There's one more step in the chain that could be a problem: bad
DMX input on the DMXi. With the full chain intact, the "COM"
LED on the
fixture should be blinking rapidly. Unfortunately, breaking the chain either between
fixture and DMXi or between DMXi and your
DMX source gives the same result of a slow blink, so you can't really tell from that where the problem is. The transceiver chips used for
DMX and Irideon
protocol are the pretty standard 75176B, but they're soldered rather than socketed, and surface-mount to boot. An oscilloscope would be really helpful for identifying whether they're working... but I'm assuming if you had one you would have already mentioned that.
As a very rough approximation, you might be able to use your true
RMS multimeter to sort of test.... it's the kind of thing where failing the test is a sign that it might be bad, but passing the test doesn't necessarily mean it's good. First,
DMX input: you should get at least 1VAC between pin 1 of the
DMX input transceiver (U4 on mine, but it might be labeled differently...closest one to the
DMX in
header) and
ground. I used pin #1 on the input
header for
ground. The actual value may be higher depending on the
DMX levels being sent from the
console. I was getting about 1.1VAC with all channels at zero. Next,
fixture output: Irideon refresh messages are sent much less frequently than
DMX, so this clunky approach to checking for a signal isn't sensitive enough to detect them unless something is actively changing. So, you'll have to connect your
meter across pins 1&2 or 1&3 of the
fixture output
header and then move a
fader back and forth on your
console. While the
fader is moving, you should see it jump up to about 1VAC. When idle (or presumably when not working) it's more like 50mV. You'll see something similar at the
fixture on pin #1 on the input transceiver (to
ground). On mine that's U3 on the board inside the arm.
I couldn't remember whether the self-test works as described, so I retrieved one from my basement and plugged it in. The behavior of the
address wheels does indeed match the
manual, so it seems there might be something off with your
fixture. There's one stupid sounding, but easy thing to try if you've already popped the
cover off the arm: the
processor is socketed and sometimes that type of
socket can have some oxidation
build up on the pins. Pulling it out and plugging it back in is enough to scrape away that oxidation. There is, of course, a proper tool for this, but some careful prying with a couple of small screwdrivers can work in a pinch.
Reverse-engineering the Irideon
protocol is tricky because you need a working DMXi in order to inspect the output. If your DMXi is working, then you don't really need to make your own controller.... unless you want to just for fun. To do that you'll need something that listens to RS-485. A USB serial port plus an RS-485 to RS-232
adapter would work (although you probably could also find that all
in one device). I don't have one handy, but at a glance on my oscilloscope it looks like you'll want to start with a baud rate of 38,400, 8bit, no parity, 1 start, and 1 stop
bit. Use a serial port
monitor program that lets you see the raw hex values and watch what happens as you move different parameters. I'm kind of assuming it's not terribly complicated, but it is a little different from
DMX. Instead of a continuous repetition of all the values, it should be individual commands like, "Pan to X position at speed Y". If you're serious about making your own controller, I could probably be convinced to cobble something together that would let me capture some of this information for you.