What bracket & servos should I use for a Servo Controlled laser?
I want to build a laser pointer connected to a pair of servos allowing for pan and tilt.
I know there are many places to buy servos but I am interested in a high level of accuracy.
I am looking for an accuracy of roughly 1mm at distance of 1 meter.
Weight of the laser pointer is negligible.
Any suggestions?
Design your mounting mechanism, calculate the angular accuracy required from the servos to achieve that positional accuracy, look for servos capable of providing that accuracy. Note that if the angular range of movement of the pointer is less than the angular range of movement provided by the servo then you could reduce the accuracy requirement of the servo by gearing/leverage in the mechanical linkage.
I am really not interested in fabricating my own bracket system for this project.
Does anyone know of a good manufactured bracket system that would fit this project?
I would need 1080 degrees of motion horizontally and vertically.
I really only need 90 deg of motion horizontally but 180 deg would be nice for future use.
Also what vendors give angular accuracy specs for their servos? Vendors I have found didn't provide that sort of information. Any advice where to look?
180 degrees is pretty much the full range of travel for a conventional model servo, so this implies that you'd need a 1:1 drive. I'd suggest the 'straw man' design could just be the laser stuck to a servo horn by double sided foam backed tape, and the case of that servo stuck to the servo horn of another servo in a similar way. That would leave you looking for servos with a positional accuracy of 1mm at a 1m range which corresponds to 1/150th of a degree, which I'd suggest is completely unrealistic for a hobby servo. This implies you need to be looking for a completely different design. The sort of mount used for a CNC telescope is the first thing that springs to mind, but they have nothing like the range of travel you're asking for. So you're going to be looking at a substantial design and fabrication job. Do you really need that range of movement and such fine accuracy?
Also what vendors give angular accuracy specs for their servos? Vendors I have found didn't provide that sort of information. Any advice where to look?
You might want to check the people at servocity. For some $$$ they might build you a custom pan/tilt to your specs. For high resolution, you might need to go with geared stepper motors like in high resolution printers.
I've got one of these from DFRobot, see below. The servos are controlled by a 2-axis joystick using a modifed version of knob, and the joystick button switches trhe laser.
Are you projecting the laser onto a plane? Because if you are, remember that the change in angle will not bring about the same linear change along the plane as the angle varies and to get your 1mm accuracy you'll need to calculate what that change in angle will be as the angle itself changes.
Also when the beam isn't normal, you'll get an oval dot and that might be construed as inaccurate....
a positional accuracy of 1mm at a 1m range which corresponds to 1/150th of a degree
I make that almost one order of magnitude out - more like 1/17th of a degree
Still probably unrealistic for a hobby servo, except over a very narrow range, with gearing.
I make that almost one order of magnitude out - more like 1/17th of a degree
arctan 1/1000 = 0.058 = 1/17 yep....
But that's when the beam is normal to a plane. Sketch attached, which I drew with a 30 degree angle since all I could find in the house was a zillion set-squares and no protractor, shows the angle produces a much magnified distance on a plane at greater angles.
And just for the hell of it, I attached the Fritz of my set-up.
AWOL:
I make that almost one order of magnitude out - more like 1/17th of a degree
You're right. Hey, I'm a mathematician - don't expect me to do arithmetic too. 
I only estimated it first - the way I do it is to complete the circle and imagine the "seconds" on a clock-face.
They're six degrees apart, and a 1m radius clock has roughly a 6.25metre circumference, so the seconds are a little over 100mm apart, so one degree is about 16mm, so one millimetre is 1/16th of a degree, very roughly.
Convoluted, but it works and it is easy to do in the head to get into the right ballpark.
I just divided 1000 by 60, which would have been a much simpler way to get the answer if I hadn't dropped a zero along the way.
Hi, There's a low-cost unit ($10 for brackets and two servos) here: http://yourduino.com/sunshop2/index.php?l=product_detail&p=250
But same issue with accuracy. What is your (probably interesting) application that needs that accuracy?
I've used the one above for ultrasonic sensor "look around" but accuracy doesn't need to be good for that.
One of my students will be putting a laser on one of these next week, and I'll check accuracy and repeatability.
DISCLAIMER: Mentioned stuff from my own shop...
terryking228:
Hi, There's a low-cost unit ($10 for brackets and two servos) here: http://yourduino.com/sunshop2/index.php?l=product_detail&p=250But same issue with accuracy. What is your (probably interesting) application that needs that accuracy?
I've used the one above for ultrasonic sensor "look around" but accuracy doesn't need to be good for that.
One of my students will be putting a laser on one of these next week, and I'll check accuracy and repeatability.
DISCLAIMER: Mentioned stuff from my own shop...
Wow that is a killer great price, two servos + brackets for $10!
Hi Terrry
Lefty
PeterH:
I just divided 1000 by 60...
Hmmm, you'll need to explain that one to me Peter
JimboZA:
PeterH:
I just divided 1000 by 60...Hmmm, you'll need to explain that one to me Peter
It's making use of the "1 in 60" rule for small angles: Tan of 1 degree is approximately 1/60 so a distance of 1mm at a range of 60mm corresponds to an angle of 1 degree. Increase the range by a factor of 15 and you reduce the angle by the same factor. It's very simple and easy, but unfortunately it seems that in the wee hours of the morning my mental arithmetic wasn't up to working out 1000/60.
It's making use of the "1 in 60" rule for small angles: Tan of 1 degree is approximately 1/60
Ah, right, relying on the fact that for small theta, tan theta = theta.
I am attempting to build a 3D Laser Scanner to scan a multitude of objects.
Precision is actually more important than accuracy.
I plan to use a line laser but the width will limit my accuracy.
I am not entirely sure the accuracy I will need since it won't be a flat plane that I am scanning.
Objects shouldn't be larger than 8x8x8in.
Distance will be somewhere around a yard depending on the focal length of the lens I use.
I imagine I need something better than hobby quality?
I thought that originally you were looking for 180 degrees travel on one axis and 90 degrees travel on the other. You don't need anything like that to scan a 200mm object at a range of 1000mm.
The greater the travel you need the harder it will be to achieve the resolution, so try to work out what you actually need.
PeterH:
I thought that originally you were looking for 180 degrees travel on one axis and 90 degrees travel on the other. You don't need anything like that to scan a 200mm object at a range of 1000mm.The greater the travel you need the harder it will be to achieve the resolution, so try to work out what you actually need.
How much travel do you think I would need?
