Film-Tech Forum ARCHIVE: JNIOR as a DMX Fixture

This is topic JNIOR as a DMX Fixture in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=003454

Posted by Bruce Cloutier (Member # 9615) on 10-05-2018, 03:31 PM:

Just a quick note as we head into the weekend. You know, something different while you guys struggle with all of the "technology" that was supposed to make things better. NAS and all that...

Steve had asked some time ago if the JNIOR can be used as a DMX fixture? More and more theaters are taking an interest in DMX but in the JNIOR generating the DMX output to control a few lights in the auditorium. For that we do have the JNIOR 412DMX available.

Turns out that the DMX512 protocol is not something that is compatible with standard UARTs. Even after you convert the RS-485 signals and figure out how to handle the 250 Kbaud data rate there are issues in synchronizing frames. It just can't be done reliably without some hardware.

Well if I am going to create a JNIOR with a DMX input I sure as hell need to understand how to pull it off. I know you can do it because fixtures do it all of the time.

I designed some simple hardware to handle it and just spent couple of days writing it up in this blog.

JNIOR as a DMX Fixture Revisited

If you are bored you can read it and comment. I need to reread and proof it some I bet.

If any of you have contacts in the "Theatre" arena to where this might be of use, let me know. I am not trying to sell it to you or them but I would bet that you agree that designing things in a vacuum is not a good idea. Just trying to listen to customers (if I can find them).

[Big Grin]

Posted by Harold Hallikainen (Member # 5405) on 10-05-2018, 05:26 PM:

Interesting! Semiconductor documentation tends to be very thorough except for the subject you need to know; in this case, recovery from a break. About 20 years ago, I designed several DMX products for Dove Systems. These were both dimmers and controllers. The first was the "Shoebox dimmer" ( http://www.dovesystems.com/shoebox ). It originally used a Microchip PIC 17C42 (this was before flash, so it was a one time programmable chip). Most of the later products used the PIC 18 series. I just looked at the UART documentation. It, of course, says nothing about how it recovers from a framing error (immediate, or take up to another character time). However, I had no problem with that particular UART. Looking at my 20 year old code (assembly), it looks like that particular UART only sets FE once and then does nothing until we go back to mark and get a good character after that. Again, I don't find anything quickly in the Microchip documentation, but that did work (and was verified under all sorts of conditions with a DMXter).

On the transmit side, though, I DID have to use a GPIO pin to generate the break since the UART would not do a long enough break.

Fun stuff!

Harold

Posted by Bruce Cloutier (Member # 9615) on 10-06-2018, 09:52 AM:

Hey Harold, that's a good point that not all UARTs are created equal. The current JNIOR series uses the Renesas RX microcontroller (RX63N specifically) and documentation for any of the peripheral silicon is completely absent. My Renesas support channels even come up empty handed. I was hoping for a configuration setting for resync. None apparently. Oh well. It is like they developed a reasonably good CPU core and then loaded a bunch of peripherals from some open source silicon block library that nobody knows much about just to get the thing out on the market. Off-topic. I know.

It was fun to throw together and external logic solution anyway. Doesn't add a programming step in production.