There’s something magical about a laser light show. Watching that intense beam of light flit back and forth to make shapes and patterns, some of them even animated, is pretty neat. It leaves those of us with a technical bent wondering just exactly how the beam is manipulated that fast.
Wonder no more as [Zenodilodon], a working concert laser tech with a deep junk bin, dives into the innards of closed-loop galvanometers, which lie at the heart of laser light shows. Galvos are closely related to moving-coil analog meters, which use the magnetic field of a coil to deflect a needle against spring force to measure current. Laser galvos, on the other hand, are optimized to move a lightweight mirror back and forth, by tiny amounts but very rapidly, to achieve the deflection needed to trace out shapes.
As [Zeno] explains in his teardown of some galvos that have seen better days, this means using a very low-mass permanent magnet armature surrounded by coils. The armature is connected to the mirror on one end, and a sensor on the other to provide positional feedback. We found this part fascinating; it hadn’t occurred to us that laser galvos would benefit from closed-loop control. And the fact that a tiny wiggling vane can modulate light from an IR LED enough to generate a control signal is pretty cool too.
The video below may be a bit long, but it’s an interesting glimpse into the day-to-day life of a lighting tech. It puts a little perspective on some of the laser projection projects we’ve seen, like this giant Asteroids game.
Thanks to [Alan Green] for the tip. He found it while researching his own laser galvo clock project, which you should definitely check out.
We use a lot of galvos at work. The expensive kind. Like $10k for a galvo. Nice galvos use moving coil instead of a permanent magnet on the armature and have either a capacitive positioning feedback or sinusoidal quadrature encoder feedback. Some are even water cooled. We have to move very fast and very accurately.
I feel like I need a tetanus shot just watching this video. Filthy.
Good info though.
I’m loving the laser damage to his camera chip, as well.
Covers each eye. Nope. His cameras fine. It’s just my eyes.
I had to get a new pump for my washer. That one plus another from a friends washer and I could make an X-Y mirror set. Well maybe and slower than most for sure. They are a 2 pole magnet and an outside core and single coil motor. Though they are 110volt, 12 volts DC makes a decent deflection angle if sprung to center. I am thinking about a mirror large enough to aim a projector at close range to move the image around. Bass audio amped in can do some interesting A/V animation effects.
Have fun with these motors, they are everywhere clothes are washed in machines. You could also make a giant VU meter out of one.
Very cool.
Back in the arcade days, I cannibalized a Pioneer LD1100 that had served it’s useful life in an arcade cabinet( Dragons Lair, Cobra Command, M.A.C.H. 3, etc. don’t remember which.) Netted me some pretty cool components. He Ne Laser tube and power supply, some cool beam splitters, first surface mirrors, cylinder lens, coil actuated focusing lens, practically a laser optics bench minus the bench! Of particular interest were the tracking mirrors, seen at 3:44 in the video above. Moving coil instead of moving magnet meant they were quite light and fast. Of course, I knew what must be done. I pretty much immediately connected them to stereo outputs, turned down the lights, and sat for hours watching the laser describe every album I had on my wall. Acoustic guitar pieces seemed to have some of the best patterns. A little smoke in the room made for some extra fun.
In the late nineties, when affordable laser diodes became available, I built a dedicated laser display appliance out of it. The base had a battery compartment, audio input connectors on the back, a pair of potentiometers on the front to control levels on each axis, and an articulated arm on top holding the laser and and tracking mirrors in their own separate enclosure, so you could set it down pretty much anywhere and point it at any wall. I was quite proud of it. It had a nice, lab appliance aesthetic to it, and was quite enjoyed by all.
My bachelor party would be it’s downfall.
We had it at my best man’s house. I connected it to his 1970s era Yamaha receiver/amp and trimmed the pots to an acceptable level for the output audio, and let it play. But my friend is one of those fellows who, as the evening wears on, gets to a certain point of saturation where he feels the need to run and increase the volume every so often, especially if some particularly nostalgic, glory days piece of music begins. This is where you begin to see the Achilles heel of the moving coil mirrors. The magnet wire attaching the coil to the base can flex only so many millions of cycles, and when you raise the amplitude beyond the rated duty cycle, that number of cycles decreases dramatically.
The next morning it was only capable of displaying a red dot.
I finally got around to attempting a rewind of the coils a couple of years ago, but the finest wire I had on hand proved to be too thick, and it’s gone back of the shelf since. I have an idea to make some balanced armature mirrors using some earbud drivers that have a habit of collecting in drawers around here. Might make for an interesting write up!
I have one of those Videodisc lasers with their X/Y mirrors, cool. I mated it with a 12volt car stereo amp of 3 watts or so. So whatever gets fed into it that won’t happen. Line feed it. I often used a stereo jambox and miked the live sound with it and got a great effect. Use a peak limiter or compressor for better effect. The jambox had that dreaded auto-level control which worked great.
The best way to drive those coils without motion feedback is with current drive. Lookup hotsprings reverb, as their coils are funky as well to drive with a voltage and not have the response go allover the place.
The ultimate plan is to include a selectable signal generator on each channel and use the beam splitters and polarizing filters to switch the path between the tracking mirrors and a pair of rotating mirrors. That way you can independently tailor each channel with a choice of cyclic vs reciprocal scanning and to inject a waveform to one or both. Add to that, a wheel with different diffraction gratings at the output end. I also picked up one of those Xmas laser projectors from the local big box store when it was on closeout pricing, so now I also have some pretty strong R, G, and B lasers to play with!
I had somehow missed [Zenodilodon]’s YouTube channel. Good stuff.
Thanks!
Are all the drivers designed for analog voltage input? I have an application to use step/dir/enable to position a laser at specific XY coordinates…