My Custom Built Camera Crane

[Jeremy_Sawatzky]

Hi everyone i recently started my camera crane build. I had no design or plan drawn up whatsoever. I was literally sitting at the computer, thought to myself id like to have one of those and walked out the door. I drove to the metal shop.

Since i cant see a project in my head without seeing available materials, i stayed there about 2 hours drawing on paper, going over what i would need all thought up on the spot in my head. (needless to say im good friends with the metal shop guys now

after a purchase of about 110 bucks in aluminum i came home with the most crucial pieces needed to start my build. I also picked up a couple boxes of 1/4in bolts and nuts along with an assortment of larger bolts, washers, nuts and bearing assembly.

My available tools at home consisted of a drill press, table saw and ratchet set so on. No band saw

I decided on aluminum for the build due to its inherent lightness. Steel would have been nice to weld but much harder to cut, drill and so on. Not to mention the finished product would have probably weighed 100lbs. Aluminum is nice and soft, quiet and not so hard on the tools

I didnt really think at the start of the build to properly document how i built it but the pictures i have are pretty self explanatory.

I started out by building this: 2 plates 9in by 7in sandwiched together with a bearing assembly, and the start of a tower mounted on top. as u ca see im going to be using alot of bolts and nuts for this project.

then i built the tower that would hold the 2 arms and incorporated it onto my base, note the larger bolts that the arms will be attached with.

Then i took my 2 5 foot aluminum arms, drilled them to fit the bolts and made a coupling piece for the end. wow a camera crane lol - oh but im no where near done

my workshop

I had decided from the beginning i wanted about 9 feet of arm on the boom but at the same time be able to carry the rig around in a small car. So i set myself a maximum length of 6 feet.

The arms would consist of 2 5 foot pieces that could extend or retract to your desired length. Or you could remove the secondary set of arms entirely and just use 2 5 foot pieces.

So i set about drilling holes in both arms 3in apart

note: when working with aluminum do yourself a favor and buy a set of nice drill bits. Dont bother that $7 set on sale. Spend at least $35 on a set of titanium or $75 on a set of cobalt drill bits. They will will not warp or lose their sharpness.

because i drilled both pieces at once i know they are identical

add the secondary set of arms and 4 pins to hold it together and ive got my retractable 9 foot reach :thumbsup:

I couldnt help but toy with it - lots more work to come!

Next i needed to replace that table saw with a proper tripod. And something very solid. I chose not to incorporate a pre-built tripod as my base because with the amount of weight was putting on it, the legs of most tripods would flex and bend, not to mention ruin the head.

This is a good place to mention from here on was more trips to the metal shop and lots of trips to hardware stores for bearings, bolts and so on. About another $150 bucks

So i started with a simple plate - and lots of measuring - the 4 holes are for the 4 bolts that extend out the bottom on the base of the crane. need something to sit on!

Dont worry - these holes arent randomly placed :huh:

These angle plates extend 3 in downwards - it will make sense dont worry

Lots of those nuts and bolts

Then the 3 tripod legs get sandwitched between the plates with bolts.

The crane sits on the plate (there is now a holding pin so the 2 cant come apart accidentally)

Obviously the legs needed some holders so they dont go flying outwards

Its astonishing how solid the tripod is, Not 1mm of give in any direction.

I chose not to make the legs extendable (for now) because the crane itself does that and going any further down will reduce how high the crane can go (due to counter weights hitting ground) and going higher would make it hard to operate

Long list of things still to do. In this pic you can also see the new weight system being built

Guess what comes next?

Yep a full 360 remote head - Now it gets tricky

note: i had the controller and receiver assembly from a rc car. But it needed some extra parts to make it work for my application, different battery, speed control, gears and so on. about $90

Starting to build the horizontal axis assembly

cheesplate

Because you have to make the whole rig balanced, you need to be able to adjust the cameras center of gravity - vertical adjustment

Next is the Rotation assembly. All of the motors, battery, reciever, speed control will be mounted on the moving assembly. therefore there needs to be a fixed gear for the low rpm high torque motor to spin itself around

Where the camera will sit approximately

And thats how far ive gotten on the build. Lots to do still, im waiting on a couple parts to be machined and for the specialty low rpm motor to arrive.

here is a couple videos i took with my iphone mid build

http://www.youtube.com/watch?v=A0SQnDFInSg&feature=related

http://www.youtube.com/watch?v=q0-gRus9k50

Il be updating!

 

[shyguyfilms]

Nice work!

 

[MBPRODUCTS]

The iphone crane shot is amazing. great job.

 

[robmneilson]

Wow, very very nice. Wanna come build one of these in my garage?

 

[treynolds]

Great crane and i love how much cmos jello the iphone shot has.

But once again amazing build quality

 

[RileyRobertson]

wow, looking great! Let us know how the pan and tilt goes! Once you get that done, add Roll to it!

 

[Jeremy_Sawatzky]

wow, looking great! Let us know how the pan and tilt goes! Once you get that done, add Roll to it!

adding roll would be a huge task not because of the mechanical aspect but because adding a 3rd channel and control to the radio setup would cost alot of money and be overly complicated.

perhaps a cable setup could work tho...

 

[Jeremy_Sawatzky]

ive got a small hurdle to overcome. I have a fully wireless rig so far so the head can spin around all it wants. I want to have my lcd at the bottom of the rig though so i can see whats going on. Has anyone use one of the Brite-view hdmi wireless transmitters on the gh2 before? The transmitter is powered by usb so it would be no problem to power that on the remote head itself, and the reciever is 5v so no problems there. anyone done this?

http://www.brite-view.com/hdelight.php

 

[RileyRobertson]

adding roll would be a huge task not because of the mechanical aspect but because adding a 3rd channel and control to the radio setup would cost alot of money and be overly complicated.

perhaps a cable setup could work tho...

Or you can go and buy a cheap $100 4-channel system like I did(only for testing) but the real cost is creating your own Transmitter/ Reciever from scratch...I'm in the stage of development. Nothing but headaches, so I hear you on that being complicated.

 

[Jeremy_Sawatzky]

ok update

Today i built the "head" of the unit that will be remotely operated.

i started with a 1 and a half in square tube cut into 2 pieces which then needed to be joined at an angle

making the joint extremly strong so there is no flex

14 bolts total haha

on the bottom of the arm extending down, i had the square pipe precision machined to fit 2 bearings which would then hold the pivot axle that would hold the camera (u can see it in the background)

making this

then it was attached to the rig using a bearing assembly i made myself

it looks off in this pic but its just due to the angle, the head is perfectly level in all directions down to half a mm

http://www.facebook.com/video/video.php?v=10150588072060455#

tomorrow i start working on the electronics

 

[Jeremy_Sawatzky]

arg - this just got a whole lot more expensive.

i thought i could get away with a single motor providing the horizontal rotation and a servo setup to handle the vertical rotation. (motor hasnt come due to seller sending it to uk by accident....)

turns out the servo (though it has 180 degrees of movement) cannot provide a smooth motion whatsoever and to top it off, the way its designed means it returns to zero (i.e. back to its original position) after a motion.

Then to top it off, the existing gear setup i had on the top piece will not work due to the pinion gear not being close enough to the spur gear.

SO im forced to build a setup with some expensive new gears and different motors and another electronic speed controller

60rpm gear reduction motors coupled to 24T 48P pinion gear will turn around a 203T 48T Spur gear providing (at max speed) 1 full revolution of the head unit every 7.1 seconds. Coupled with a speed controller i should be able to get speeds around 5-10 rpm on the motor giving me 1 full revolution of the head every minute.

these are the motors im using
http://cgi.ebay.ca/ws/eBayISAPI.dll?ViewItem&item=360347383800&ssPageName=STRK:MEWNX:IT

and the gears
http://www.servocity.com/html/48_pitch_gearmotor_pinion_gear.html
http://www.servocity.com/html/48_pitch_aluminum_hub_gears___.html

and the speed controllers
http://cgi.ebay.ca/Ace-Veloci-Pro-R...Control_Parts_Accessories&hash=item4cf71280fa

2 of everything... lame

Custom motor mounts, and more machining...

hehe this is just more of an overview for myself...

 

[RileyRobertson]

Its looking good though! Nice work, keep it going. Are you going to make this a Wireless setup or a wired?

 

[Jeremy_Sawatzky]

Its looking good though! Nice work, keep it going. Are you going to make this a Wireless setup or a wired?

This setup is going to be wireless. Unfortunately i still have to run hdmi out on the cam and that will have a cable running to the base. Ive been looking at wireless hdmi options but that just means more batteries and more electronics.

 

[Paul F]

Jeremy,

i've just finished building one and I have some suggestions.

The motor you have selected is 12V, but the controller is rated for 8.4v max. I suggest you consider the Jrk series. http://www.pololu.com/catalog/category/10. They are programable and you can set input scaling, output scaling. It also has a graphical interface to show you wnat the joystick and motor are doing in real time. That is helpful in tuning the controller. Once set, you can disconnect the PC. About the same price as the one you selected.

60 RPM will be ok for the fast speed, but unsatisfactory for slow. These motors do not start moving until there is about 15%PWM. I bought a 60RPM (too fast at the slow end), a 6 RPM (too slow all around), and now a 12 RPM (not tested)). None are satisfactory at the slow end due to the fact that the motors do not start moving until the PWM is at 15%. I don't know how zoom lenses do it. They can creep at the very slowest rate with tiny motors.

I'm using a 120 tooth hub gear and the same pinion gear (same pitch).

I've talked to companies with much more expensive motors and planetary gears and they get the same results; they need about 15% before they start moving.

 

[RileyRobertson]

Yea I here ya on the batteries. My setup is going to take 6! But I'd rather have to swap out batteries(when they die), then have a bunch of wires everywhere. And with something like a Pan/Tilt it doesn't drain your batteries quick.

 

[Jeremy_Sawatzky]

Jeremy,

i've just finished building one and I have some suggestions.

The motor you have selected is 12V, but the controller is rated for 8.4v max. I suggest you consider the Jrk series. http://www.pololu.com/catalog/category/10. They are programable and you can set input scaling, output scaling. It also has a graphical interface to show you wnat the joystick and motor are doing in real time. That is helpful in tuning the controller. Once set, you can disconnect the PC. About the same price as the one you selected.

60 RPM will be ok for the fast speed, but unsatisfactory for slow. These motors do not start moving until there is about 15%PWM. I bought a 60RPM (too fast at the slow end), a 6 RPM (too slow all around), and now a 12 RPM (not tested)). None are satisfactory at the slow end due to the fact that the motors do not start moving until the PWM is at 15%. I don't know how zoom lenses do it. They can creep at the very slowest rate with tiny motors.

I'm using a 120 tooth hub gear and the same pinion gear (same pitch).

I've talked to companies with much more expensive motors and planetary gears and they get the same results; they need about 15% before they start moving.

thanks for the suggestions, i will test my setup (when its built) with the speed controller i already have. and see how that performs. I tested the controller with a 12v 500rpm motor i had and it turned quite slow when i put the least power too it. If the 60rpm motors start turning at 15% that means i will get around 9rpm out of it. Which is perfect with a 24t pinion gear and 203t hub gear that will make the rig turn at 1 rpm. exactly where i want it. Although its no big deal to buy a slower motor. I dont think il have any problems with the controller but rather with the radio im using. Im afraid its signal is too coarse.

now the jrk controller, the analog control, could i just hook 2 of those up to a joystick like this?

http://www.servocity.com/html/2_function_joystick__thick_sti.html

that would be probably the nicest setup. But at the same time i might be going backwards by using such a simple potentiometer with an esc. and another question, what is input/output scaling?

if you look at this rig

http://www.servocity.com/html/pt-2100_accessories.html

they are using 20rpm motors with 10t pinions and 60t hub gears. there is video of it in action and it is butter smooth. I did the calculations and while 60rpm might be a little faster with my gearing, it will be pretty darn close.

i think the reason the other companies can get such smooth motion is due to their controllers, the motors they are using probably have a stall voltage around 5v. The motor wont move at that voltage but its just about to. with a proper controller set to operate at 5v + you could get very smooth motion.

 

[Bante]

Nice build, Jeremy!! Looks great!

I built a similair jib with motorized head a couple of years ago, and one of the issues I had was the backlash of the gears. I tried different solution and settled for a frictionbased drive, where the motoraxle was just pressed into a big rubber wheel mounted on the pan or tilt axle. The motor itself had a planetary gearbox so the inherent backlash in the motor was minimal. I controlled everything with a basic stamp 2, and it was pretty smooth.

Here's some pictures, sites in swedish.

http://bantheboy.se/kran.html

I'm actually thinking about picking up this project again and try stepper motors.

How do you attach your bearings? Are they pressed in? No chance they can fall out?

 

[Paul F]

....
now the jrk controller, the analog control, could i just hook 2 of those up to a joystick like this?

http://www.servocity.com/html/2_function_joystick__thick_sti.html

Yes, that would work well.

that would be probably the nicest setup. But at the same time i might be going backwards by using such a simple potentiometer with an esc.

I didn't follow this. What is esc?

what is input/output scaling?

A joystick as you probably know is two potentiometers. They provide a voltage reference to the motor controller (typically 0-5v). The Jrk has a learn mode where you set the joystick full left, then full right, then center. It then learns what the input voltage is for those 3 settings and uses that to scale what each voltage represents. You can then set the scale of the output; how much voltage the controller will put out for a given input voltage. This let's you make sure you take full advantage of the joystick movement and the motor's capabilities. It also makes sure that the motor is stopped when the joystick is in the center position. You can also override the learn mode and tune the input and output.

if you look at this rig

http://www.servocity.com/html/pt-2100_accessories.html

they are using 20rpm motors with 10t pinions and 60t hub gears. there is video of it in action and it is butter smooth. I did the calculations and while 60rpm might be a little faster with my gearing, it will be pretty darn close.

i think the reason the other companies can get such smooth motion is due to their controllers, the motors they are using probably have a stall voltage around 5v. The motor wont move at that voltage but its just about to. with a proper controller set to operate at 5v + you could get very smooth motion.

Smooth is not an issue. You will have very smooth motion with your setup if you have nice bearings with no slop in tolerances of the fittings.

You probably know this, but for everyone:

Motor controllers work differently than servos. A servo joystick position represents a specific amount of rotation of the motor. The farther you move the joystick, the more the motor moves clockwise or counter clockwise. The controller sends a pulse whose width that is proportional to the desired position. Then there is a built-in potentiometer that the servo reads and it keeps moving until the POT is in the position that the pulse width represents. Servos have a somewhat jerky movement.

The position of the joystick for a motor controller represents speed, not position. The farther you push it, the faster the motor goes (there is an exception I note below). Motor controllers send a pulse whose width that is proportional to the speed desired rather than the position desired. That pulse is repeated at a very fast rate (5khz or 20khz in the case of the Jrk) and is in effect, an analog voltage to the motor. Pulses are used instead of a DC voltage for efficiency. So you should be happy with the results you get from the controller you purchased.

The issue I'm having is that I want to be able to start a head pan move (with the jib stationary) with a very smooth rate of change. This is not possible. There is always a sudden start to the move as the controller gets to the 15% PWM point and the motor starts moving. If the motor started moving at 1% or 5%, then the start of the movement would be much more subtle and not have the sudden start of motion.

I also want to be able to creep very slowly with a long lens. For my needs, 1RPM is not nearly slow enough even for a medium focal length.

One more thing - the Jrk and other motor controllers have an optional feedback input that let's you operate them similar to a servo, but with smooth movement. You add a POT to the thing that is moving to indicate position (the pot is rotated with the pan or tilt). The allows you to have repeatable motion control that can be driven by a PC for special effects such as shooting a background plate and then repeating the move with the foreground. That same technique could be used to control the pan and tilt of the jib (with much bigger motors and gears) for a full motion control rig.

 

[Paul F]

Nice build, Jeremy!! Looks great!

I built a similair jib with motorized head a couple of years ago, and one of the issues I had was the backlash of the gears.

I was concerned about backlash. With the brass pinion Jeremy is using and a nylon hub gear, I am able to achieve zero backlash.

I tried different solution and settled for a frictionbased drive, where the motoraxle was just pressed into a big rubber wheel mounted on the pan or tilt axle. The motor itself had a planetary gearbox so the inherent backlash in the motor was minimal. I controlled everything with a basic stamp 2, and it was pretty smooth.

The nice thing about your rubber wheel solution is that you are not concerned about stripping gears with accidental hits to the head.

How do you attach your bearings? Are they pressed in? No chance they can fall out?

For my head, I bore out the holes about .001" oversize. I turn the shafts to a hand-press tolerance. I can hand press the bearings onto the shaft. They don't move.

 

[Paul F]

Here is a sample of my jib with no head pan/tilt movement; just jib pan/tilt. You will notice that the camera is not level when at the high point. It also wobbles a bit. This was my first test. I found out that the stabilizer that holds the head level had an interference that caused the whole head to tilt when it was up near the top. So the whole head is at an angle and wobbling around. I've fixed it, but have not run another test.

In that shot, it is going 16' high, 14mm focal length (28mm equivalent on 35mm camera)

http://www.youtube.com/watch?v=qCx0azIM1e0&feature=channel_video_title