[Matthew Robinson]

Hi,

We've been interested in building electronics relating to DMX for a while- leading up to a lighting desk. We decided to start small with a DMX buffer, and have got a nice one nearly planned out. However, after discussion last night about electronic kits that people sell for things such as amplifiers, etc. We discussed making up some modular kits for DMX applications, but before we dive into designing such kits we wanted to do a bit of market research.

The kits we would be likely to design would be the following kind of stuff:

• DMX to TTL/CMOS level converters
  
• TTL/CMOS level to DMX converters
  
• DMX decoders (DMX in to a parallel output for one or more address)
  
• DMX encoders (parallel input to a DMX output)
  
• Intelligent stuff to process DMX- DMX combiners, readouts of DMX values, etc

For example, you decide that you want to make a microswitch on set turn on a light, you could buy a DMX encoder, DMX combiner, DMX to CMOS and CMOS to DMX boards and wire them together in the following way:

              DMX from lighting desk
                   |
              DMX to CMOS converter
                   |
DMX Encoder - DMX combiner
                   |
              CMOS to DMX converter
                   |
              DMX to lighting rig.

Any comments, suggestions, etc welcome.

Thanks,
Matt Robinson & Greg Hargreaves

[timd]

Designing a nice Arduino DMX shield turning DMX into serial for a specified address and block size would be a good start, from that you can extend to pretty much anything.

[timsabre]

Matthew Robinson, on 02 February 2012 - 06:53 PM, said:

• DMX to TTL/CMOS level converters
• TTL/CMOS level to DMX converters

This is just a single 30p 8 pin chip ... SN75176 or DS75176. Not really worth having a separate board for.

A DMX combiner as in your example is not a simple thing to do. If you combine your system into the main lighting DMX, suddenly your kit is responsible for keeping the show running...

The most useful DMX gadgets are break-outs, such as a DMX relay board or DMX transistor switch board. However people like Milford Instruments and Velleman have this fairly well covered already.

[alistermorton]

Matthew Robinson, on 02 February 2012 - 06:53 PM, said:

Any comments, suggestions, etc welcome.

DMX to R/c model servo? I know that the usual suspects already do these though.

[Matthew Robinson]

timd, on 02 February 2012 - 07:02 PM, said:

Designing a nice Arduino DMX shield turning DMX into serial for a specified address and block size would be a good start, from that you can extend to pretty much anything.

Although we'd be more likely to use a PIC rather than an arduino, it's a fair idea and something I've considered many times before- it is, after all, an important part of any fixture with a DMX input.

timsabre, on 02 February 2012 - 07:50 PM, said:

Matthew Robinson, on 02 February 2012 - 06:53 PM, said:

• DMX to TTL/CMOS level converters
• TTL/CMOS level to DMX converters

This is just a single 30p 8 pin chip ... SN75176 or DS75176. Not really worth having a separate board for.

Yes, I have some samples of ds16f95a's on my desk. The point of these boards would be to provide a CMOS/TTL level output with optical isolation and its own power supply to keep the DMX line separate from the CMOS/TTL circuitry. These boards would be pretty much the same, but for a couple of jumpers, placement of the opto-isolator and a few passive components. This was the basis of the buffer that I was originally planning- you would build one board as an input board and your others as output boards to make a DMX buffer as large or small as you want (within reason). Eg: you want a long DMX run, so you could build a single input/single output buffer or you could build a large DMX distribution unit to distribute DMX around a venue.

timsabre, on 02 February 2012 - 07:50 PM, said:

A DMX combiner as in your example is not a simple thing to do. If you combine your system into the main lighting DMX, suddenly your kit is responsible for keeping the show running...

But that could be related to anything involving DMX. We'd eventually like to build a desk, that most definitely will be our kit being responsible for keeping the show running!

alistermorton, on 02 February 2012 - 08:30 PM, said:

Matthew Robinson, on 02 February 2012 - 06:53 PM, said:

Any comments, suggestions, etc welcome.

DMX to R/c model servo? I know that the usual suspects already do these though.

Thanks for the suggestion. It's a possible product.

[Nicktaylor]

To a great extent I believe most kit you now need is available on the open market. I remember many years ago thinking of building a lighting desk. We are talking late 1970s here. I have great bag of sliders in a box at home. IT eneded up being no worth my while except as a project which ends up looking like that. Anything built professionally has to be better than kit I can make. Good luck with the ideas though

[Samwise]

How about a DMX to MIDI converter? You could make it programmable with a PC or just notes to channels etc. Then you could have a lighting desk control music/video software such as logic or older versions of resolume.

I could then for example have my desk hooked up to Native Instruments Traktor on my laptop and use cue stacks to perform complicated playback procedures, including FX (imagine the strobes linked to a sound effect! Well actually that would probably be pretty annoying but still...)

While the bigger, better desks are capable of this already (Magicq) certainly is. It would add MIDI capability to many smaller/older desks. Would be great to launch SFX from your GO button!

Such products do exist, but seem quite pricey and it's fun to build things yourself!

Sam.

[timsabre]

OK, I see what you mean about the DMX-TTL interface, I hadn't considered optoisolating it.

The key to making a DMX product useful to lots of people is to make the software easy to configure/change without the user needing to know C or PIC assembler or whatever. If you could create a "DMX manipulator" board which was USB programmable using some simple graphical software (like MIDI-OX or something like that), I think people would find that useful. Arduino has done this simplified programming to some extent.

[Matthew Robinson]

Nicktaylor, on 02 February 2012 - 09:25 PM, said:

To a great extent I believe most kit you now need is available on the open market. I remember many years ago thinking of building a lighting desk. We are talking late 1970s here. I have great bag of sliders in a box at home. It ended up being no worth my while except as a project which ends up looking like that. Anything built professionally has to be better than kit I can make. Good luck with the ideas though

Our eventual goal would be to become a competitor in the lighting desk sector.

Samwise, on 02 February 2012 - 10:09 PM, said:

How about a DMX to MIDI converter? You could make it programmable with a PC or just notes to channels etc. Then you could have a lighting desk control music/video software such as logic or older versions of resolume.

That's actually quite a good idea, and hopefully not too hard. Perhaps something else to consider.

timsabre, on 02 February 2012 - 11:06 PM, said:

OK, I see what you mean about the DMX-TTL interface, I hadn't considered optoisolating it.

I was thinking of a standalone board that someone would put into their project without having to think about the interface between CMOS and line levels.

timsabre, on 02 February 2012 - 11:06 PM, said:

The key to making a DMX product useful to lots of people is to make the software easy to configure/change without the user needing to know C or PIC assembler or whatever. If you could create a "DMX manipulator" board which was USB programmable using some simple graphical software (like MIDI-OX or something like that), I think people would find that useful. Arduino has done this simplified programming to some extent.

Now that is a good idea. My flatmate (there's 5 of us- 4 EEE students and a physics student) has shown me some DSP manipulation software, and now you come to mention it, we could build something like that which has the potential to be good.

[alistermorton]

Matthew Robinson, on 02 February 2012 - 08:57 PM, said:

alistermorton, on 02 February 2012 - 08:30 PM, said:

DMX to R/c model servo? I know that the usual suspects already do these though.

Thanks for the suggestion. It's a possible product.

Or, along similar lines. DMX control of pneumatic or hydraulic actuated servos. Then you could have actuators that open and close winows or doors or move bits of set under the control of the lighting desk or a dedicated DMX surface.

Matthew Robinson, on 02 February 2012 - 06:53 PM, said:

Any comments, suggestions, etc welcome.

A really small wireless DMX link, to get DMX into a battery powered special without the need for a cable. If they were cheap enough I bet the market would be large - I haven't researched whether such units already exist.

EDIT - unsurprisingly they do, but I still think there's a niche there.

This post has been edited by alistermorton: 03 February 2012 - 09:28 AM

[timsabre]

alistermorton, on 03 February 2012 - 09:22 AM, said:

Or, along similar lines. DMX control of pneumatic or hydraulic actuated servos. Then you could have actuators that open and close winows or doors or move bits of set under the control of the lighting desk or a dedicated DMX surface.

But DMX isn't really suitable for moving big stuff. R/C servos OK, moving set or doors/windows, too much chance of injuring someone.

[alistermorton]

timsabre, on 03 February 2012 - 09:26 AM, said:

alistermorton, on 03 February 2012 - 09:22 AM, said:

Or, along similar lines. DMX control of pneumatic or hydraulic actuated servos. Then you could have actuators that open and close winows or doors or move bits of set under the control of the lighting desk or a dedicated DMX surface.

But DMX isn't really suitable for moving big stuff. R/C servos OK, moving set or doors/windows, too much chance of injuring someone.

Sure, I wouldn't want to use sometihng open loop like DMX for moving a heavy or dangerous bit of set, but rather for something like a small fanlight or a simple effect where the lightweight actuators would work better than the small r/c stuff to move something in the set, like a bathroom cabinet door or something.

[alistermorton]

alistermorton, on 03 February 2012 - 09:25 AM, said:

A really small wireless DMX link, to get DMX into a battery powered special without the need for a cable. If they were cheap enough I bet the market would be large - I haven't researched whether such units already exist.

EDIT - unsurprisingly they do, but I still think there's a niche there.

The ones I've seen are quite bulky. A very small DMX receiver that could be fitted into small practicals would have a real market if it was cheap.

[boatman]

alistermorton, on 03 February 2012 - 12:00 PM, said:

alistermorton, on 03 February 2012 - 09:25 AM, said:

A really small wireless DMX link, to get DMX into a battery powered special without the need for a cable. If they were cheap enough I bet the market would be large - I haven't researched whether such units already exist.

EDIT - unsurprisingly they do, but I still think there's a niche there.

The ones I've seen are quite bulky. A very small DMX receiver that could be fitted into small practicals would have a real market if it was cheap.

A much more robust solution to that problem would be a static DMX decoder which converts the DMX to a proprietary serial command protocol for the wireless signal. DMX over wireless is very greedy for bandwidth and error prone, hence the cost and complexity of the commercial units. However, basic RS232 to 868MHz wireless modules are readily available (see here), which makes it possible to build in some error detection and correction on the wireless link.

That would be an ideal final year project for a an Electronics Engineering student.

[timsabre]

I've done quite a bit of stuff with the easyradio modules you linked to, and they are quite slow relative to the DMX speed. If you need to keep up the refresh rate of the DMX (e.g. for fading a battery powered lamp or whatever) you can't send very many control channels.

If you only needed a couple of channels sending over radio, this works fine.