remote focus

i like the design but you need to think about where you will be housing the power for this unit. I would also think about some form of support for the lens. Adding all of this onto a longer lens will put a lot of strain on the lens mount and tripod mount on the camera which they were not meant to hold. It will also put more weight on the front of your rig making it front heavy on a tripod. If you were mounting the camera rig with an extended lens support it would also help reduce any vibration from the two servos.

I would also think about the distance between the two moving rings. In your mock up it looks like you have a 70-200 but think about the distance between these two rings on other lenses. It wouldn't be too hard to slide the servos forward and back a little bit but i don't know how much room for adjustment there would be.

The only other thing i have heard when looking into similar products is that a lot of servos for remote control cars, plains, etc. are rather loud and would kill audio in most situations.

Just food for thought, overall i think its a really rather elegant design
 
Some thoughts on your design:

RC servos are noisy. Is that an issue for you?

How are you going to mount an .8mm pitch gear to the servo? I've not found a way to do that the last time I looked a couple years ago.

RC servos tend to be twitchy and may not be smooth enough for focus and zoom.
 
Well. thanks for the feedback. to answer your questions, somewhat in order :
1- the power unit will be a small 1000mAh battery, about 200g, for which i must say i haven't thoght of, but i will now. the rest is only a small receiver which weighs abut 10g
2- the overall weight of this should be no more than 200 g, for all the parts that go on the lens, do you think that is too much strain ?
3- the servos don't vibrate that much or twitch, expecially if the reduction is high, at least 2.5/1 .
4- yes, there will be some issues with sound if the camera is closer that 2m from the target, so this sistem won't be so good for close up shots. I will consider a design that uses polulu micro gear motors, which are really quiet, but on the downside is more expensive.
5- for mounting the gears to the servo, i was thinking on using custom made gears, glued or bolted on the small discs which come with the servo.
6- yes, it will probably not fit every lens, so i will probably end up making two of the motor mounts, one for focus, one for zoom, which will add another 50g tops.
7- and before i forget, if i will go on with the servos, i will use digital servos, which are more precise, i.e. less twitchy than the usual servos.

And i also have a question. My design has a range of 50-80 mm for the diameter of the lens, is 80 mm large enough ? thanks.
 
Last edited:
The servos big advantage is that lens focus point proportional to the POT position, which is very intuitive. It allows you to make a big focus dial with variable stops or a marker that behaves just like a regular focus puller. You could even calibrate the dial for each lens and put footage marks on it. You can't do that with the Pololu motors unless you add a feedback POT.

I hadn't seen the Pololu micro gearmotors before. Nice.
 
I started searching for the micromotors from pololu, and i found some nice ones on they're site, on which i can put an optical encoder, and so it will behave like a servo, i.e. with feedback, but with more than 90degrees of rotation. I will hook up the motors and encoders to a mini arduino, and at least for now in my head, the range will probably be controlled with a pot on the camera. I was thinking on force sensors, but i think i will just complicate the thing unnecessarily. I'l start designigning tomorrow, because today is sunday :)), and probably in a week or so i'l have the mechanics figured out. Oh, and i'l scrap the zoom+focus, and make a 1mount-1motor deal, and that coupled with carbon fibre, will be very light. Also the weight of the motors is almost nothing, so i don't think it will be a problem for the lens. And please answer my question about the max. diameter of the lenses, because as I said, i don't know if 80mm is enough. Thanks.
 
Good choice. If you want to do this right, your own custom design with motors and feedback is the way to go. This gives you control over rpm, torque, power consumption etc.
 
i would love to see your arduino code when you get to that phase.
After i saw this post of yours last night i started looking around and found this. I thought you might like to take a look. The build isn't as sophisticated as yours but the code is good.
instructables
http://www.instructables.com/id/Arduino-Lens-Controller-Synchronized-ZoomFocus-/
code
http://playground.arduino.cc/Main/WiiClassicController10

So far as the weight goes i was thinking about it and here is my logic:
when using a long lens that has a tripod collar you don't mount the camera but instead mount the lens, this means that the weight on the mount is not that of the lens but the body. Now these are designed to work with cameras up to the size of the 1d which are significantly heavier than smaller cams like the 7d. That being said, assuming you are adding 200-300 grams to the lens it shouldn't be a problem for the LENS side of the mount. The other half of that mount is built however strong the camera body is. On the t2i body i would trust it a lot less than the 7d. The 1d series bodies are build to withstand the weight of that camera so it wouldn't be a problem.

The other thing to think about is the weight at 200-300g is also dependent on how far out away from the mount it is mounted. I would go to your local camera shop and check out some filters and lens hoods, start to see how much those weigh. I would bet my cokin system with 3x 3"x4" plus a hood filters would be about 150g. I can weight those at work tomorrow if you are interested. I just don't have my filters here.


tl:dr
the weight is probably ok.
 
thanks for the weight and code tip. i think i can calculate the weight this week, but a quick calculation says : 10g 1motor, 15g 1 carbon clamp, 10g 1 large gear, 10g 1spacer, 4g 1small gear, and 10gmax wires, comes to a total of 130gonly mechanics( the receiver, battery and controller will be somewhere else, probably below the camera, or attacheable to a rod), +-10g. I need to make exact drawings to figure that out. Btw, that code looks rocket science to me :)), i think i need something less complex. I am making now a diagram to see what i need to connect where, and because i'm a noob on arduino, i will probably swear a lot until i get the code working .
 
as somebody who is just starting with arduino myself i can promise you that much swearing, and frustration. But a lot of fun. I made a laser trigger, or i am working on a laser trigger for water drop photography
 
One thing you may be disappointed with about gearmotors and PWM (if you use PWM) is that there will always be a slight jerk to the start of a zoom. PWM doesn't get the motor moving until the pulse width is about 15% of max. So it has a little jump when it starts. If you find a way to get rid of that, please let me know.

Please keep us posted on your project.
 
Hy. I am studying now the arduino documentation, and don't really know yet the details. But the motor will be hooked up to a motor shield, and i will somehow have to write the code do give variable pwm to the motor, in order to sync the input from the motor's encoder, to the input from the transmitter. I believe it will use variable pwm to achieve the counts coming from the encoder, and fast reverse if it overshoots, or just code in some logarithmic progression for the pwm. On the other hand, last night i couldn't sleep and I had a conceptual breaktrough, it is a bit weird, so bear with me on this : as i started designinng the focus device, i checked on the net to see some examples, and found a really nice one, from hocus focus( video here : http://nofilmschool.com/2013/04/nab-video-hocus-wireless-follow-focus/) on which i found the auto limit detection extremely nice. So,i've given it a lot of thought, and nothing came up, until last night, when i remembered that every motor, has a specific amp consumption for a given amount of torque, lets say a micromotor uses 70mA for free run, but if you stall it, it can use even 10 times as much, and this is (i hope) the clever part: i want to use an amp monitor at the start of the code, so that when the code runs to make the limit detection on the lens, the motor just goes and goes in one direction, wuth the amps beeing measured constantly, but at a low rev, and when the motor stalls due to the lens limit, amp consumption rises, arduino detects it, shuts down the motor, reverses it by 2 mm, and records the position as a new limit. The proces repeats itself for the other end of the lens, and i found a piece of code that maps out the new range to the input range of the transmitter, and in theory, it all works just fine. If I didn't make myself understood, someone could translate this into plain english :)).
 
Yes, that should work. My mill does the same thing. It has very big motors. If it sees anything funny going on with the feedback loop (such as the encoder gets too far behind the command to move) or if it sees the current go past what it should be based upon what it is doing at the time, it will shut down.

If you really want to do it right, don't use PWM. Use a variable DC voltage. This is what the Fujinon and Canon broadcast lenses do. That's the only way I know of to get that really smooth and slow zoom. I've played a lot with gearmotors and PWM. I've never gotten them to behave correctly. I've talked to manufacturers of high quality DC motors. They said the same thing I have found. You can't get the motor to start smooth with PWM. I've taken apart Fujinon 'servo' lenses. They drive the motor with a variable DC voltage to get it to start out really smooth and to get it to creep along as slow as you like.
 
Thanks. I'll look into the variable voltage drivers. But if i don't find some driver compatible with arduino, i'll use the pwm type, because i don't know how to build one. Although i believe a pwm can be used in the following scenario : lets say want a slow zoom, and the arduino sends 40% pwm to turn the motor 50 steps( the gearbox ratio is 100/1, with encoder mounted on the back of the motor, the pololu ones i've found). i'm sure there can be a way to make the arduino move the motor slowly without jerking, just by rapidly stopping the pwm. well, a lot more documentation required, but i'll sure look up the variable voltage.
 
Those encoders are nice. Pololu didn't have those when I was working on this. I just drove straight PWM without feedback. You may be on to something. Give it a go and tell us how it works. That would allow you to do all kinds of nice things - software driven focus stops, pre-programmed zooms, etc.
 
well, probably they didn't have them, but encoder driven motors have been around for a while, although mostly found in expensive cnc servo motors. anyway, i found this while searching for the variable voltage regulators(which i haven't found, only some big boxes which don't really work with a small camera on a crane or whatever). I think there is a 95% chance i'll go with pwm drivers, since they can be found on every fence(romanian expression, don't really know how it translates), and are relatively cheap 25-60 usd. so if a microcontroller+a crappy encoder can do this(watch the video on the guys site, it's somewhere on the bottom part), i'm optimistic that i can get it running. hopefully in a month i'll have enough $$$ to buy the parts, and test some code. till the, i'll keep you updated on the progress, and as always, keep the constructive arguments flowing.
 
Come to think of it, i could use the same encoder technology to fully automate my cablecam, and probably connect it live to a laptop and control it in blender via wi-fi interface. the optimism of you here gives me good ideas :smile:
 
I would think that the code for the Arduino has been written several times for robotics. It's just a matter of finding someone willing to share the code to get started. Then modify it to suit your needs. I'd go to the robotic forums to see who might be willing to share their code.
 
You must log in or register to reply here.
Back
Top