MIDI-controlled Projector Douser

Forum 2005-2010 (read only) General Exhibition
1

Show I am working now was using a video projector but didn't want to see the dim light left during "digital black." Newer projectors have an internal metal gate or douser. Since I had a servo lying around, I thought this was a good excuse to finally getting around to hooking it up to make a DIY retro-fit.

We were hosting the actual videos via Qlab, a Macintosh show playback software that already gives robust MIDI and MSC commands in and out. On a previous show we were using Qlab to send "Go" commands to an Expression lighting console. So it made sense for the same software to trigger the douser.

Douser itself was a simple flag of black Cinefoil. Had to teach myself how to sweat brass connections with MAPP gas to make the mechanism. Used a 4N33 as my local electronics store didn't have the opto others had recommended, and the project deadline was too short to allow mail order. With some component value adjustments it worked fine.

To give enough "oomph" for the servo I bypassed the Arduino's power regulator completely and added a jack for an external 5v, 1 amp wall wart. Also, in case MIDI scared the client too much, a 1/4" phono jack to allow me to hook up a simple "douse/reveal" switch instead (which is patched to an analog in port for future expansion).

As currently configured the MIDI note number sets the desired servo position, and the MIDI note velocity is (mostly!) patched to either insert a delay between servo steps or to slew at maximum rate. Still to code is pitch wheel CC to real-time position, and "sleep" mode (there's a Phillips ODC5 in there in case I want to stop sending power to the servo for a while).

Well..I stuck the box on top of the projector with double-stick tape, ran the data and power lines, and it worked like a charm. I also wrote a Qlab show that doused the projector every 2 seconds, and left that running for over an hour. So I know it works!

But we cut the projector from this particular show, so the box is back on my workbench now. That's show biz.

2

But we cut the projector from this particular show, so the box is back on my workbench now. That's show biz.

All that hard work - that stinks!

Got any pics of the setup?

:slight_smile:

3

nice work.

i recently made something similar, except mine was dmx controlled.

instead of a separate power supply, i just used a really small servo. since all i had to do was move a small bit of plastic.

how exactly did you make the circuit to disable the power to the relay?
i suppose i could just do it with a power transistor. this is one thing i'd like to add to my design.

unfortunatly i only had 3 pin xlr connectors in the workshop so i committed the shamefull sin of not using 5 pin xlr for my dmx.

the problem i had was that many theaters actually don't like to give you dmx on stage(we used the projector in a black box on the front of the stage) so i had to run annother dmx cable from the desk to the stage. and to further complicate matters my lightfactory licence only has one universe and i didn't have a dmx splitter. so i had to set up annother (little, chinese)lighting desk just for the douser.

i think that there is a better way of solving the douser problem though:

wouldn't it be much better not to have to lay another cable?

maybe it'd be better to build a circuit that detected black video and automatically activated the douser?

i wonder if there are any preexisting vga-in circuits for arduino?

or for manual control it would be possible to use the ddc pins on the vga cable, (ddc is just an i2c buss that is used to transmit information about the display.)

i may look into this.

4

Heh. 3 pin DMX is one of the sorry secrets of the industry. Non-compliant, yet seen on so much gear!

This was my first servo project and I didn't have a good instinct as to how much power I'd need. I did want it to be over-engineered, because I wouldn't be there when they were using it.

My power cut was a Phillips ODC-5. That is one of a series of integrated modules with built-in opto-isolator and solid-state relay. All you do is connect the leads. Looks like the non-moving power draw of the servo is small enough I didn't need that, either.

I've been through the DMX daisy chain more than once myself. As it happens...although again it is non-compliant, I built MIDI-to-XLR adapters. So the cable run is XLR. I also have been running TTL through XLR and the house snake/wiring; close a switch at one end of a 200' run of house wiring, and an Arduino at the other end detects it.

I did think about detecting light. I even went a step further; you could code the light, or put it in a small spot in an upper corner like the old reel-change lights. This seemed simpler and easier to debug for its first outing. As i mentioned, MIDI was drifting around already so it made sense to detect that (so was DMX, but we had a limited number of channels we'd already maxed out).

5

My current thought project towards developing new theater technologies is Dracula's Coffin. Helped a friend of mine install a "Dracula" a few shows back. The thought problem is how you would automate a creaking sound to "sweeten" the lid opening.

The coffin was on wheels and pushed around the stage; that means wireless. Dracula rattled the lid before opening it properly; this means a contact switch won't work, and with him inside the coffin a proximity sensor might fail as well. Possibly a photocell that acts when it can "see" the stage lights? It even occurs that you might have a coded infrared source in the coffin that is revealed to a detector mounted in the booth and pointed at the stage...

Thing is, it is very hard to improve on a skilled board operator and stage management team, and the traditional development of an effect through rehearsals. Outside of doorbells and gunshots, I've been having trouble practically integrating Arduino-trigged sound effects into ordinary stage productions.

6

the way i look at it is that it's ok to use mic cable in a pinch, but it's a bad habit that shouldn't be encouraged.

i toured a show for a whole season with 4 movers and 2 dimmers all on mic cable once. ocasionaly the movers would start twitching and i'd have to hunt the dodgy cable. it wasn't really that bad. since i was expecting it and kept the mic cable that i used for the dmx separate from the regular mic cables.

before i started making things with pics and arduinos i made loads of stuff that was just a switch connected through a mic cable. i think that these days i tend to over engineer things.

the "old skool" dowser is an old cd rom drive whose tray has been covered in gaffer, with the "open" button connected to a switch.

i'm actualy upgrading my douser right now, i'm just using a bd139 power transistor and a diode, that i found in my spare parts box.

i'm going to add code so that if the servo hasn't moved for a few seconds then it get's it's power switched off.

for dracula's coffin why don't you rig up a potentiometer to measure how open the door is? with the arduino you can do all the signal processing you like. "debouncing" acceleration, whatever... you could even make it creak in either direction.

a cheap and simple way to do wireless would be to play the sound effect into a wireless mic transmitter. i'd trust a ?1000 sennheiser a bit more than i'd trust a ?40 garage door opener. and the theater will already have radio mics.

i have a friend who does acrobatics on a trapeeze dressed like a camp vampire. i've been trying to convince him that he needs a coffin that he can rise up out off like max shrek does in nosfreatu. if i ever make it i think it'd have to creak like yours.

7

I'm fond of over-engineering too. I like to think paranoid, and build beefy. Plus the less a circuit looks like a breadboard and gaff tape, the more confidence it gives to the (often techno-phobic) stage management and running crew.

And as importantly, building it in a nice box with internal connectors and lots of extra I/O already wired makes it easier to re-purpose for some other project.

I got into Arduinos with one specific need; to compose MIDI messages on command. Sure, MIDI is old hat, and slow, and so on and so forth, but it is already in the theater. You can run it down a length of XLR or send it wireless. Sound boards and light boards already speak it, plus there are always keyboards and stuff in the pit.

It does change how you approach problems, though. Instead of trying to engineer the thing so it works out of the box, you start making it only an approximation on the physical layer (servos, sensors or whatever) and do the fine tuning in software. I'm doing a mechanical effect right now and if it wasn't so late in the game I'd drop all my work with springs and things and just replace it with servos and software. And this is me, who built a "Little Wonder" some shows back with a spring-loaded rubber dagger for the surprise (and for a friend, a set of hand-claws that was deployed via a biceps trigger).

I'm on about my third iteration of a "make a MIDI note event when a button is closed" box. Already it has; been a remote "Go" button for Qlab when there wasn't room by the Stage Manager for a laptop, triggered a train whistle from the pit (to be on-tempo), triggered a baby cry from the pit (it repeats and is in the score), buzzed an intercom from both on stage and off stage, been an actor-triggered doorbell (in this case played through the sound system), and (when rigged with a cheap RF keyfob) triggered a gunshot sound under actor control.

On the latter, my "dream" machine is an Xbee radio link and a Hall Effect or photo-interrupt to detect the actual hammer fall of a prop gun. For extra points, something clever would triangulate the actual gun position and adjust the pan of the playback!

But, really, outside of the perennial gunshot sound problem, my next iteration is going to be more of the same; a central box that is USB powered and spits MIDI or MSC on receipt of sensors. But this time, based around XLR connectors and providing phantom power so the remote stations will both display status and report status back to the main station between cues. Detecting the small current drain through an LED accomplishes both tasks. With bicolor LEDs, the remote switch boxes can even show successful triggering of the base station.

(On the "baby cry" show I had to deal with some unlabeled cables during a strike-and-restore, and it was a lot of fun tracking down the right line!)