A nice mechanical drive
  • If, like me, you have a liking for clever mechanical approaches to KAP you might like the Geneva drive, developed for early cinema projectors, allowing the film to be held briefly in the gate before the next frame is moved on. A splendid way to create a geared intermittent rotary drive from a slow rotating continuous source like a dumb servo.

    I can certainly see a KAP use for such an elegant old mechanical design. I know current autoKAPs do this with electronics but this is just so lovely.

    And thanks to Wikipedia for such a well prepared animation.

    Geneva Drive

    PS I'd love someone to fabricate a few of these for AutoKAP. (I'd pay for a couple myself.)
  • I'd be happy to do the fabrication.

    Sorry, I hadn't read your post thoroughly, and assumed you were planning on using some sort of spring drive. After a few minutes of head-banging trying to figure out how to pack enough winds on a spring to make it worthwhile, I finally re-read what you'd written and saw you mentioned a dumb servo. This is entirely doable, and should give hours of run-time. It would be neat to either use the geneva cam itself, or to put a second cam on the servo shaft to actuate the shutter so it's all driven from a single motor.

    An alternative to the dumb servo would be to use a low-speed gear motor like any of the Solarbotics GM series motors. The smaller ones have a shaft that's easy enough to adapt for this purpose, and they're pretty forgiving on voltage range.

    In any case I think this would be a fun project, and I'd be more than happy to do the design and fabrication for it.


  • The dumb servo (servo converted to simple low speed geared motor) would be an ideal drive for such a device, otherwise I may use my simple AutoKAP drive with elastic band and silly putty friction bearing set for a reasonable revolution rate to match the drive.
  • The lost art of pure mechanics. Killed by the microprocessor. I mourn it's passing. I found that one in a 1904 book titled "Mechanical Movements and Mechanical Appliances" that my grandfather left me. Amazing what you can do without silicon. :-)
  • Very inspiring.
    Here's a link to a PDF with a drawing on the second page that will help to construct this.
  • Aaaaah, not entirely lost. For what it's worth, the first time I ever worked on a geneva mechanism was on an infrared camera mounted on the back-end of an adaptive optics system. The thing never really did work out, unfortunately, because it's actually quite hard to get a zero-backlash lock on a geneva, and that's what was called for. For the replacement I voted for a solenoid driving a ratchet wheel with a sprung ball engaging a V-shaped detent to lock it in each position. The arrangement kinda looked like a steam engine. Unfortunately interest in the instrument waned before I got to build the steam engine wheel design.

    For KAP, I think the geneva would be a lot better suited than a racheting solenoid because of the ramp-up and ramp-down on the rotation rate. It should make for very little residual rotation in the rig at the end of a move, and might not even need additional damping.

    Seriously, I'm in. I've got some 25mm diameter Delrin rounds lying around looking for something to do. If I can make everything work out with that, it'd be easy enough to make up several sets with the material I've got on-hand.

    I'll design around a Futaba spline, but I may also try an under-sized 3mm hole to use on a gear motor.


    P.S. If anyone's interested in seeing that pure mechanics are still alive and well, take a look at "Home Shop Machinist" and "Machinist's Workshop", two publications from Village Press. A third publication of theirs focuses on CNC machining, so it's not PURE pure mechanics, but a lot of the stuff that gets produced doesn't have a smidgen of silicon in it. Neat stuff. But I'm getting way off-topic.
  • Count me in for a small "6 station" version (ie as per the wikipedia animation and illustration below), rather than the 4 station type if you can do it. Happy to have in Futaba splined and another shaft adapted for 3mm. I love the creative enthusiasm and depth of knowledge and skills available around here.

    6 station Geneva mechanism
  • This could be great for a motor free concepts.

    I'm not sure about the positives and negatives of the different designs. The dwell time, angular velocity of the driven wheel, and contact surface during dwell will be different depending on the size and number of stations. The driven wheel for the four station model looks very fast to machine. You could gear the output however you wanted. I think you would have to make a simple jig to complete the slots in a 6 or 8 station model. Not difficult. Especially if you are making a few.
  • To some extent the dwell time and angular velocity can be adjusted by changing the speed of the drive motor. But yeah, it'll depend on the geometry of the geneva.

    I have to admit, I was going to cheat somewhat on making the thing. Rather than using a jig or rotary table, I was going to CNC the geneva wheel and the drive cam. Making duplicate sets shouldn't be a problem.

    Hey, quick question: What should the exit shaft look like? Or should that be left as an open-ended question for now? I'm tempted to make it something that'll play nice with Brooks's picavet cross since that's what I'm using on my rig. But I can leave it pretty generic.

  • For me, the exit shaft to picavet or other suspension would be fine as a 3mm hole which if necessary could be drilled out to a little larger then I'd probably go for threaded shaft and locknuts . So Generic rather than specific for me....
  • Tom, please put me on the list for one too. Thanks.
  • For the moment, I'm figuring out a system for taking motorised ground shot panorama's. This system could be very interesting, because it has 6 positions, for which I need to take 6 separate shots.
    If you have more info on your project, Tom, please let me know.

  • 3mm through hole it is, and I'll stick it in a 5mm or wider shaft so it can be re-drilled or tapped as necessary.

    Yvan, I'm guessing for ground shots you'd need something beefier than I'm designing, but I'll happily post all the drawings and details once the project is done. Glad you mentioned this, because I hadn't even thought about extending this to a ground-based panorama setup! It'd be perfect, especially if the rotating mount is tripping the shutter.

    Peter, no problem. If anyone else wants one, let me know.

    I haven't had a chance to start work on the design yet, but I'll post some prototype drawings as soon as I've got them done.

  • Computer animations, CAD drawings, CNC machining? Seems an odd way to eliminate a microprocessor on a KAP rig. I'm of a mind to design an electronic controller with a small LCD that can display that wikipedia animation while it's controlling everything electronically. ;-)
  • Hahahahaha! Touche, David. And when I've finally built that Holtzappfel Rose Engine, I'll be in the market for an LCD animation of that, too.


    P.S. No, I'm not planning to build a Rose Engine. But cripes what a gorgeous machine.
  • Electronics and mechanics; two completely different languages. Combining them is one aspect that makes KAP so interesting.
  • Alrighty, here's the first pass:


    5mm diameter output shaft with a 3mm hole, and I did my best to put the output shaft at the symmetric point on the servo so you can just rotate the servo 180 degrees, bolt the Geneva on top, and still be in balance.

    The top and bottom plate will be 1/16" 6061 aluminum, and the Geneva wheel and cam will be Delrin. The pin will be a 1/16" dowel pin. I've still got some work to do, like beefing up the lower plate a little around the cam wheel, and making the broach to cut the Futaba splines in the cam wheel. But otherwise it should work as-drawn. Also, I doubt it'll need all four bolts through the plate, so it should be possible to just use two diagonally opposing bolts and not interfere with the bolts and standoffs that hold the pan servo onto a BBKK rig.

    One bit that's not drawn in is any way for the cam pin to hit a switch to trigger the shutter. A lever switch would probably be the best bet, set on the opposite side of the cam away from the Geneva wheel. But I'll try to leave enough meat on the plate for that to be added later.

  • benedict:

    If you're taking orders, now or in the future, I'd love to get on the list.

  • David just to answer your point, "Seems an odd way to eliminate a microprocessor on a KAP rig."

    For me KAP is an adventure in ideas, it is a ever-fascinating field of play in which to try out ideas, creativity, expansion into other fields. it is an area in which thankfully, there are few rules and the people who practice it are utterly commited to sharing their concepts and ideas, building on them, and congratulating each other on the results of their efforts.

    A fact that I endlessly delight in is that there are no right and wrong ways to do KAP, and whilst microprocessors and complex electronics will do it and may be the ideal approach to consistency and accuracy, other approaches using wooden cameras, or mobile phones from stunt kites, or elastic bands and silly putty all have validity. Strangely enough too, each slightly different approach ends up being a little niche within the whole, and that collection of niches and approaches creates that fabulous diversity of KAP we see today.

    I realise your comment was not really an attack on these fundamentals of KAP, more a sideways look at the computer machining technology and bringing it to bear on such mechanics.

    I was thinking anyway to write something like this from for some time to remind all KAPpers that there are many ways to KAP heaven, and each time the latest KAP electronic invention comes out, it also has the potential to stifle the inherent creativity in KAPpers. If all the thinking is done for you by an electronic box, it is possible to miss the chance to do it a simple way to similar end results. And likewise, whilst I delight in the hundreds of KAPpers brought to KAP by Brooks' kits, I also sometimes feel a little pang that some of those folks have missed one of the greatest pleasures of KAP (for me) - learning workshop tools and techniques and building rigs from scratch.

    I'm delighted we have someone here that has the potential to machine such parts, because although I could surely do it, most of my fabrication on a geneva mechanism would be labour intensive and take me some hours to do.

    When I see something like the geneva mechanism, or a solar powered houseplant turner in a garden shop, my whole mind leaps ten feet to the left and screams, "you could use that for KAP !"

    I have KAP rigs "enough to cobble dogs with " (wonderful old saying from my mother) and in truth need no more, but as long as I have hand and mind, I'll go on thinking up new ways and approaches. There is little need - just the fun of doing it.

    I know from your sensational ring rigs that you have a similar inventive approach to KAP problems Dave, so I know your comment was not really pooh-pooing simple mechanics.

    I think the geneva mechanism is going to end up on two rigs one a ultra light elastic band powered, and the other with a dumb servo and some other ideas incorporated.

    Love the 3D Tom. Looking great.
  • Wow. I'm as blown away by the model as by the upcoming device.
    and of course, add me to the list too, esp after the shut switch jumps aboard.

    As to Simon's chagrin that lots of newbies are missing the creative part, I agree -- but selfishly, that gives me a lot of exploring and learning and tweaking to enjoy myself. If it weren't for that part of KAP, I wouldn't be here several times a day.
  • Simon, I think you recognized my comments as the good-natured jab they were intended to be. And lest there be any doubts, I do admire an elegant mechanism as this certainly is. Why else would I carry around your autokap thingy that I bought at auction at KAPiCA 2006?

    Tom, I'm impressed by how quickly you volunteered to do the fab and put together a cad drawing. Methinks I've made a serious error not posting a sketch of a next-generation tilt ring mechanism to this forum. Perhaps someone here would have built it and shipped it to me by now.
  • Tom,

    Could you post a 2D drawing of the wheels with measurements? I'd be interested in trying out a few different variations for a wind-powered rig I'm building around the gear box of a disco ball. I love the creativity here.
  • Tom,
    Please put me down for one. The compactness of your drawings of the device are great.
  • The Geneva design is brilliant, but how do you plan to synchronise firing the trigger with the still moment of the rotation?
    Tom, put me on the list as well as I really would like to test such a device. I have been using a small electric motor ( 1RPM) for a few month combined with a 5 seconds intervalometer on a RicoH GX 100 but I have to go through hundreds of shots for a decent panorama. This design may be the answer.
  • Great piece of work Tom. Put me down for one too!

    One thing slightly worries me though. The drive is going to have to turn the Geneva mechanism very slowly for this to be comparable to a TUCIT or Gentled360 (i.e. one step every 5-10 seconds say). Will this be possible without a lot of extra gears?
  • The dumb servo I have here does some 15 rpm on a 1,5 V battery. Or it takes some 4 seconds for 1 rotation. With a "6 station" Geneva that would end up to some 24 seconds for 360 degrees for the camera.
    At 3 V it does 32 rpm.
  • Just some random thoughts...

    Do we have any idea how long a 1.5 V battery would last running continuously?

    I wonder how the vibration and friction would be between the drive wheel, the driven wheel, and the servo.
  • I've run a small dumb servo on an AAA for over an hour. Certainly with rechargeables and a couple of spares in pocket, thats enough for me.

    I read somewhwere that high speed Geneva drives need good lubrication and I can imagine foreign bodies could interfere with them. I expect for the proposed usage this shouldn't be much of an issue. There is potential for a fair bit of friction between the flat surfaces.
  • Tom, please add me to your list; though I'm still figuring an application, I can not pass up such a thing of beauty.
    Thanks for your efforts.
  • One thing slightly worries me though. The drive is going to have to turn the Geneva mechanism very slowly for this to be comparable to a TUCIT or Gentled360 (i.e. one step every 5-10 seconds say). Will this be possible without a lot of extra gears?

    If the pin on the driving wheel (or another bit rotating in sinc) simply triggers an "on button" for the Gent360-LED at or just before the beginning of the dwell period it should work fine. A simple gentLED shutter or equivalent would be all you really need though. With the right camera and set up you could take exposure bracketed bursts during each dwell period.
  • I think one of the nice properties of Delrin is that it's self-lubricating. So hopefully friction won't be much of a problem in Tom's design.
  • Wow! This is exciting! I would also like to be on the list to get one.
  • I think one of the nice properties of Delrin is that it's self-lubricating. So hopefully friction won't be much of a problem in Tom's design.

    Ah, thanks, Dave. I had no idea what Delrin referred to. (I should have looked it up.)

    I've run a small dumb servo on an AAA for over an hour.

    Thanks, Simon.

    Solar power anyone?
  • tgranulosa, I can certainly post the 2D drawings. The dimensions are all a little oddball because I'm matching metric and imperial parts, but more because the dependencies are all at funky angles rather than being nicely rectilinear. But I'll try to get the 2D drawings up today.

    Pierre, tgranulosa pegged it: When the rotation is locked, the drive pin is 180 degrees away from the Geneva wheel. So if you put a switch at that end, it'll trigger whenever the Geneva is locked. Turning that into something that'll trip a shutter is another story, though. (Is this something the GentLED already does? Since my camera doesn't do IR, I'm really not all that familiar with it.)

    Yeah, rotation speed may be a bit of a bugaboo. I like the idea of running off a lower voltage source to drop the rotation speed. Another possibility (though a lot less fun on the construction end) is to make a Geneva with twice or three times as many stations. Set up two switches, one to index off the cam's drive pin, and the other to index off a set of bumps or divots in the Geneva wheel that indicate the positions where you really want the shutter to fire. It's oogly, but because of space constraints that's how that Geneva mechanism inside that IR camera was set up (you can have a much smaller cam this way). Given the choice, I'd drop the speed on the motor.

    Far as friction goes, it's Delrin on aluminum. Not ideal, but David's right: Delrin self-lubricates. I stuck a 2mm wide boss at the top and bottom of the Geneva wheel, so it should have a wide enough contact area to wear well, but not so wide as to be draggy. Time will tell.

    Man, the list is getting long, and I haven't even cut any parts yet! ;) Just to let folks know, I've got the Futaba spline worked out and can make a broach to do the job. (But egads, NO ONE has specs for that spline! If anyone can find a drawing of the Futaba spline, let me know. I've got 6mm OD, 25 splines, 65 degree angle, and two different radii, one for the spline tip and another for the spline root. It's a weird shape. I'd love to see a real drawing since mine's all based on guesswork.) The only remaining design issue is how to make the thing brain-dead easy to fabricate. Given the size of the parts, material costs aren't that high. And since it's Delrin and aluminum, tooling costs are minimal as well. The biggest bear will be the time required to make everything. So despite the whole, "I won't use a jig" thing I said earlier, I'm trying to design some jigs so I can make about five sets at one go and try to get the total time per set down under an hour. It's looking like it'll be a mix of CNC and manual machining, so I'll be busy the whole time.

    Simon, thanks for finding this in the first place! Even though I've worked on one of these before, it never occurred to me that KAP was a good application for a Geneva. (DOH!)

  • This is a verrrrrry interesting topic. Got both eye-balls on it...
  • I also sometimes feel a little pang that some of those folks have missed one of the greatest pleasures of KAP (for me) - learning workshop tools and techniques and building rigs from scratch

    Don't worry. I'm a newbie and used the Brooxes kit and found myself purchasing a dremel, hack saw, new drill bits, glue, etc. to put it together. I felt quite crafty adapting the shutter servo parts to my camera. Later when I got the Kap Feather I used shape lock to create my own lock nut. I purchased a kit for the dremel that included a bunch of bits for working with metal. One of the bits is great at smoothing rough edges on metal and is small enough to smooth the points on the wire snap used to connect the line to the kite. Not bad for a city slicker with no back yard, garage, or basement. :-)
  • I have been playing with 2 plastic disc for a while now, trying to work out how best to use them. When placed on top of each other, they will slip one way, but luck the other way. I hope to find an easy way to set it up, so I can use 1 unmodified servo to do both 360 deg one way, and trick the shutter. Each click is about 12 deg.

    Has anyone played with this kind of setup
  • Simon, I'm sure you popped an idea into our friend in Brasil mind. With his cnc, there is nothing he can't do.
  • tgranulosa, I can certainly post the 2D drawings. The dimensions are all a little oddball because I'm matching metric and imperial parts, but more because the dependencies are all at funky angles rather than being nicely rectilinear. But I'll try to get the 2D drawings up today.

    Tom, don't worry. After a little geometry refresher I think I found general solutions:

    Radius of driven wheel (distance from center of driven wheel to pin when it engages): A
    Radius to pin on driver: B
    Distance between axes: C

    Making the assumption that the motion of the pin is aligned with the groove when it engages, ABC is a right triangle and A^2+B^2=C^2

    Solutions (A:B:C):

    4 stations (1: 1: 1.414)
    5 stations (1: 0.727: 1.236)
    6 stations (1: 0.577: 1.155)
    7 stations (1: 0.482: 1.110)
    8 stations (1: 0.414: 1.082)

    The radius of the journal surfaces is flexible (smaller than B) but defines the thickness of material around the groove and therefore the strength of the driven wheel. I think if you follow this solution the cutout on the driver should be 90 degrees for 4 stations, 108 degrees for 5 stations, 120 degrees for 6 stations, etc.
  • Whilst I admire the mechanical elegance of the Maltese Cross, I can't help wondering if this is really the best use for it?
    As I understand it, it was designed for applications where the driving wheel is rotating continually - as in a cinema projector. To have a continually rotating wheel in a KAP rig implies heavy battery use. To use a servo as the motive force is surely overkill as the servo itself can already do the job so perfectly - with the aid of a simple timing device. It's not that I don't like mechanical solutions per se, but I just wonder if this a good one?
    Incidentally there are other mechanical delights to be found at: http://www.dalefield.com/mm/Manual_1934_Std_Mechanisms_UK/index.htm
  • Nigel, I think it's most interesting for something like wind power or spring power that's mechanically difficult to turn on and off.

    I suppose there might be something else you can do with the power during the dwell phase. Operate a mechanical shutter switch via another driven wheel, signal the operator, make tea?
  • re - wind-powered rigs - exactly! Timo Noko had some photos of his propellor-driven rig on Flickr (no longer there) which would be just right for using such a mechanism and Marooka-san's wind-driven rig in Aerial Eye 2.3 (page 7) uses the same technique.
  • As Nigel, I am wondering wether putting this beautiful Maltese Cross on the pan servo is the most useful application. Especially when one considers that you can easily shoot sharp pictures from a continuously rotating rig.
    (e.g. http://ostro.ced.berkeley.edu/~crisr/discuss/comments.php?DiscussionID=1061 )

    However this geneva mechanism is so clever that there must be an application : I was thinking about coupling tilt to pan through a Geneva drive : pan 360
  • Good point, Nigel. That's one reason I was also kinda leaning toward one based on a Solarbotics gear motor. But I love the idea of a wind-powered one as well. Hey, that's how the kite's powered, might as well power the rig the same way!

    Oddly enough, the Geneva was originally designed for extremely intermittent motion of the cam wheel. If you leave out one of the slots, it basically makes a multi-turn stop on a shaft. Like... winding the spring on a watch. Five turns max? Make a six station Geneva and leave out the last slot. You only get five turns before it stops. (Which is one reason why these sometimes get referred to as Geneva stops.) Adapting it to the motion picture projector was a whole 'nuther story that came later.

    tgranulosa raised another point regarding the mechanical strength of the Geneva. Yesterday while I was poking around I ran across an IBM paper from the mid 60's that went through the mechanical strength aspect of Geneva design. Apparently they've been used in everything from slow intermittent drive motion all the way up to extremely high speed continuous drive motion. I don't think there are a lot of rules on how it can and can't be used. Looks like just about everything's been played with. (But I'm going to go back through the paper to make sure the design I've got won't break if someone slings an SLR under it!)

    I've got some testing to do on my mill before making parts (like how to clamp the little bits down to machine the cam and Geneva profiles) and I do want to go back and look at the mechanical strength of the thing. While I'm doing that, might as well ask the question that's being raised already: Is a servo the right thing to stack this on top of? Or would it have better application if it was designed to use some other driving mechanism like wind or a small low-power gear motor that wouldn't drain batteries too fast?


    Here are some 2D drawings:

    Geneva Drawing

    Kinda weird... It's easy to get JPEGs of 3D models out of the CAD software, but I'd never tried to get a JPEG of a 2D drawing out of it. Two ways: One is to print to a PDF and change that into a JPEG, but it wound up looking like a really bad photocopy of a very old blueprint. The other was to save the current viewport as a JPEG, which is how I made this. But it wound up being pretty big, so I re-sized it to 800x500. Smaller than that and it was illegible.

    I'd describe the geometry, but tgranulosa pegged it in an earlier post. I left in the layout lines, which basically followed that math.
  • As Nigel, I am wondering wether putting this beautiful Maltese Cross on the pan servo is the most useful application. Especially when one considers that you can easily shoot sharp pictures from a continuously rotating rig.

    I thought we were winding film.
  • It's a fairly ugly but you could make the geneva mechanism in three parts. An aluminum cam, the journal surface of the driver (with cutout), and the driven wheel.

    The latter two could simply be discs. This way you can create the 2D pattern with a milling machine in the end of a 1 inch or 1.5 inch delrin rod, transfer it to a lathe, adjust the exterior diameter, drill the center, and then part off several discs. At ~1/10th of an inch you should be able to make ~3 driven wheels or ~6 drivers at a time.

    I found a geometry that works well for a four station model:

    The driven wheel is made from a 1.5 inch rod with 1/4" slots extending from the exterior to 0.3" from the center along along the x, -x, y, -x axes and four cutouts made with a 1" mill at 0.75,0.75; -0.75,0.75; -0.75,-0.75; and 0.75,-0.75''. The journal surface of the driver is made from a 1 inch rod with a cut out made by a 1.5" mill centered 1.061" from it's center. No adjustments to the exterior diameter of either piece, no small parts in the vise, no resetting of the vise during machining.
  • @Tom:

    A third way to set a JPEG from a 2D drawing is to do a screen capture and paste it in to a graphics editor (such as PhotoShop PaintShop, and so on).

    If you go through the PDF route, you might lose a lot of detail unless you have your PDF distiller settings set just so.
  • Quick update: I finalized the gear geometry and placement, so I'm going to cut a set of gears and make a pair of plates. This gives me a prototype to test. But I could use help on the switch:

    Right now my one and only KAP rig is a BBKK with an AM receiver on it that can be swapped out for an AuRiCo. My camera won't do IR, and isn't on the list of USB-operable cameras. So none of James Gentles's hardware will work for me. (RATS!) Soooo... What does the shutter synchronization switch on the Geneva need to look like so it can be useful? A simple switch? A TTL 0-5V output? A servo PPM signal that swings from 1ms to 2ms to move a shutter servo?

    By far the easiest would be to sink a small magnet into the cam wheel and use it to trigger a hall effect switch or a magnetic reed switch. A slightly more complicated one would be to have the cam's drive pin hit the lever on a micro switch, still providing a simple switched output. To have it drive a servo directly would take more hardware, but would still be entirely doable. What's the best bet?

  • benedict,

    What camera are you using?

  • I currently use a Nikon Coolpix 5600. But I was asking in general about the switch since other folks have expressed an interest in one of these beasties.

  • Hey Tom,

    You have talked about running this with a power supply and a servo. So why not a power supply and 2 servos. 1st servo runs the geneva mechanism, the 2nd servo trips the shutter, have a micro switch tripped by the pin on the geneva mechanism trip the shutter servo?


Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion