Analogue Protection Filtering for ILDA XY/Galvos

[mixedgas]

~For the record, nothing will stop a determined galvo user from frying the galvos, hopefully the polyfuse, CTC, or output fuse trips first. For a music synth, the low pass and clipping is a good start.
~
Abstracts work on frequency ratios, amplitude ratios, and phase, a 3:5 ratio frequency generates a five pointed star or pentagon for example. However, the lowest frequency for an image to be flicker free to the eye is ~20-30 Hz. So if I divide 1000 Hz by 30Hz, I get 33... A 33:1 ratio, so if the user's base or master circle is 30 Hz, they have more then enough possible patterns to get started. A ordinary 6 dB per octave filter rolloff is fine...
~
No need for a steep skirt, the resonance test is done with the Galvo amplifiers high and low frequency damping controls set to zero. When the damping is turned back up, you get additional resonance suppression. So you have good protection from rogue frequencies at above 1 KHz, and yet plenty of ratios to work with for complex patterns... 1 KHz is a actually a tad short for but will be very safe, even at 12 or 20 KPPS... If your response is ~ 2-3 dB down at 1.5 KHz, your doing fine.
~
Just get as flat as possible phase response, because a phase shift between two quadrature sine waves of more then +/- 2 degrees, starts to show as an ellipse when projected.... You may need an "all pass" network to correct for major phase errors in your filter...
~
500 Is just an arbitrary number at this time.
~
Remember your ILDA connector must short pins 4 and 17 for the projector safety interlock... It helps if that interlock can be a "Big Red Switch" of the panic shutdown style, or "key Switch" to conform with international standards for projector safety...
~
Most RGB color inputs for projectors with diode lasers these days are good to 10 KHz or more.. Often to 100 KHz, So other then clamping and making sure the color signal output NEVER goes more negative then 0.6V volts, you should be fine. Ie Schottky diode for a reverse clamp. When there is no color output, it helps to have a weak 4.7K pulldown to ground at the signal source. If the panic button is pressed, the RGB outputs should go off for a few modern projectors that omit the mechanical laser safety shutter. Which is most Chinese projectors, and they dominate the market.
~
Most synth users would probably be happy if you had an optional absolute rectifier, and multiple input summing amp for each of the color channels... As well as a way to add a DC offset to each of the final color outputs. And more then three color channels, serious abstract people are likely to have a second blue or red wavelength, or maybe an additional yellow channel for color.. At least four color channels is a good idea. In the old days, my projector supported eight wavelengths, and that is coming back as more diode lasers are developed across the spectrum. However the best software on the market supports just six color channels, so we tied a few channels together. Bump buttons to full 5V for each color for projector tests and effects to the beat of the music are always nice... Oddly when making abstracts, black/dark/blanking is a useful color.. Almost all projectors use a single ended 0-5V color signal, with 5V as full brightness.
~
ILDA's recommended output circuits are in the document I listed. TL082 or TL084 are the op-amp of laser choice, simply because of their internal +/-20 mA current limiting when driving the cable.. Most experimental products socket the output op-amps to make for a quick repair in the field.

~
It helps if the X-Y outputs have individual DC offsets, and a common image size control. An audio taper dual potentiometer makes a great image size control. The audio taper works better then linear, believe it or not. As you never know your desired image size until you get to the show venue, a master image size control is invaluable.
~
If your customers ever ask if they can use the DC offset control potentiometer built into the scanner amps, the answer should be no.
!

!
PS: FOR FUTURE READERS OF THIS THREAD: Setting the scanner amp damping to TOTAL zero is for Galvo master users during notch filter tuning only!!!!!!! Even then, near zero servo gain must be used...
!
Steve

[james]

Remember your ILDA connector must short pins 11 and 17 for the projector safety interlock... It helps if that interlock can be a "Big Red Switch" of the panic shutdown style, or "key Switch" to conform with international standards for projector safety...
Steve

That's 4 and 17.

Pin 13 is a +5VDC shutter actuator.

[mixedgas]

Whoops, corrected, thanks James...

[creatorlars]

However, the lowest frequency for an image to be flicker free to the eye is ~20-30 Hz. So if I divide 1000 Hz by 30Hz, I get 33... A 33:1 ratio, so if the user's base or master circle is 30 Hz, they have more then enough possible patterns to get started. A ordinary 6 dB per octave filter rolloff is fine...

OK, I'm following you. I have worked a lot with vector rescan of video rasters using XY displays and so forth, so the visual stuff I get. It's the mechanical galvo stuff where I've got a lot to learn.
So let me get this straight: A 6 db/octave lowpass with -3dB attenuation at 500Hz/1000Hz cutoff frequency, design for the least phase distortion as possible (a Bessel topology, maybe?) with possibly an all-pass network to correct phase errors after the fact. Identical signal paths each for X & Y.

Remember your ILDA connector must short pins 4 and 17 for the projector safety interlock... It helps if that interlock can be a "Big Red Switch" of the panic shutdown style, or "key Switch" to conform with international standards for projector safety...

I have the interlock A & B pins tied together, and a switch for either 0V to 5V going to shutter. On the laser scanner I tried (and blew up) last week, the shutter 0V vs 5V command didn't seem to do anything. Sounds like a big red safety switch for the interlock connection is a better method, and to just send 5V to the shutter pin via "digital output" standards as shown in ILDA specs.

My differential drivers for RGB are protected with a 5.1V zener and follow the ILDA diagram schematics. Two zeners each for the X & Y differential drivers. On this last proto, the zener clipping is on the output driver, but on this next version I'll place it before the filters. Master size controls I will place right before the differential drivers (so input+attenuator+DC offset --> zener clipping --> low pass filtering --> master size control --> differential driver.) I used TL07X series on this last proto, but will switch to TL08X for production.

I've been trying to sort out how to handle RGB. The laser I was testing with only responded to TTL control (on/off) per color channel. Is this typical? If so, I'll probably do a 10KHz-20KHz analogue ramp generator, with a pulse width modulator each for red, green, and blue outputs. That should provide a full range of hues. I have a few synth modules I designed for HSY to RGB colorspace conversion and color matrix mixing for video, which will provide lots of options for driving the RGB in a larger system. I reckon most of the users of my module will be buying low budget chinese RGB ILDA scanners like the one I tested with. (TVS VS-11, I think?) Am I wasting my time with PWM?

Full wave rectifiers for color is always a good idea. I do this on my video modules as well.

Kill color outs when panic/interlock switch/button is engaged... easy enough. I'll probably use a tri-state output driver if I'm talking TTL signals, and just set it to float when the switch is engaged.

Currently X, Y, R G & B all have DC offset controls, single inputs + attenuators for each input. I may consider more inputs, but in a modular system you always have access to lots of different mixers (like an RGB matrix mixer as mentioned above) so my main concern is the interface itself. I'd love to do an abstract generator/console type module release later on that could accompany this module, for those who want a "voice" with modulation inputs rather than patching directly from the array of synth modules out there, but that's a project for another day.

Master image size/zoom control with dual gang audio pot... deal!

[mixedgas]

Oh, you hit probably hit one of the few remaining Chinese "Inverted TTL" projectors that don't follow the standard and need a pull down to less then 0.6V on the RGB inputs. You perhaps got one of the new ones where they went to 3.3V chips on the inputs. If its any consolation, I tried adapting that design to the standard and found even with level shifters and optoisolation, the current mode logic they used is a disaster...
~
We don't know why they exist, but once a design gets built once in China.... It never goes away... Only gets cheaper and worse...
~
When we got to actual RGB modulation, you just asked me about something else I work on.. Good luck with your project..
~
Just make sure your target pricing is high enough, and your design is difficult enough to make, with copy proofing, that the Chinese don't copy you, or all this will be for naught...Within two weeks or less of going on the market.. Don't do the development work for them for free... In this Laser market they copy everything, to the point of pulling the lids off microprocessor chips and probing them...
~

Steve

[discothefunkyhippo]

I reckon most of the users of my module will be buying low budget chinese RGB ILDA scanners like the one I tested with. (TVS VS-11, I think?) Am I wasting my time with PWM?

I plan on purchasing at least one of your units and all my projectors have analog color modulation.

-Sal

[mixedgas]

PWM is a disaster with laser graphics. You'd be better off with a 555 tunable from 1 Hz to 4 KHz , followed by a divide by 2 to square it...
~
Funky, lets wait till we see it scanning, shall we? He has some stuff to learn, especially about targeting the low cost market...
!

Steve

[discothefunkyhippo]

Funky, lets wait till we see it scanning, shall we? He has some stuff to learn, especially about targeting the low cost market...
!
Steve

I am not sure if he is targeting the low cost market. He is targeting the modular synth market, which by no means is low cost. I have interest in these modules because I am collaborating with someone who has a modular system and would like to extend the possibilities of collaboration.

I already have this bad boy, but would like a few more input/ output options.



I actually contacted Lars about this module before he posted here about it. I even offered to test it for him, but he didn't even reply to my offer...

-Sal

[swamidog]

I am not sure if he is targeting the low cost market. He is targeting the modular synth market, which by no means is low cost. I have interest in these modules because I am collaborating with someone who has a modular system and would like to extend the possibilities of collaboration.

I already have this bad boy, but would like a few more input/ output options.



I actually contacted Lars about this module before he posted here about it. I even offered to test it for him, but he didn't even reply to my offer...

-Sal

bananas for scale?

[Draco]

I have been wanting to build some simple form of an analog generator for some older open loop galvo's but this turning out to much more complex then i had thought, just for basic patterns what type of signal generators should i be looking for or are there some online schematic that i could build my self for as cheap as possible?

I dont need differential inputs as the amp i am using for the galvo's is only single ended input, the amp's output to the galvo is differential though.
I normally just feed it stereo left and right but would like more complex patterns, the galvo's and big and slow and i cut out high frequency with a mini dsp board. this device is also for my own use i will not be doing public shows with it