Servo controlled large format 3d printed leaf shutter

[jimskelton]

I made a 3d printed leaf shutter that's servo actuated and controlled by Arduino. I posted the project at https://github.com/hackaninstant/Servo-Shutter

Fastest shutter speed is 1/8, so it's really only suitable for paper negatives, X-Ray film, or Lith film. The electronic parts cost around $10. The same servo control unit can be used on multiple shutters if necessary.

[reddesert]

Now that is pretty cool!

Do you have any videos showing the shutter in operation?

[John Layton]

Actually...a range from 1/8 sec to 10 minutes would be just about right for much (most) of my landscape/seascape photographs!

[Tin Can]

Glad to see progress!

10 years ago I tried Robot servos with Packards as my hands could not squezze the bulb

I got my hands working again!

https://www.jameco.com/?utm_source=bing&utm_medium=cpc&utm_campaign=bingbrand&msclkid=afaebaac0a591f052bc5dec13617ae40

[ic-racer]

Actually...a range from 1/8 sec to 10 minutes would be just about right for much (most) of my landscape/seascape photographs!

Same here. Even though I have beautifully functioning Copal 3 and Copal 2 shutters, I'm using "B" and counting most of the time.

[acrobatic_citron]

Nice, what is the limit on the shutter speed? Weight of the shutter blades? Could you go a bit faster say 1/60 with a light blades?

Edit: nevermind I read the Read Me!

[domaz]

Nice, what is the limit on the shutter speed? Weight of the shutter blades? Could you go a bit faster say 1/60 with a light blades?

Edit: nevermind I read the Read Me!

I'm wondering if you could achieve a higher speed by having a spring close the shutter blades, and have a servo inhibit the blades from closing for slower speeds. Sort of how a Compur Electronic works.

[Dan Fromm]

I'm wondering if you could achieve a higher speed by having a spring close the shutter blades, and have a servo inhibit the blades from closing for slower speeds. Sort of how a Compur Electronic works.

Good suggestion, and along the lines of a comment I wanted to post, except that I don't see a need for a servo. FWIW, I've been inside two different so-called electronic shutters. Ilex uses a solenoid (pull plunger when energized) to open the shutter rapidly, timing circuit to tell the solenoid when to let go, and a strong spring to snap the leaves shut. The AGI F135 has two over center springs and two solenoids (extend plunger when energized) and a timing circuit. One solenoid and spring to whack the blades open, another pair to whack them closed.

Also, those printed shutter blades seem heavier than necessary. The F135 shutter's blades seem to be metalized mylar. Thin, light, little inertia. I didn't ask the Ilex what its blades are made of, probably blackened brass shim stock. Again, thin and light.

[jimskelton]

I was thinking about what could be done to make this shutter design faster and here are a few ideas:

1. Make the servo gear bigger so the servo doesn't have to spin as long to open the blades (currently it has to travel 69 degrees)
2. Get a faster servo. There are servos that operate around twice as fast (but cost 20X as much).
3. Totally redesign the actuating mechanism, replacing the servo with a 2 spring system and solenoid for holding the shutter open.

If you did #1 and #2, you could possibly get speeds of up to 1/30, though the shutter body itself may need some tweaking to beef up the pivot points which can break off easily under too much stress.

Trying #3 means thinking about and designing the actuating mechanism differently. It also likely means you won't be using 3D printed parts since the stress of the materials by strong springs would probably break things. And the point of this was to quickly and easily make a shutter I can use on a process lens. After the parts were printed, the build took around 3 hours to put the shutter and circuit together, ready for use on the field. Making the shutter blades lighter is not an option if using 3D printed materials. Currently they are 3 .2mm layers of PLA and any less are not light tight, and the weight doesn't seem to affect the servo speed anyway.

[rayfor]

use a electrical magnet instead of serveo just like most eshutter. Modern eshutter(like the one in XCD lens) can work at 1/4000s.

[John Layton]

Naive question here about getting faster shutter speeds...and one which counter-intuitively assumes larger aperture values (physically smaller lens openings) and higher speeds. Would it be possible to design a shutter which would allow for wide-open aperture for focus, and when the lens is again closed prior to exposure, a physical stop would then limit the amount that the shutter opens (for exposure) to a diameter which would coincide with what would be the minimal physical size appropriate to the set aperture value - meaning less shutter travel and therefore higher possible speeds? Possibly graduate this maximum speed to coincide with the specific choice of lens opening? Yeah, I know, this runs counter to common practice - but within certain limits it might be applicable? Hmmm...could need lots of light - kinda silly idea, I guess!

Or...maybe a shutter with two sets of blades - one set opens and the other one then closes...with the assumption being that as each set has only one direction, they could be more tightly sprung?

[Peter De Smidt]

Terrific project!

[jimskelton]

Yes, this would be possible in the software. Since this shutter is controlled by a servo, you can limit how many degrees the servo rotates during a shutter cycle. With this shutter, the typical value is 68 degrees. The software would just need to know the aperture used and could limit how far the shutter opens. If the shutter only has to open 1/3 (around f/22 on a 300mm lens) or around 22 degrees of servo rotation, you could achieve a shutter speed of 1/30. Combine that with perhaps higher gearing and a faster servo, and you may be able to achieve something like 1/250.

A shutter with 2 sets of blades could work if the blades were operated by springs. Though, the second set may be unnecessary if the opening spring was stronger than the closing spring.

The point of this shutter was to use it for paper/lith negatives shooting around ISO .75 - 6. So far, I haven't used any shutter speeds over 1/4. I'm thinking about a high speed shutter design, but it wouldn't be a leaf shutter. It would likely be a curtain or guillotine type which actuates with springs and is held open by a solenoid.


Naive question here about getting faster shutter speeds...and one which counter-intuitively assumes larger aperture values (physically smaller lens openings) and higher speeds. Would it be possible to design a shutter which would allow for wide-open aperture for focus, and when the lens is again closed prior to exposure, a physical stop would then limit the amount that the shutter opens (for exposure) to a diameter which would coincide with what would be the minimal physical size appropriate to the set aperture value - meaning less shutter travel and therefore higher possible speeds? Possibly graduate this maximum speed to coincide with the specific choice of lens opening? Yeah, I know, this runs counter to common practice - but within certain limits it might be applicable? Hmmm...could need lots of light - kinda silly idea, I guess!

Or...maybe a shutter with two sets of blades - one set opens and the other one then closes...with the assumption being that as each set has only one direction, they could be more tightly sprung?

[jimskelton]

Posted some videos on the Github project page.


Now that is pretty cool!

Do you have any videos showing the shutter in operation?

[jimskelton]

Just finished designing/producing a circuit board for this project, which is posted on GitHub (link in first post). This makes it much easier to assemble the parts. I'll be designing a 3d printed box for it in the future. It's able to do shutter speeds up to 1/13, and can do speeds up to 1/20 if using a smaller aperture. Flash sync works with electronic flash only, though you may be able to drive a flashbulb unit with an additional transistor.

240351
240352

[Tin Can]

Interesting project

how small is everthing?

I may have already commented

I used RC Servos with Packards making similar speed




Just finished designing/producing a circuit board for this project, which is posted on GitHub (link in first post). This makes it much easier to assemble the parts. I'll be designing a 3d printed box for it in the future. It's able to do shutter speeds up to 1/13, and can do speeds up to 1/20 if using a smaller aperture. Flash sync works with electronic flash only, though you may be able to drive a flashbulb unit with an additional transistor.

240351
240352

[jimskelton]

Here's what I wrote up on my Instagram (@hackaninstant):

Need an inexpensive large format shutter for your lens? Here's one you can build yourself for around $10CDN. All of the resources to build it are free and available at my Github page (link in profile), including 3D printer files, C++ code, schematic diagrams, build instructions, and even gerber files for producing a PCB board!

It's intended for shooting low speed film/paper (ISO 12 and lower) and not high speed film. It was mainly conceived to allow accurate shutter speeds between 1/10 and 1 second.

You'll need:

- Access to a 3D printer
- Soldering skills
- Some electronic parts
- Around 2 hours of build time

It features a 50mm diameter servo operated leaf shutter with shutter speeds as fast a 1/13 fully open (1/20 around f/16), self timer, and configurable servo parameters, as well as X and M flash sync (though the opto isolator currently doesn't pass enough current to fire a flash cube).

This is an open source project which you are able to modify to your heart's content, so before you ask whether the shutter can go faster or whether it can be modified for a different diameter lens, take the files and see what you can do! Then share the results with the open source community.

In case you're wondering, here's the $10 breakdown:

$3.00 -- Arduino Nano
$1.70 -- 128x64 OLED display
$1.30 -- 9g servo
$1.10 -- rotary encoder
$0.30 -- TLP127 opto isolator
$0.16 -- button switches
$0.80 -- breadboard
$1.50 -- PLA filament
$9.86 -- total (not including shipping -- adds up to around $3-$8 per item depending on quantity)

240490
240488
240489

[Tin Can]

I will keep using Packards

I bet I have a bakers dozen