Jump to content

DIY Crystal Sync Motor?


Jay Taylor

Recommended Posts

Hey guys,

 

So this is probably over my head, but has anyone ever built their own crystal sync motor?

 

I have a Fries Mitchell, so I'd like a motor that could do 1-120 fps, all crystal speeds. How are most motors built? Do most use some sort of DC servo's, or something else?

 

Like I said, probably over my head, but I'd still be interested in learning how motors are built.

 

Jay

Link to comment
Share on other sites

  • Premium Member

Bruce McNaughton built my Fries motor. It is muliti-speed in 100ths increments. So, it can handle all those strange flicka' (you know, my friend, Flicka?) speeds. As I understand it, it's all servo motors and stepper motors these days with digital controllers. Clive Tobin makes motors and controllers. I assume he's digital since his motors have precise control resolutions. I do know digital is easier and has more, precise controls than crystal.

 

I guess I should ask, how handy are you with motors, controllers, writing controller code, machine tools, soldering wire, etc? Because these kinds of projects can get pretty involved. Another thing is, accept that it will take a few prototypes to iron out unforeseen problems. Guys like Bruce have piles of conceptually flawed prototypes laying around. They don't let a product out the door until it has been refined to a well tested and working model. I have a small pile of parts and pieces to my DIY scan rig in a box from ideas that just didn't pan out.

 

I'm not telling you, "Don't do it," because I'm a, "Give it a try," kind of guy. But I am saying you could save yourself a lot of head scratching and cussin' if you get something already worked out by someone else who does this stuff for a living.

Link to comment
Share on other sites

Hello,

 

Paul, I sent you a PM.

 

The only thing I'd really be worried about is having to program the controller. As far as just hooking up a servo to the camera to make it go, it's probably not too big a deal. But programming it for different speeds, intervalometer functions, and all that? no clue.

 

I know I sound naive, but I'm sincerely interested in figuring out how this stuff works.

 

 

Jay

Link to comment
Share on other sites

  • Premium Member

I did build a demonstration unit about 30 years ago. IIRC, I just barely got it to work once before I fried something in the electronics. Back in those days, the idea was to use a CMOS 4046 phase locked loop chip to lock the frequency from a light chopper disc on the shaft to the crystal by pulse width modulating the power to the motor. So, you had to run the motor thru some hefty switching transistors. IIRC, I used 9.6 KHz as the pulse rate, so the thing made audible noise under load. The tradeoff is that going to a higher frequency means switching more often and making more heat in the power transistors. Going below the audible range would give you mechanical chatter.

 

I started from the schematic for an early CP crystal motor for the Arriflex. I just happen to have pulled the paperwork out on that, since I got a request for the schematic. They ran at 576 Hz and made their own PLL out of a quad nand gate.

 

Today, stepper motors could be driven at the chosen rate without the complexity of a servo. In any case, a crystal will still be your frequency standard.

 

 

 

-- J.S.

Link to comment
Share on other sites

Hey guys,

 

So this is probably over my head, but has anyone ever built their own crystal sync motor?

 

I have a Fries Mitchell, so I'd like a motor that could do 1-120 fps, all crystal speeds. How are most motors built? Do most use some sort of DC servo's, or something else?

 

Like I said, probably over my head, but I'd still be interested in learning how motors are built.

 

Jay

 

I upgrade of many type of DC motors of cine cameras on crystal sync speed.

Give me more technical inforation about your motor, pictures ( on personal e-mail ).

What kinds of type of motor ?

The motor have any feed back ? digital, optical, analog, taxo ?

What kinds of camera you will use ?

You have any mechanical parametetrs and characteristics of camera ?

 

P.S

The Micropower Phase-Locked loop chip 4046 - the design of past century.

Yes, this is very good chip and can be use, but, the procedure of signal conversion of chip included many procedures of calculation of analog quantity.

The modern algorithm of PID speed control inclued more wide rate of parameters and data and have more high accuracy of control.

This is like, if i will compare of compact calculator and modern Personal computer.

Any case, this is my personal opinion and personal experience of research.

Link to comment
Share on other sites

  • Premium Member

Crystal motors are digital, just as not elaborately so as stepper motor and microspeed controllers. There are some real difficult design criteria in motor control. It's not enough to just control the motor's speed, there are all sorts of issues to be addressed like ramp up to speed, braking, tendency of servo loops to "hunt" back and forth past the desired speed, etc.

 

I'm about to build a controller and power supply for an Arriflex 3-phase blimp motor. But I'm not going to start from scratch, I'm going to try to modify a commercial 3-phase variable speed controller to run at crystal speeds. That way I'll inherit all the work that supply's designers had to solve with respect to generating the 3-phase AC power and making that circuitry reliable.

Link to comment
Share on other sites

Hey everyone,

 

Hal, I've consider a lot of the things you've mentioned. Ramp speed, braking, etc. I've also read a lot of arguments over servos and steppers.

 

I guess I'm oversimplifying, but the controller itself would be responsible for timing, and that's it. The speeds, the ramping, braking, it's all timing. Sounds simple in concept, but I realize it's probably quite complicated in practice to build a controller capable of all that.

 

Another type of motor I had in mind would be responsible for the variable shutter. Something that would allow to program fades and dissolves. I'm assuming with something like that it would need to be synced somehow to the main motor. This getting into full blown motion control territory!

 

Well, is there anyone you guys would recommend that builds these types of things for a living?

 

 

Jay

Link to comment
Share on other sites

Hey guys,

 

Okay, trying to simplify things a bit. Rather then being able to punch in any speed to two or three decimals, I decided to have about ten preset speeds ranging from 6 to 120. I could use a rotary switch to select the speeds rather then some sort of numeric pad.

 

Would it make it easier if only 24 were crystal, or would it be no more difficult to make all speeds crystal?

 

As far as ramping is concerned, could there be a flat rate applied to all speeds? Each ascending speed would then take slightly longer to ramp up then the previous speed. Of course, how is this done?

 

I'm assuming you'd want a servo connected to the 8:1 gear shaft, which would require 900 rpm's for the top speed of 120fps, and each frame would require a turn of 45 degrees. I suppose you could have the motor connected to the 1:1 gear shaft, but then you'd need 7200 rpm's for the top speed. Wouldn't vibration be a problem at that speed?

 

So what I'm imagining, as far as the controls are concerned, would be three rotary switch's. One would be for the live action speeds, 6-120.

 

I suppose you'd have a switch that would change the control to the second rotary switch, which would be for single frame exposures. It would also have about ten settings from 1/8th of a second to 1 minute exposures.

 

Then there could be another switch/button that would activate the third rotary switch for the interval settings. I'm thinking about 12 selections from 2 seconds to 1 hour.

 

Then you'd need a power button, a switch for reverse running, and a start/stop button. Also, a power connection, and an outlet for possibly a remote.

 

Sound feasible?

 

I thought that you could have two separate motors inside the casing. A servo coupled to the 8:1 shaft, and a stepper attached to the 1:1 shaft. That way when you enabled the single frame control, it would change over to the stepper. Of course, I'm not sure if you could fit two side by side given the small amount of room. Also, I'm thinking it may not be possible anyways due to one of the motors being stationary when not being used. Would this cause some drag, forcing the working motor to compensate?

 

Alright, tear it apart if you want, or suggestions on how to make this a reality!

 

 

Jay

Link to comment
Share on other sites

  • Premium Member

Do you have access to a machine shop? From my perspective, 95% of any serious modifications to a camera or home-made accessories always ends up requiring a machinist who knows their way around precision machinery. I was connected by family for a few years to Cadillac Motor Division's retired chief tool and die expert. Unfortunately he's no longer with us - another of America's Finest Generation gone to their eternal reward.

Link to comment
Share on other sites

  • Premium Member

Hello Jay,

 

I am definitely in favor of you learning this stuff. At the same time, what is your goal? Do you want this gear because you want to make movies or do you like widgets and technology. James Beverly and I have chosen the path of self-equipping because we want to make movies. But, when you think of the time we've spent getting equipped, we could have actually made at least one movie if not more.

 

What do you really want to do?

Link to comment
Share on other sites

Hello,

 

Hal, I will need to find someone who has the equipment and expertise for any of my machining needs.

 

Paul, I've been interested in learning electronics/motion control/animatronics for quite a while, but like I said, very limited experience with any of it. I'm wanting to learn this in order to aid my film making endeavors.

 

After talking to a few people I'm starting to think you can't have a motor capable of high speed, and single frame/intervalometer functions, but certainly it's possible.

 

The Lynx C-50 can do this (with the optional quicksync controller), the Fries motor can do this (with the intervalometer remote), the Jackson Woodburn can do it with all functions completely built in!

 

The Fries is a dc servo, I'm assuming the other two are as well. Would they be coupled to the 1:1, or the 8:1 drive shaft?

 

First thing I'm trying to figure out is what kind of motor (servo or stepper), and which drive shaft to couple it with.

 

http://www.tobincinemasystems.com/page14.html

 

That article claims steppers can be hard on a cameras gears. Of course, everything else I've read suggests that creating circuits to control steppers is quite a bit easier. So?

 

I realize I'm at the bottom of a very steep hill.

 

 

Jay

Link to comment
Share on other sites

  • 2 weeks later...

Hello again,

 

Does anyone know of any forums that might discuss electric motors, and how to build them?

 

I'm wondering if there would be anything wrong with having a dc servo coupled to the 1:1 drive shaft if I intend on shooting at high speed. I figured that would be the only way to have a motor capable of high speed, and single frame/intervalometer functions. Although, if making two separate motors would be way easier, I have no objections.

 

Any resources on writing code for the controller?

 

Would it be best to use an eprom, or a microcontroller? Or something else?

 

So many questions. Having a hard time finding the answers!

 

 

Jay

Link to comment
Share on other sites

  • 2 weeks later...

Hey again,

 

Not much discussion in here! I'm not surprised. This motor business is super complicated.

 

I stumbled upon an old thread over at the stop-motion message board. Someone built a single-frame/intervalometer using a superior electric stepper motor, and a stepper controller from applied-motion. The controller comes with some software to help program everything, and the person that did this posted a video showing how the software works. Looked quite non-programmer friendly.

 

Anyways, applied-motion also carries servo controllers, which include the same software. However, the specifications for the software claim that only 8 inputs are programmable. I'm assuming, in the case of a live action motor, I'd only be able to setup 8 speeds. Is this correct? I had worked out ten speeds I'd like to use, and now it looks like I may have to take another couple off the list.

 

Out of curiosity, how are the old mitchell "wild" motors put together? Of course, those are AC, so that would probably be too dangerous to mess around with.

 

 

Jay

Link to comment
Share on other sites

  • Premium Member

I wish I had answers for your q's. I do know this, servo-motors are often used for continuous speeds even if those speeds can change incrementally via an encoder. Smoothness is the reason. Steppers are great and certainly perfect for stop motion. But they wobble. Some, a lot. They can shake enough on a lighter camera to effect the image quality. Bruce put a whopping big, rare earth, stepper on my scan rig. But, the camera is mounted to an even whoppinger big chunk of steel to eat the vibrations. It would be no good on a tripod. Unless something has changed over the last few years, servos are used for continuous speeds and steppers are used for intermittent speeds. It may be beyond motors to do both for you without some other compromise.

Link to comment
Share on other sites

Bruce put a whopping big, rare earth, stepper on my scan rig. But, the camera is mounted to an even whoppinger big chunk of steel to eat the vibrations.

Is your scan rig an Arri II or similar? And did he put that whopping big, rare earth, stepper on the 1:1 shaft?

If so, that was a big whopping mistake. When you use a stepper, the torque of the motor must be matched to the torque of the load. The 1:1 shaft of an Arri is a light load.

Replace the whopping big motor with a smaller motor and you'll be much better off. Maybe it won't pass the penny test, but it might pass the nickel test. :rolleyes:

 

The Nickle/Penny Test

 

2008-rev-test.gif

Edited by Dominic Alt
Link to comment
Share on other sites

Oh, I see, now. You didn't even ask what camera I was using before putting me down.

No offense intended, either to you or Bruce, but that would be my remote diagnosis.

In any case, Mitchell or Arri, basically the same thing. Not much torque is required.

 

If the vibration is a problem, try a smaller motor, or less current through the motor. If not, no need to change anything.

 

I've only seen the old motors in pictures. Very impressive. When running at normal speeds, it is nice to have a giant motor for fast acceleration. I'm not sure what size magazines were used with those cameras, but I think 1000' magazines were available. That puts a load on the camera.

 

I just bought a wild Mitchell MKII motor door. The attached motor is 1/6th horsepower! Not football-sized but it is bigger then a soda can.

I will be replacing it with a much smaller motor, but for single-frame only. No sync.

Link to comment
Share on other sites

  • 11 years later...
  • Premium Member

very old thread but I just wanted to add my 2 cents because this thread pops up in google searches very often when trying to find information about crystal sync motors. For "Crystal Sync Motor" I mean a brushed DC camera motor which RPM is phase locked to a stable crystal oscillator signal so that the motor speed follows the oscillator reference as closely as possible. The motor speed is adjusted with closed loop controller so that it stays as close the Crystal reference as possible no matter how much the motor load changes.

 

DIY crystal sync is fully possible. "Possible" does not mean easy but it can be done. I started a camera motor project this January which quickly turned to full crystal sync with multiple speeds and display and everything. At the moment the system works correctly and I am designing the circuit boards having the schematics ready and components chosen.

First I thought I wanted to just make a simple servo controller with simplicity in mind and no need for high accuracy speed stability. Just using an Arduino for simple speed measurement and just using the low speed arduino pwm for controlling the motor. No displays or anything. But eventually it turned out I really wanted to have a real crystal sync system which phase locks to a crystal oscillator reference signal to be truly accurate. And the display seemed to be really useful so I wanted that too. Multiple speeds were a must as well. 

By my opinion, you can start simple and try to make the simplest system you can. Then it may even work correctly. But the simplest shoestring system is never enough as a final solution and eventually you will want (or need) more features. Then you need to redesign the  whole system. After some time you need to redesign it again completely.    Either you add all the needed features from the start OR you will end up in a constant loop of updating and redesigning. It is a nightmare for people who just want to shoot with the cameras, NOT build them. And designing a really useful and possibly complicated system from ground up takes lots of time and testing and even money. You will lose your nerves countless times and without enough motivation you will abandon the project and move up to something more interesting. Not trying to be discouraging, it's just the way it is. 

 

I tested the 4046 chips too. They were working pretty OK but I had serious problems getting the locking range good enough so it was simpler to just write a microcontroller code which does similar style of phase locking task but has practically infinite locking range. It took lots of time of course but it is just a more advanced way to do these things. 

 

So the actual phase locking function seemed to be easiest to do by programming a microcontroller. Then you need of course the speed sensor which reads the motor speed. And a preamplifier which conditions the sensor signal so that it can be used for the phase lock. Sensors can be optical, inductive or Hall sensors. Or sensorless though it is tough to do. Inductive sensors ("AC generators") can be the original Pilot Tone generators some camera motors have built in. For optical one can use a slotted or reflective encoder disc with some kind of LED light source and some kind of photo sensor (light diode/transistor) or a "fork-style" assembly which has both the LED and the light sensor in the same package. For Hall sensor one needs magnets attached somewhere to the motor axle so that the changing magnetic field can be measured by the Hall sensor and used as a speed measurement signal after conditioning.  Sensorless speed measurement uses the current spikes which the carbon brushes make when they are switching from rotor winding to another. Sensorless measurement needs lots of additional components and signal conditioning so for most systems it is probably much easier to use the other sensor styles.

 

You need a crystal reference oscillator of some kind to generate a very accurate speed reference signal which the phase lock can compare to the motor's speed sensor signal. So some kind of Crystal oscillator section is needed. One can use traditional frequency dividers/counters like binary or bcd dividers to get the multi-megaHz crystal oscillator signal down to suitable frequency to be used as a reference. For example something like the CD4020 binary divider can be used for testing purposes and some of my first prototypes used partially that circuit. This task can be done by programmable dividers too. Or one can use microcontroller as a custom frequency divider ( ctc interrupts) . On my designs I am using all of these depending on which one is the most practical for that particular system. Depending on the system you may need to use multiple techniques so it is best to be familiar with them all.

 

At first I used low speed PWM directly from the Arduino programmed microcontrollers. After some time I learned how to do analog PWM controllers and updated to 38kHz PWM which one of the current designs uses. Digital high speed pwm is also possible but the analog system was pretty interesting to do so I wanted to try it and got it working pretty well. Some other systems will use digital depending on which style is the most practical in that application. You will NOT want to use low speed PWM signal in these systems. It works for control purposes but the noise is a nightmare and you will want to update it to high speed as quickly as you can. Use something which is higher than the human hearing range. For example the 38kHz my current prototype uses works just fine.

 

You will really want to use mosfets when driving the motor. Bipolar transistors are not very good for that application (heating, power loss, possibly a bit more complicated to control with low power circuits, typically physically larger sized for same wattage) and mosfet is even a bit simpler solution in the end. It needs a higher level control signal to saturate fully (it is switched fully on/off quickly by the PWM  signal and you will want to avoid the intermediate states so that the internal resistance of the mosfet stays as low as possible to avoid power loss and uncontrolled heating). So you may need to amplify the pwm signal before you can feed it to the mosfet's gate depending on your design and operating voltages. A bipolar transistor will do for this amplification stage or other means will work as well as long as the gate signal is high enough level compared to the motor's actual supply voltage. 

By my opinion, these Crystal Sync projects are really not suitable for persons who want to only shoot with movie cameras. If wanting to make these, one must be very interested in electronics designing too as well as building cameras from ground up and making mechanical parts and endless number of versions and prototypes which takes many months. Most people are not interested in what is inside the camera or a car engine as long as it works... but for this type of stuff you really really need to be interested in everything technical and mechanical and electronic aspects of the filmmaking products. 

Another thing is that it is definitely much faster and easier and much much cheaper to just order a crystal sync system from someone else than to build something by yourself from scratch including doing all the research and testing by yourself and learning to program this type of stuff and so on. 

 

All in all, I would say:  Fully possible to do DIY real Crystal Sync if you are really motivated and have from about 6 months to one year worth of spare time. You need to be really interested in programming and electronics and mechanical stuff. Just wanting to shoot with film cameras is not enough motivation and the project will definitely fail. Use high frequency PWM from the start, you will need it anyway and it may be difficult to update later without redesigning the whole system. The phase lock will probably be a bit easier to make by programming than with logic circuits. Mosfets are the way to go for motor control circuit. For speed reference you can use various techniques as long as they are all based on real Crystals (no external or internal resonators, no 555 style lower accuracy circuits). Check with your oscilloscope to verify the function and accuracy of your reference signal.

Oscilloscopes are a must when making these systems. You can't get you system working well without having a good oscilloscope (or two. or three. At least you need one which has two channels or more.) and know how to use it properly. Every signal stage needs to be checked and adjusted with the prototype using the oscilloscope measurements. Without the oscilloscope you can't even know if the system is working correctly or not.

If someone is interested in my DIY crystal sync systems, I am making videos of them every now and then. I am updating them here and you can find information in other forum threads as well.  https://www.youtube.com/playlist?list=PLXvIUtmF3OxvOYgAHWEapJQzL9e_DCWvX   

 

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...