresponds to 1 sample,
and a gap is required between time codes to permit detecting
the start bit.

The time code for 00:00:00.00 is to be represented by the value
5760000 (2 minutes at 48000 samples/second). This probably is
a leftover from the ADAT tape days, where the tape had a
lead-in (and indeed negative time codes) before 'time code 0'.

Pin 3 and 7 carry a word clock signal as long as the HD24 is
powered (even when it has been powered off from the front panel).