Beam Director

[loopee]

Hey Planters

The gimbal motors might be a good option as you can get them in several sizes and can be quite cheap, however you may need to take into consideration the inertial mass or the rotor plus mirror

These guys have a selection
https://aerialpixels.com/product-cat...gimbal-motors/

Not sure the best method for driving them for this situation though

On a side note for gimbal controllers, the one on my quad will shudder slightly when the quad is sitting on a bench, so assume the PID loop error or something

[mixedgas]

Planters, I have three complete sets of 6350 large mirror galvos with amps. The amps are class I servo loops with adjustable slew rate limiters. This means they love to drive big mirrors. Have everything. Amps, PSUs and interconnects. AMPS have the 12 bit DAC chip with parallel input and by adding a jumper they switch to single ended analog. These things LOVE big mirrors. You could easily go over Cambridge's speced Arpeture size. My sets are from a layup machine for fiberglass, have spent their entire lives in clean enclosures. They guided a human in laying up airplane wings. I bought them used for a Canadian guidence system project that was unfunded half way thru.
Would just like to recover my money at this point.

Neat thing about a slew rate limiter, these galvos will try to scan the ILDA test pattern, and do so quite well. The limiter keeps them from scanning the portion of the image that is too fast to follow. Which means even with a huge Y mirror, these things will actually scan a decent graphics show.

Mill your own large mirror mount and be very happy.

Attached are some docs.

Steve

[planters]

Steve,
Wow! This will take me a little while to wade through, but your description makes it clear these would work. Short of bending the shaft, these will simply refuse to move too fast if I load too much mass or demand too much speed. I like that. Let me scan through this and I'll send you a message. Thanks.

[planters]

I want to bump this thread with a request.

I would like to try out my idea of a modified moving head mount as a beam director. I am looking for suggestions for a manufacturer or better yet, for offers from anyone that has a nice, solid, fast and functioning yoke style mount. I do not need the head or light source, the optics or any of the electronics that drive the head and it seems like a waste (and expense) to buy one of these things and scrap this. I can fabricate the mirror and its mount and the large size of this 2 axis mount means that I can remote this beam director from the laser which is much too heavy to move to say an elevated location and not worry that the beam has to be dead on center otherwise it will spill off the sides of the mirror.

[loopee]

Would this work

http://www.lasershop.de/second-hand/...irror-352.html

[lightlinked]

If 1.8 or 0.9 degree steppers with micro-stepping is good enough angular resolution for your application, two mirrors would most likely be less mass than a single mirror on a yoke or a yoke holding a fiber launch. that is, unless you want a very wide scan angle, then a yoke could do 360 on one of the axis. they real key would be accelerating and decelerating the mirror at a rate that matches the motor and moment of inertia of the mirror assembly to avoid missing micro-steps without encoders/servo loop.

a thing not on this CAD model would be hard stops for dumb homing upon initialization.

steppymirror-2000.jpgsteppymirror-2001.jpg

you could do this with surplus parts for pretty cheap.

[planters]

loopee,
That would work if the speed is acceptable. I don't know what the degree/second is. The mirror size is good so even if I have to replace it with a mirror having a higher damage threshold and it is a thick mirror, this should be able to move it.

lightlinked,
This should work as well. The torque of the steppers and the speed should be more than enough. The angles I want to cover are on the same order of a typical laser projector and the resolution should be a few mrad. The beam will cross free space to the tripod mounted beam director and there be reflected to the target. The motors would have to operate with micro-stepping or be geared because even 400 steps/revolution is 1.8 degrees with the 2:1 optical reflection. The hassle is I need everything. I need motors, drivers, a controller and mirrors and then I have to integrate it. This isn't impossible, but it is a fair project.

[mixedgas]

Anaheim Automation makes my preferred microstepping driver, cost is reasonable, and they are a stock part.

Watch out for motor/load resonances and the need to ramp the driver speed up and down. Steppers do not start/stop on a dime. They have a nasty nonlinear load/speed curve, and the thus torque falls off at high speed. (Which I am now fighting on a motion stage development project)

you might be much better off with a gear train between your motor and mirrors, with a brushless motor.

The Cat's Meow of fairly easy to use motion control chips are made by LSI:

http://www.lsicsi.com/products.htm

Steve

[lightlinked]

The two mirror option is more ideal if the laser aperture is close to the assembly since it is ideal to enter it at a precise angle. If you wanna hit this device from far away, a single mirror on a yoke becomes a little more practical. However, it will be slower on the axis that holds the second axis and mirror due to having to move the mirror motor, yoke frame and mirror. however, it is still quite quick.

are we to understand that this application has nothing to do with art/entertainment? ie are you trying to avoid using lightshow parts? ie. you want something more industrial or R&D oriented?

how are you going to "computer control" this? what are you doing for software?

[planters]

This application is not LIMITED to art and entertainment, but that's where I'm starting. I think heavier duty components whatever the source are more appropriate. Unlike a typical scanner application where the mirrors are light and fast this will be a "slew,stop, settle,pulse...repeat" at a rate from less than once/second to 4Hz. One way around the precision entry challenge is to fix both the laser and the scan head and then fire the laser at a fixed target and a known scanner orientation. This then establishes a parking position from which X and Y slews are measured.

I am planning to use open CV:
http://docs.opencv.org/trunk/modules.../tracking.html

Steve may have a good solution with the scanners because they are digital and hopefully can interface with a PIC or Arduino and I know that there are algorithms in open CV that will output a signal based on frame position to operate a camera scan motor. They could as easily output a 12V signal to the laser to fire after a sufficiently long period for the scanners to reach a target location.