DMX code for renard???

[Deast]

I was reading a while ago, and remember seeing something about a version of DMX for the renards. Is anyone still working on this?

[RavingLunatic]

Phil has the Renard-DMX code done (more or less). I know he is hoping for more testing to be done to make sure that there aren't any bugs.

I stopped testing when I lost faith in my Enttec Pro Dongle at short event intervals. I need to put together one of RJ's dongles and try the code with that dongle.

[Deast]

Thanks!!

I'll have to pm Phil... or he can get a hold of me first.

[ben]

so Phil, can you post the firmware for DM to renard so I can do some testing with it? I am interested in going Renard-> DMX

Ben

[stempile]

I am planning on going DMX this year as well. I have 3 Renard64s that I am going to convert over. I haven't selected a DMX dongle yet. Still catching up on what I have missed since not being active on this since Feb '08.

THANKS!
ms

[tconley]

I am interested in this as well

[tconley]

Has anyone got the update firmware for DMX yet?

[darko]

*bump*

Argh. First timer here, ready to take the plunge, but making the hardware choices is difficult not knowing which will be DMX compatible in the future.

[P. Short]

OK, here is a version for more testing. It should be taken as Beta code.

A few comments, which I'm sure partially duplicate the other thread on this topic.

1) Connector Pinout

The Renard PCB RJ45 pinout is different from the standard DMX512 RJ45 pinout. You will need an adaptor cable or adaptor board to connect a Renard board to a DMX512 cable.

2) Multiple boards

There are two different ways for you to connect multiple Renard boards to a DMX controller.

The preferred method is to connect all boards in parallel to the DMX512 cable. The first PIC on each board will be programmed with the desired DMX address for that board, and all other PICs on the board will be programmed for address '1'. In this situation you will only use the input connector on each Renard board, and the output connector will be left open.

The alternate (non-preferred) method is to daisy chain the Renard boards in the same way that you would with Renard non-DMX firmware. The first PIC on the first Renard board will be programmed with the desired DMX address (1-512), and all of the rest of the PICs (the remaining PICs on the first board, and all of the PICs on the remaining boards) will be programmed with address '1'. The advantage of this scheme is that it is easier to set up (fewer PICs with non-default addresses, fewer adaptor cables/boards required). The disadvantage is that it is much more difficult to intermix Renard-dmx boards with other types of DMX devices, and there is less flexibility in assigning DMX addresses.

3) Terminator Resistors

The default BOM for all of the Renard boards has a 120 Ohm terminator resistor across the RS485 lines. This doesn't work well with DMX512 in most situations, so you should remove these resistors from the Renard-DMX boards, and use the prescribed DMX termination methods.

4) Configuring the PIC firmware

With this firmware you will be able to configure each channel individually for PWM or non-PWM operation, as well as for current-sink vs current-source outputs. This is a little bit different from the previous Renard firmware, and is described in the .asm file.

The usual disclaimers apply...this is beta firmware, and use it at your own risk.

[darko]

I had completely missed the other thread ;)

Thank you very much!

[tconley]

Thank you.... This info will be very helpful

[ben]

3) Terminator Resistors

The default BOM for all of the Renard boards has a 120 Ohm terminator resistor across the RS485 lines. This doesn't work well with DMX512 in most situations, so you should remove these resistors from the Renard-DMX boards, and use the prescribed DMX termination methods.

Phil, first of all let me say thank you for the beta Release. When I get a chance I will run some Renard boards and test it out.

I am curious about #3. The termination resistor for DMX is 120ohm. Are you saying the placement on the RS485 line is not good for the signal or that is flat won't work with the DMX signal? I do know the termination pins for DMX are (more than likely) not that same 2 RS485 pins.

Ben

[P. Short]

You need to follow the DMX rules for terminating the line. This, I believe, means that there is one termination on the cable at the furthest possible location from the signal source (the host controller). In most cases this rule would be violated if you left the termination resistor on the Renard-DMX board (using the preferred connection method from above).

The problem with leaving the resistor on the Renard boards is that the load on the line would be too high for the RS485 drivers. The load would be effectively n+1 resistors (one per Renard plus the usual line termination resistor).

[P. Short]

One added comment -

Caveat:

The front-end circuitry on the Renard boards was not designed to be DMX512 conformant, and has not been run through any of the DMX512 conformance tests. It is likely that each board will have an input impedance and sense levels that are slightly different from the DMX512 standard. It may be possible to bring the boards into conformance by careful selection of the RS485 receiver chip and the voltage divider resistor chips on the front-end, but I don't have the facilities to do this testing (and it would have to be done separately for each board, probably). On the other hand, I think that this difference would only cause problems in very large setups (many dimmers spaced over many hundreds of meters of cable in electrically noisy environments).

[ben]

One added comment -

Caveat:

The front-end circuitry on the Renard boards was not designed to be DMX512 conformant, and has not been run through any of the DMX512 conformance tests. It is likely that each board will have an input impedance and sense levels that are slightly different from the DMX512 standard. It may be possible to bring the boards into conformance by careful selection of the RS485 receiver chip and the voltage divider resistor chips on the front-end, but I don't have the facilities to do this testing (and it would have to be done separately for each board, probably). On the other hand, I think that this difference would only cause problems in very large setups (many dimmers spaced over many hundreds of meters of cable in electrically noisy environments).

So..... don't coil my Cat5 ontop of my subwoofer ? ;)

Thanks for the info Phil. I think Franks board will work great for termination and transfering of pins

Ben

[RavingLunatic]

3) Terminator Resistors

The default BOM for all of the Renard boards has a 120 Ohm terminator resistor across the RS485 lines. This doesn't work well with DMX512 in most situations, so you should remove these resistors from the Renard-DMX boards, and use the prescribed DMX termination methods.


It will be interesting to see what happens when using Frank's Renard2DMX board and connecting a bunch of Renard boards to the DMX line. I'll wait to do any de-mods til after I do some experimenting to see what happens. I'm not too worried about sticking to the "DMX code" as long as it works.


Education Time:

After your comments and actually thinking about the circuitry (instead of just accepting it as is), it appears that the data input circuitry is setup the way it is so that it can handle the RS232 signal. Is that correct?

If it is, then it would be perfectly acceptable to omit those components on a future board if you intend it to only handle RS485 (DMX) and not RS232, right? Just feed the input directly into the 75176 chip (or suitable sub).

Just thinking ahead to next year.

[P. Short]

Yes. I had forgotten about the two zeners. They can be left off for DMX use, thereby reducing the capacitance on the RS485 inputs.

The purpose of the two 1K resistors and the two 27K resistors is to provide fail-safe biasing for the inputs. This means that the output of the RS485 chip goes to a guaranteed high state if the inputs to the board are either left high or shorted together. This is from one of the TI application notes, and I'm not sure if the resulting design meets the DMX512 specs.

[darko]

.....

2) Multiple boards

There are two different ways for you to connect multiple Renard boards to a DMX controller.

The preferred method is to connect all boards in parallel to the DMX512 cable. The first PIC on each board will be programmed with the desired DMX address for that board, and all other PICs on the board will be programmed for address '1'. In this situation you will only use the input connector on each Renard board, and the output connector will be left open.

The alternate (non-preferred) method is to daisy chain the Renard boards in the same way that you would with Renard non-DMX firmware. The first PIC on the first Renard board will be programmed with the desired DMX address (1-512), and all of the rest of the PICs (the remaining PICs on the first board, and all of the PICs on the remaining boards) will be programmed with address '1'. The advantage of this scheme is that it is easier to set up (fewer PICs with non-default addresses, fewer adaptor cables/boards required). The disadvantage is that it is much more difficult to intermix Renard-dmx boards with other types of DMX devices, and there is less flexibility in assigning DMX addresses.


I was just staring at the renard24 DMX converter board schematic, and I finally see what you mean by "Parallel" wiring.

Does this neccessarily mean that, with some future design, we could have a renDMX board that can be either daisied OR wired with a star-like (parallel) topology? Is it possible to have hybrid layouts with "subcontrollers" spoking off of more centralized daisy-chained controllers?

If so, yikes, thats cool :)

[P. Short]

DMX512 only supports a linear topology (unless you use some sort of active repeater to provide for branches). While you could use the Renard-DMX board as an ad-hoc repeater, I don't think that this would be a good idea. You would end up driving yourself crazy handling the addressing complexities.

[darko]

By addressing complexities, are you referring to the need to have to individually program each controller with a start address?

If thats the only issue, I think I would actually be comfortable with that. Perhaps I could sharpie the start addresses onto each enclosure. Maybe ICSP can be used too.

EDIT

Let me clarify. By "Daisy Chain", I was simply referring to the DMXin and DMXout ports on the converter board. I wasn't referring to wiring the out port to the last pic in the onboard chain.

I guess since the DMX board itself has the DMX In and Out ports wired in parallel, there really is no distinction anymore right? Everything is parallel, so does it require a linear topology?

For future designs (which incorporate the ren + DMX coverter circuitry) we could daisy chain (in a parallel sense), but also use passive splitters where we need to produce branches?

http://www.ebest24.co.uk/images/products_images/unfurl/ebest24b273em.jpg

[tconley]

you can always use RJ's REnard splitter.

http://diylightanimation.com/index.php?board=24.0

[rrowan]

you can always use RJ's REnard splitter.

http://diylightanimation.com/index.php?board=24.0

I think you mean RJ's DMX splitter since he doesn't make/design Renard hardware.

Cheers

Rick R.

[tconley]

you are correct sorry

[darko]

Forgive me, but I cant find the schematics. Has he posted any?

[scorpia]

RJ doesnt normally post his schematics. do a search on the intrernet and you should find schematics to basic dmx splitter/isolator

[jrock64]

I have something something in mind for this year and I want to build a lot of them, so I am looking for the simplest design possible.

Phil, is it possible to tweak the Renard DMX code to use the internal clock so the oscillator can be omitted.
Unless its already in there and I missed it.

Just asking, if it can not I will just have to include the part.

JOel

[P. Short]

The code was designed for operation at 16 MHz (i.e. using the internal oscillator). The timing is very tight, though, to the point where adding one extra instruction at certain critical places would put it over the limit.

[smartalec]

you can always use RJ's REnard splitter.

http://diylightanimation.com/index.php?board=24.0

i found this as a DMX splitter
http://users.skynet.be/kristofnys/files/booster/dmxbooster.pdf
i dont know if its the same one from RJ's or not, as i couldn't find his schematic anyware

[Matt_Edwards]

i found this as a DMX splitter
http://users.skynet.be/kristofnys/files/booster/dmxbooster.pdf
i dont know if its the same one from RJ's or not, as i couldn't find his schematic anyware

We have one of those as our very own. Check this One (http://doityourselfchristmas.com/forums/showpost.php?p=48902&postcount=1)
I use two of them in my display, operation was faultless. I don't have very long cable runs nor a lot of devices in one chain.

[smartalec]

that looks sweet, do you know if boards are still floating around for sale an how much?


We have one of those as our very own. Check this One (http://doityourselfchristmas.com/forums/showpost.php?p=48902&postcount=1)
I use two of them in my display, operation was faultless. I don't have very long cable runs nor a lot of devices in one chain.

[Matt_Edwards]

that looks sweet, do you know if boards are still floating around for sale an how much?

PM wjohn

[smartalec]

Allready thought of an done.. thanks
im thinking of running dmx on my linux box, something to play around with an test.
since i have the renard64's working great with vixen on its own laptop.

There's a lot to learn with the visual basic an dmx code stuff, looks intresting, I mainly like the idea of writing the code an then sending it via dmx to dmx made lights.

question, is there code for the pic's on the ren64 to make them DMX. that would be so cool if thats possable.

never done pic coding so i was hoping of using someone else's code if its out there.

[P. Short]

The firmware source can be found in the wiki...

http://www.doityourselfchristmas.com/wiki/index.php?title=Renard_Firmware