I put together a 3D model of this device to show a family member (who's a machinist) so that he could get a good understanding of how the different parts of the device fit together. He is going to get it built for free / cheap. I just pay for materials.

http://www.frankladner.com/images/vibro_adapter_3D.gif

However, I am now thinking more towards the 4-pole version, since it would allow for a smaller unit.

Anyhow, I just thought I'd post it and maybe it would help some of the guys that are new to this project understand it better.

...now that I look back over the image, it may not be the best presentation...maybe a little confusing...but hopefully after comparisons are made from the various images/angles, it will fit together and make sense.

Thanks!

that is a lot better.
But i still bo not understand why you add this plate in the middle with the ball bearings.
in theory just two plates, one static , one moving should be ok ?

Excellent images! I dont want to speak for Franks design but your definately right in thinking that the concept will work with only two plates.

BTW Im also going with a 4 shaft system. Its not only smaller but its easier to build with less mass on the plate which will hold the GG.

The idea of having 3 or 4 motors instead of a pulley system introduces some possible problems you might want to consider. Each of thoughs motors will have to always be running at the EXACT speed for smooth/quiet motion and you just reduced your battery run time by 1/3 or 1/4th. Not to mention its also likely to be 3 to 4 times noiser. For these reasons Im making mine with a pulley drive.

Very nice work everyone. Were getting there.

P.S. Rai. I dont want you to give away your design if its going to turn into a money maker for you but could you at least give us a hint at the basic concept for the curious like myself? Being that I've been working on this project for over a year I've gone thru at least a dozen ways to make a mini35 like system work. My bet is on something similar to one of my design I posted awhile back. Its a adapter without any moving parts that is grainless (because there isnt any gg), flips the image right side up and works on the concept of front projection instead of rear projection. I decided against the design only because of light loss problems.

>The idea of having 3 or 4 motors instead of a pulley system introduces some possible problems you might want to consider. Each of thoughs motors will have to always be running at the EXACT speed for smooth/quiet motion and you just reduced your battery run time by 1/3 or 1/4th. Not to mention its also likely to be 3 to 4 times noiser.

No your wrong. We had build both, 3 axis with external pulley drive motor and 3 CD-ROM like motors instead the axis. and the last way was noiser and works with less power, because it work without a pulley. A CD-ROM motor is a 3-phase motor, it work with AC and you can connect every phase of the 3 (or more ) motors and with every puls all 3 motors move the exact same angle. But it will works also with 3 DC motors. The exact speed is given, because there is a little mecanical transmission over the moved platform, like the wheels on a old steam train are moved. Think on the pulley, it work also not exacty, but it work because there is also the mecanical transmission.

Brett, i know you design, but the light loss is the problem.

I know also a different design, maybe the best off all, but we can not >The idea of having 3 or 4 motors instead of a pulley system introduces some possible problems you might want to consider. Each of thoughs motors will have to always be running at the EXACT speed for smooth/quiet motion and you just reduced your battery run time by 1/3 or 1/4th. Not to mention its also likely to be 3 to 4 times noiser.

No your wrong. We had build both, 3 axis with external pulley drive motor and 3 CD-ROM like motors instead the axis. and the last way was noiser and works with less power, because it work without a pulley. A CD-ROM motor is a 3-phase motor, it work with AC and you can connect every phase of the 3 (or more ) motors and with every puls all 3 motors move the exact same angle. But it will works also with 3 DC motors. The exact speed is given, because there is a little mecanical transmission over the moved platform, like the wheels on a old steam train are moved. Think on the pulley, it work also not exacty, but it work because there is also the mecanical transmission.

Brett, i know you design, but the light loss is the problem.
I know also a different design, maybe the best of all, but we can not produce the necessary part. It is like a light-wave cable with 2 - 4 millions of threads, bundled on one side to the ccd format and at the other side to the 35mm format. It can mounted on the ccd and so it need no lenses.

We work since many years on different 35mm DOF solutions and other parts. (All for a movie production and it was before P+S make the adapter). I like to tell our last "secret", but this will cost my head. But i can tell some good and some bad ways. Maybe i can post some pictures of our older (but good working) parts. Also we can sell parts, that will be okay.

Giroud: I thought at first about just removing the bottom plate. But then I thought about the pulleys where the belt is attached underneath. I figured that there would be tension from the belt squeezing the poles together and that they would need support. I guess this depends on how much slack there would be in the bearings. Likely none but would it wear things down faster not having a bottom support plate? (But then again, with the top plate for support, they couldn't be 'squeezed' in too much, eh?)

However, removing the MIDDLE plate, as you suggest, might work. I guess the pulleys would then be placed in the middle. That would simplify it and make it cheaper. Hrrrmm..... This sounds like something to try.


Thanks for the feedback, guys!

Frank, try it, but think about fine mechanic like a clock, and not like a toy. If you remove one plate the axis "hang" only in the bearings. The slack in the bearings is not a defined parameter and the axis will swing and the bearing will not work long times. The only possible way is you use a thicker plate and mount two bearings in place of one (one directly over the other)

Rai: I completely understand your thinking. But would the top plate act as a support also, or would it not due to the fact that it is offset (with the motion - although small - and weight moving the poles)?

When I actually get my hands on some bearings, I'll have an idea of their characteristics and will better understand these principles.

For the optical fiber idea, it already exists in night vision device (generation 2 and up).
They put a lot of optical fibers together, then make a 180 degree turn and glue all. this gives a really lightweight system.
http://www.us.schott.com/fiberoptics/english/products/nightvision.html
unfortunately it cost probably the hell, but it would be a good solution because you got two problems solved in one shot.
vibrating the picture and putting it upside down.
if somebody is in touch with some military guy who could provide a dead model, we could check if resolution is usable (i doubt).

for the 2 or 3 plate question:
If you put a ball bearing at each side of the plate and this one being thick enough (approx. 3 or 4 mm), i thing there will be no more problem while keeping only two plates.

for the pulley and belt, i do not undersand why you need it.
if you drive one hole of the top plate , the other ones should follow as their are physically drilled in the same plate.
it put a little bit of stress on one of the axis.
It should be even better as the motion created is not more exactly an ellipse but can slightly vary.

Sorry to challenge the design , but i am not a mechanic guy and i try to think to all these guy who are building the stuff on the kitchen corner and hope for a really simple design.
On the other hand , if you sell it finished for a good price , i do not really care how it works.

I just came across this page and wanted to share it for the guys like me that are new to bearing nomenclature.

http://www.dynaroll.com/catalog/pag005.htm

Pretty helpful information, I think.

Frank, every transmission , every wheels and every axis need two bearings. It´s basic mechanics. If you try those with only one, it will work inaccurately and only short time.
The first two plates in my disign stand still. The only think they do is hold the 6 bearings (and housing and motor).

Giroud,
1.) optical fiber idea: Cool. I newer see this before...

2.) 2 plate, your idea is good. You have to bearings on a axis, but only one plate, okay.
3.) pulley and beld. You need it, because the moving only is to small.
4.) really simple design. Oh, yes, there is a realy simple design, but ... ...mmmmh

I am about to order the following ball bearings, and wanted to share the source with you guys.

MR105-ZZ (5x10x4mm) $2.00 each or $6.95 for 4 Pack (BWK-06)
(10mm diameter, 5mm bore, 4mm thick)
MR63-ZZ (3x6x2.5mm)$3.50 each or $11.95 for 4 Pack (BWK-40)
(6mm diameter, 3mm bore, 2.5mm thick)

http://www.bocabearings.com/cart/cart1.cfm

I am getting 2 packs of the 10mm (8 total) and 1 pack of the 6mm (4 total) the final price comes to $25.85. (Free shipping on orders > $10.00.)

The fiber optics idea was explored a while back but the research on the subject revealed that the resolution of current systems on the market was not up to par with even standard def. But perhaps they have inproved on the technology since then. Its also used by some plummers to check a pipe for a clog with a video probe. Same low res problem though. It would seem that you would have to either have every fiber optic strand perfectly lined up with each pixel on the CCD. If not your would need to have a much finer array of fiber optic strands to make up for any misalignment. Either way it sounds like is not a option for a DIY project. Pretty interesting stuff though. On a interesting side note they are thinking about using fiber optic cable instead of flourescents in office buildings during the day. The sunlight would be captured by a parabolic mirror on the roof and sent down a array of fiber optics cables to each room. No power needed. Cool stuff.

-Brett Erskine

Brett, that the point. The fiber optics idea work only if perfectly lined up with each pixel on the CCD (like the RGB filter on one chip sensors). My company make opto mechanics and also a very fine 35mm GG solution (last week we tested it on a optical bank with a 25000x1500pixel sensor and it show a high sharp test picture with no grain) but the fiber optic idea, we can not realizable.

Rai,

Is this GG that you make for sale? How much?

Did I also hear you saying a while back that you make follow-focus devices for still lenses? I would be interested in knowing the price on that too.

Thanks

Regarding fibre optics. They are called coherent fibre bundles when their arragement at one end is the same as the other so an image at one end is reproduced at the other.

Following is information kindly provided to me by Electrophysics Corp who make the night vision devices which provided the Iraq war news night-vision footage.

"""""
Clearly the prism/mirror approach would be lower cost and better performance. There is no existing production of image twisters having 35mm format. In addition, performance will be degraded in several areas:

1)

There will be a significant reduction in effective F-number from the coherent fiber bundle (they usually have about 30% transmission);

2)

Resolution (MTF) will be significantly reduced due to the limited size of each fiber;

3)

They will be costly.

"""""

Justin, in principle yes. Fact is, up to now we had make all products only for use in productions, not for sale, so far. At this time we make some examinations with our products (ISO, CE ...) A must for the market. This need a little time.

Arround the 35mm solutions, i think this products are possible:

Complete units and/or

Self-made-GG-sets with m42 on the front side, maybe in two versions, static and with vibration drive (because vibration make the best GG better). A low-cost version for full PAL and a special HDTV version but this not for self made.

Prism-Set, suitably for the backside for UP-Side-Down correction

Relay-Lens-Set for XL1 or other

Motor driven follow-focus system for all kinds of still camera lenses. Controllable over a poti on the side of the GG unit, or wireless (For steady cam users)

All part made for our rail mount system.

The products are here on my table, but yet i dont know details about prices or delivery.

the problem is there are already several company making such product and we are not purchasing from them because it is too expensive. I do not really see why you would give yours for less than theirs.
The first thing that your market Dept. will do is to compare with equivalent product and price your stuff in order to get the money back, then still too expensive for us.....infinite loop, as soon as you are trying to do a business with something, it does not interest people that are not willing to purchase finished products.
My static mini cost me less than 100$ and i will probably add 100 for the moving stage. if it should cost me even less than 1000$ (that is actually pretty nice) i would not even started to read all these posts.
Probably you are in a different category, but i am in the "nice to have , as long it doesn't cost me too much". Additionally the reward of doing it by myself is bigger than just spare money.
And if i am able to justify it professionally, i would probably simply rent the real stuff.
I think we should focus on this because it is the start of everything.

ok , that was few hours ago, now i am back home and found the perfect roller for the project.
initially it is a metallic roller about 12mm dia.that is used to drive a belt. the axis is about 12 mm long , 2mm dia , threaded on each end. when you remove the flange, surprise... it contains 2x 7mm roller bearing. in fact not really... the axis has thread of 2mm dia on each end , but is larger in the middle (about 4mm) and serve as core for both the rollers bearing. the greatest thing is that there is 2mm of this axis available between the two roller so i can put the belt here. The three plate design (with a spacing of only 2 or 3 mm) takes all the meaning here with almost no work than drilling holes. additionally, i will use one of the threaded end to fix the motor on one of the axis.

On the other hand, if you can build a business on it , good luck for you.

Rai,
It seems you would have to hit a very low price to interest people on this forum. I offered a kit for a 4 shaft 12 ball bearing kit, with motor,belt, misc parts to make a complete GG orbiter for $350, and there was ZERO interest. It's not a pipe dream on paper on my end, as my posted pic have shown.
Perhaps a price of $100 might work here. Or maybe we let them try to make one for a while, and then see what happens :)

It's *not* the cost of the parts. The labor of making the custom parts is the price here.

-Les

Les,
If you want to sell one for $100, give me a shout.

Hey Les,

I am interested in you orbiter. But $350 is a lot of money for no guarantee.

You see, I've been following these 35mm adapter threads from the beginning. Every time someone came up with a new idea, I went out and bought something different hopeing I'd finally have the holy grail of adapters. I've spent a lot of time and money on aluminum oxide, filters, filter rings, condenser lenses, old camera lenses, the list goes on. I still don't have a working adapter. I sure wouldn't want a $350 vibrator to be the next dead-end purchase for me.

I think you might have more luck if you lower your price to start with. This would help you get a user-base established who can then vouch for the validity of your product. Once there is more demand, you can raise the price to more effectively cover your labor.

Anyways- I am interested. How big is the target size on your vibrator? Will I be able to use medium format lenses on it? What kind of battery does the motor use?

Jesse,
I am sorry that you wasted your money on some of the other GG attempts.
Let me just say something about evaluating them:

1) Posting still images of a GG video is WORTHLESS for evaluating the grain. Period. You need to see a slow panning shot of the GG in use, preferably with some of the scene out of focus a bit and shadows and highlights. Only then do you have a chance to see the grain pattern because it remains in the same place on your monitor while the image moves past it.
So you shouldn't ask for frame grabs of the GG for evaluating it for grain.

2) Posting tiny video of some highly compressed format is also worthless for seeing grain effects. The first thing that a compression codec throws away is the grain pattern! The poster should use a 7 megabit mpeg4 or Media9 video to show the details. Then you can see possible fixed pattern grain effects.

OK,
Having said that, how about I post a small demo video, at a high bit rate Media9 ? My camera is an HD-10 JVC, if it works for that, it will work for DV res. no problem. The HD video will be 1280 X 720 pixels. I can also crop into it and make a smaller version for playback on slower computers. I'm open to suggestions on what to do for the Mac. I hear they play media9 now, but with some difficulty.
So, isn't the proof in the pudding, as they say? ( when the pudding is properly compressed )

-Les

Jesse: I understand exactly where you're coming from. Although you don't see many posts from me, I have been following developments and have spent money on glass, aluminum oxide, wax, etc... only to discover a couple of days and few posts later that there's a better way to do something.

HOWEVER, Les does have an actual device (not just something on paper), and I think the price he's asking is easily justifiable for the parts and labor involved, even if it's not in a housing. I don't see a reason why a 5 micron ground glass mounted on a rotating device would not work. However, I am building my own since it is just fun.

Les, have you tried advertising this on other forums where there are less technically minded people? (ie. consumer type people) Well, I don't mean that in a bad way. I just mean...people that don't like to build these things themselves, since that is what a majority of us guys in this thread are, I think.

Les-

A video clip would be great. I can handle high-res.
Also how big is the target area? Is the vibration isolated in the adapter? Does it shake the camera at all?

thanks
-jes

The cutout size right now is about 42mm X 30mm, but I want to switch to a 4 shaft system that will leave a little larger area. Right now the hole is not centered top to bottom either, another reason I want to change the design. The motor seems to run fine on 6V , four AA batteries are adequate.

There is a small amount of vibration, it's impossible to get rid of it all. I'm going to soft mount it in the housing, to minimize external vibrations. This must be done in such a way to avoid any z axis motion, which would ruin focus.
I haven't attempted to advertise anywhere, as this is really more for the type of person that can adapt it to their own use. It would not ship with a GG, for legal reasons.
I'll try to make some demo video in the next few days.
-Les

Les do we all get royalties for our contributions? ;-)

Im only kidding. Good luck.

-Brett Erskine

my roller ball
I will try to get more of them.
http://www.giroud.com/minidv/roller.jpg

Where did you pick up your bearings/pulley wheel?

frankly , i do not know.
i got a friend who got a box full of this, a few month ago, a take five of them to see if they could fit my needs for my video crane.
since the crane is going to a no-cable version, i will probably use these little marvels to the vibrating gg project.

Les,

How's that video clip coming? I'm anxious to see it.

It's finally the weekend, I'll get a chance to work on it!

-Les

<<<-- Originally posted by Jesse Rosten : Les,

How's that video clip coming? I'm anxious to see it. -->>>

Update: I tested with a lens and GG and discovered a very slight ( about 0.0015 inch ) off axis motion that causes the image to move laterally a little. I think one of the shafts is off center, I have to take it all apart and check, I'll probably make a new backing plate for the bearings. This thing has to be very precise for sure to get a good image. Also, I think that a fairly slow rpm will work, the grain moves by fairly fast with my 1.5mm radius.

So no video clip yet.
-Les

Les.

If your device is floating on rubber mounts, you may find you need counterbalance on either side of load centerpoint on each shaft, not just one side, otherwise the counterweights themselves may be creating the abberant motion you describe. It may not be a shaft running out.

Depending on how much of the oscillating mass you manage to balance out you will still get vibration but it will occur directly and not obliquely across the image axis. Make sure wires and other attachments come off at the load centre as well.

If your device is rigidly mounted, then you may have clearance in the bearings being taken up, again by the same offset counterbalance problem.

I am assuming that the bearings are a secure fit on their shafts and crankpins and the housings they fit into and cannot float. If not then you will need to limit movement along the image axis with some sort of guide frame. If this is already part of the design then ignore this commentary.

Don't take my ramblings too seriously as I am not an engineer. I do know that small 2 stroke engines with a half-crankshaft (like the oscillating groundglass devices) vibrate cruel-bad compared to a full crankshaft version of the same power and cubic capacity.

Les -

I'm still interested in your oscillator. If I disappear for the next three weeks it's because I'll be on location in Africa <plug plug> :)

peace
jes

Jesse,
I worked on it a bit this weekend. It still has some Z axis deflection that I'm trying to minimize. I can test it by looking at the mirror reflection of the GG shiny side, at a grazing angle. Even the smallest deflection, under 0.001 inch is enough to move the image around, indication that there is some Z axis motion. I'm thinking one of my eccentric shafts is off, I think I'm going to remake them, out of brass this time ( easier to machine ).

I want to post some tests soon. HD is very demanding, I don't want to diminish it's sharpness with my DOF adapter!

Africa... be safe, it's rather lawless in many places!

-Les

<<<-- Originally posted by Les Dit : Jesse,
. I can test it by looking at the mirror reflection of the GG shiny side, at a grazing angle. Even the smallest deflection, under 0.001 inch is enough to move the image around, indication that there is some Z axis motion.
-Les -->>>
Les,
Try to use a small laser pointer and look at the reflection on a screen or surface away from the disk. The "amplification factor" will make the measurement of the deflection easier.
Good luck

0.001... laser...
you start to understand now why these gadget cost the hell to build. I think we reach the point were the usual workbench in the garage is not more of any use.
try some small springs to keep all the stuff in place.

<<<-- Originally posted by Giroud Francois : 0.001... laser...
you start to understand now why these gadget cost the hell to build. -->>>

Giroud,
A laser pointer can be purchased from gift shops or night market stalls for ~3-5 dollars or for less euros. They come in various flavors, even as keyrings.

... and i already got so many of them.
They are fine to kick the cat, but i doubt they can be of any use
for chasing the .0001 inch.
When you start to deal with real fine mechanics (i used to deal with micron when i was in tool machining), you need some serious tools.

The laser idea is good. Just looking at a distant object using the GG shiny side does the same thing, really, but the laser would make it easy to actually measure the thing.
Springs: I am already taking the bearing lash out by using a pair of magnets and steel disks.
I don't think I need 0.0001 tolerance. Just a bit better than what I have now.

-Les

hi everyone I just recived a piece of 5x9 ground glass from ebay seller .It looks real good. he has quite a few pieces. my winning bid was about ten dollars. I tried to have a circle cut once ,it has to be perfect. but maybe this could be of use. please let's share this.




http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=15247&item=3830724092&rd=1

I now have my GG orbiter working. I redid the shafts and made them longer. There is no noticeable image vibration or shift, and it looks like it holds HD focus now. My GG has some rather gross pits in it when viewed static, but those look like they are gone when it's moving. I have no corrective optics to minimize the center hot spot, and I flip the image in the computer.

This weekend I will post a link to some video which I'll shoot on Sat.

-Les

Les-
Whats the target size your actually focusing on?
It should be 36mm X 24mm but without corrective optics you might have to zoom in farther than that.

OK, my first test footage is up. Media9 and Mpeg2, both the same test. See the new thread:
http://www.dvinfo.net/conf/showthread.php?s=&postid=209863#post209863

-Les

Heey!!

I'm preparing to give this oscillating adapter a try, but I'm unsure if it's possible to oscillate my GG condenser, since it might be too heavy..

.. but let's say I get around this somehow ( a more powerful motor ? ), do you think I would see some wierd optical distortion because I'm oscillating my last condenser? .. I'm thinking it will scatter the incoming light all over the place .. resulting in a hotspot .. but I'm just guessing! .. anyone know for sure?

Martin-
Yes it will. Dont do it.

alright, brett, thanks for clearing that up!

by the way, do you have any footage with your orbiter in action ? I'm very curious!

I've been too busy to work on it lately. Trust me as soon as its done I'll post it for you guys.

About vibrating the condenser dont do it for all the reasons you mentioned. The way to do it is make the GG and the Condenser two seperate pieces and rotate the GG only. This will solve all of your problems. This is also how the Mini35 works.

How have you guys been working out how much mass to be offset for the counterbalances? What are you basing these calculations on? As small as they seem to be, it seems too easy to over compensate and cause instability and vibration.

I just guessed on my first pass. I could see calculating the correct amount, but there are other things I would redesign before I make a new version.
The vibration I had on my first test, which I posted a while ago, is reduced when it is mounted in a more secure manner.
-Les



<<<-- Originally posted by Brandt Wilson : How have you guys been working out how much mass to be offset for the counterbalances? What are you basing these calculations on? As small as they seem to be, it seems too easy to over compensate and cause instability and vibration. -->>>

Is it possible to mount the motor on the same stationary plate as the three other bearing/shaft/counterweight assemblies? Sort of taking the place of one of the corners in the 4 shaft design. One possible problem is that the motor may get in the way of some of the optical components. You could always make the spacing between shafts a bit larger, but because of the larger overall design, that may defeat the purpose of having all of the stationary components on a single plate. I'm also assuming that the motor would have to have an offset shaft, or be fit into some sort of pin which would offset things appropriately so that the motor wouldn't get beat up. P+S probably went through all of the variations that have been talked about in this thread, and ended up with the one they're using because it seems to be the most efficient and stable design.

Les, your footage looks really nice. Especially under the conditions which the assembly was mounted. Are you using a condenser in that shot, or are you just shooting through your GG and primary lens?