So I would like to start a thread on an idea I have had kicking around. I love many of the aspects of the existing Renard designs but compared to another common DMX controller "out there" we are a bit feature lite at this point. I think if we took a Renard design to the next level, we could avoid the problem (you know the initials of that problem) of that other design. The "RE" stands for Renard Express.
Here are the features I would like to consider:
1. Base on the Renard SS16 design
2. Focus it on DMX - be able to handle Renard protocol as possible
3. Add a 7 segment display (3 or 4 digits) for Status and in field parameter changes
4. Make the DMX start address be setable (that is stored in flash) in the field using buttons and the 7 seg display . On start, the unit reads that address and responds appropriately. When needed, the address can be changed without using a PIC programmer.
5. On board tests. Using the buttons, trigger a test mode that runs through all the channels and signals status (did the diagnostic pass, was zero cross detected. was DMX data detected, etc.)
6. All on mode. With a button press, trigger an all channels on for setup.
7. New wireless capability OPTION built in. Xbee is OK, but there seem to be better and faster and potentially cheaper wireless solution\s available. The Microchip MRF24J40MA-I/RM is $10 in qty 1 from Mouser and can receive at 115k baud thus handling a full DMX universe. This means needing to design a new tranmitter and the Microchip MRF24J40MB-I/RM (also 115Kb capable) and is about $17 in qty 1 from Mouser.
Additional ideas- remap unused channels in the field. If a channel fails, you could move the lights over to that "extra" channel and logically remap the channel on the unit so the sequence does not need to change. Example: strobes are on channel 3 and the triac fails. You go into the setup and tell the unit to map channel 3 to 16 (your spare) and then move the strobe over to the channel 16 output which responds to channel 3 command.
Again this is just an idea at this point, what do you guys think?
If you are going to take the Renard out of the controller, then perhaps you should take it out of the name as well. I have no objection to a new controller design or DMX, but I would have a problem with using the name for something that it purposefully designed as "non Renard" It would create a great deal of confusion. After all it is the protocol that makes it a Renard, not necessarily the hardware, no?
I don't do DMX, but do the current boards running DMX firmware not meet the needs?
Well we could rename it to something like RenardDMX RE16. The big answer is that the current designs are close (usable) but are missing the features that make it ever so much more usable especially when setting it up. A Toyota Corolla is usable, a low end Lexus adds quite a few bells and whistles. My plan would be to add the missing bells and whistles to a RenardDMX without have to go to a Lexus price. Keep in mind the design is primarily DMX, but would be able to be Renard also.
Kirk
Do you (personally) have any interest in the Renard capabilities, or are you solely interested in DMX?
My first thought was to try and incorporate these features inside one slightly larger PIC. This might be either a 20-pin PIC (something like the PIC16F689) if the granularity is eight channels, or a 28-pin PIC if the granularity is 16 channels. However, you are requesting a lot of features that chew up a lot of pins, so it would be difficult to do in any clean fashion. Any clean implementation would probably need a 40/44 pin part for 16 channels (or maybe 24 channels, I haven't counted the pins), which is either SMT or an enormous chip in DIP format.
Another thought that came to mind is to use the existing PIC designs, and add an up-front PIC for the user interface functions. This might be either a 14-pin or an 18-pin part. I'm not sure if the cost would be higher with this approach than one of the above approaches or not.
As far as naming goes, I'm sort of with ky. If the primary focus is DMX, a different (and unrelated) name would be a good idea.
