Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

help with film scanner circuitry

Mr.We

New Member
Hi Guys,

I'm working on building a film scanning machine, I have a pretty good idea of how the mechanics
should work but I don't have a whole lot of experience with circuits so it would be amazing if anyone here could offer some advice. I'm hoping in explaining my idea someone could help me figure out what specific components I should buy and the basics of how it would be wired.

The idea is based on this machine:
**broken link removed**
which is a movie film scanner for 3 different formats - super 8mm, 16mm and 35mm, i'm basically just using the same platform structure of the kinograph but using other software to control the camera settings. The camera is triggered to capture an image by my own design for the circuitry.

Essentially I want the scanner setup to be a 'projector' of sorts, except it can take 3 different film gauges, and it doesn't project an image, only advances one frame at a time, stops it for a moment - illuminated by a LED light - triggers the camera, then advances to the next, etc. The scanner should advance the film across the LED at 3-4 frames per second. Since all 3 film gauges are different sizes, i will need to be able to switch between 3 motor speeds. They don't have to be 3 frames per second exactly since the camera is triggered when the film is advanced sufficiently for each frame.

In my mind the circuitry would be something like as follows:

a) a laser shines through the sprocket holes onto a light sensor. It has adjustable height depending on the film gauge.

b) Depending on the film format, a 2 way switch is moved to one of two 'counters'. One counts to
1 (the amount of sprockets per frame for super 8 and 16mm film), the other counts to 4 (the amount
of sprockets per 35mm frame).

c) Then (somehow), a counter resets once it reaches its limit (1 or 4), and the act of reseting sends 2 signals:
1 to trigger a Machine Vision camera, the other signal instantly stops the take-up motor, maybe including some sort of breaking mechanism, since the timing is critical.

d) This pause in the take-up motor is perhaps a timer (preferably with easily changeable times, but
probably set around 1/50th of a second for the sake of argument), once the timer is up, the motor starts moving again and the whole cycle continues.

photo(4).JPG

My main concerns are:

1) Should I use stepper motors or continuous motors? The issue is around how to ensure once the desired sprocket is counted, that the motor (and hence the film) stops immediately. In terms of choice of motor my understanding is that a stepper motor might not stop in the same position for each frame. It seems most stepper motors are around 200 discrete steps per full revolution, which might not have the level of accuracy/control needed for all 3 film gauges, since the super 8 film will need smaller increments than 35mm film to advance.

On the other hand, my understanding is that whilst a continuous motor has infinite positions it can stop
in, that once it gains momentum, cutting the power won't necessarily stop it dead in it's tracks, that it will
slowly come to a halt. Maybe this isn't such an issue since the motor will be stopping and starting 3 or 4 times a second? Bare in mind the smallest in differences is enough to cause the resulting video file to shake all over the place and/or be out of frame.

2) Can I make a counting device reset at a preset number, and send a signal upon such a reset? Is this easy enough or perhaps there's an easier gadget to get the same effect? Perhaps for the 4 counts for 35mm for e.g., could I have 4 gates of some sort in series that close one by one, thus completing the circuit upon the 4th sprocket signal?

3) How to power and control the speed of the take-up motor, with 3 or so speeds? Also how to power and control the non-takeup motor.

4) BONUS if anyone here can help though it might be the wrong forum: Do you think I can get away with using a rolling shutter machine vision camera or do I really need a global shutter? Since it's only 3 or 4 fps, and i'm guessing the camera shutter can be set around 1/100th of a second shutter speed, and the film won't be moving
when the image is taken, I feel it should do the trick? Otherwise i'm looking at spending over double to get an equivalent camera with a global shutter.

Any thoughts would be great appreciated!

Cheers,
 
Tidbit: Dynamic braking can be done mechanically or electrically to a DC motor. You basically short the windings. Som motor control IC;s have fwd, rev, free wheel and brake modes.

I would guess that a proper servo controller could maintain any position as well as linear move.

Steppers have a torque limit, so reel size could be an issue. Do multiple film sizes definitely makes the design harder.
 
Welcome to ETO, Mr.We!

ALL timing/control issues can be handled by a uC (take your pick - I like the Arduino series [e.g., the Micro would more than adequate]).

Obvious, the uC would need programming. If that's an issue for you, there are some very sophisticated programmers on this forum that could provide assistance.
 
Thanks guys.

Great idea Clyde! That design would seem to solve many issues.

To trigger the camera, I could place a little paddle that gets pushed onto a microswitch right on the opposite
side of the maltese cross to ensure it triggers when the film is static.

And in terms of making 3 different sets for each gauge, it would just be
a matter of making the circumference of the cam and maltese cross larger or smaller, right?

Sounds like a DC motor is the way to go. In terms of torque, does it makes sense that I might get away
with just running the motor at one set speed (ie not having a mechanism to regulate the torque) as long
as I make sure the motor can handle the torque load of the largest reel of film i'd be scanning?

Assuming the above is correct - and imagining the intermittent mechanism is built to drive the film
take-up platter - would that mean the motor takes the exact amount of time to advance one frame
at the start of a roll of film (when the load is greatest) as it does to advance one frame towards the end
of the roll of film (when the load is the weakest)? If it will still gradually slow down as the load increases,
I would just need to test that the camera's shutter speed is fast enough to not capture motion blur when
the motor advances the film the quickest, right?

Othewise I could look at building some sort of torque controlling arm/lever to regulate the torque whilst
running the motor at a constant voltage? Or perhaps one of these torque motors that sound like they
manage torque within the motor?
**broken link removed**

Hopefully I can solve the motor issues mechanically, but failing that it's good to know there are
programming options!

Thanks again!
 
Personally I'd drive the film with a stepper motor based drive using the edge notches triggering a sensor as feedback.

Slowly ramp up speed to max speed at half frame, ramp speed down then microstep until frame is in exact position, trigger camera, repeat
 
Picbits, would you be computer programming the motor to do this? Since it'd need to be taking photos of 4 frames a second i'm not sure
it'd have time to ramp up and down in speed inbetween every frame, it would hopefully chug along at a contant speed and stop after
exactly the same amount of steps as the last frame once it senses the sprocket. I'm still at a loss as to how i'd wire up the whole circuit
if I were to go down this path.

I really like Clyde's idea, but now i'm not sure how easy it would be to make the mechanism advance the film 1 frame exactly.
 
I just read this on this page:
https://www.sparetimelabs.com/animato/animato/movements/movements.html

"The specifications of 16mm and 35mm film and perforations (by the SMPTE) state a tolerance of 0.01 (one hundredth) mm, or 0.0004 (less than half a thousandth) inch, and a film movement should have the same, or preferably even better, positioning accuracy. That's why a single Mitchell or Bell&Howell movement can cost US$ 10.000 ..."

Looks like it might be a long shot to think I could craft such a mechanism myself...
 
Stepper motors can be had for around £20 for a decent one (~$35 USD).
A controller capable of 1/128th microstep can be had for around £15-£20 from our Chinese friends and they seem to work reasonably well.

That would give you on a 200 step motor potential for 25600 steps per revolution. In reality, bargain on around 1200 accurate steps per revolution.

Gear the motor down - not too much as their maximum speed isn't huge anyway and you have mechanically increased the resolution.

You can't make a motor go from zero to max instantly, you need a ramp up and down time otherwise the motor would stall (stepper / servo / bldc etc) - you'd also find that accelerating your film too rapidly would possibly snap or stretch it. You could potentially run the film at a constant speed but have a motorised mount for the camera which moves the camera at exactly the same speed as the film, takes an image then returns back to it's original position and tracks the next frame - complicated but "doable".

Certainly slowing things down a bit would make life a lot simpler. I would suspect one to two frames per second would be fairly easily achievable with just a fixed HD webcam, a stepper controlled mechanism and optical sensors on the film sprockets to trigger the movement and stop positions.

Not a simple project though and fraught with complications but worth playing about with if you have the time and knowledge. The electronics side while simple for many would be a minefield for a beginner though.
 
If you can assemble the machanics for a maltese cross shutter arrangement then that will be the simplest electronics wise solution, allthough you would still need accurate speed control of the motor.
If I was faced with the task then a microcontroller would have to be used, maybe pull the film with a dc motor and then use a solenoid to stop the film at the correct point, current limiting the motor at the same time.
I have just been working on a project that involves accurate dc motor speed and position control, but its not up to starting and stopping at that rate.
 

Latest threads

Back
Top