Patent: Dual Motor For Autofocus on STM Lenses

Canon Rumors Guy

Canon EOS 40D
CR Pro
Jul 20, 2010
10,779
3,158
Canada
www.canonrumors.com
HTML:
<p><strong>UPDATE

</strong>This is what we’re told this patent actually is.</p>
<p><em>“This is not a focus drive patent, it’s a SPRINGLESS LEAF SHUTER mechanism. EF Leaf shutter lens that will allow a true 1/1000s are on the drawing board, they just can’t make the Cmos do a clear scan and dump faster than 1/250s no matter what they have tried , and HSS make you loose way too much power.”</em></p>
<p><strong>Original Post

</strong>A patent showing a dual motor autofocus for STM lenses has appeared. There is one motor for the STM and a secondary DCM motor. It looks like one motor assists starting the second motor for a faster and smoother operation. The patent information below is Google Translated. If anyone can clarify the patent, please do so in the forum thread associated with this post.</p>
<p><a href="http://www.canonrumors.com/wp-content/uploads/2014/09/newstm.png"><img class="alignnone size-full wp-image-17272" src="http://www.canonrumors.com/wp-content/uploads/2014/09/newstm.png" alt="newstm" width="219" height="426" /></a></p>
<ul style="color: #444444;">
<li><strong><span class="goog-text-highlight">Patent Publication No. 2014-164106</span></strong>
<ul>
<li>Published Date 2014.9.8</li>
<li>Filing date 2013.2.25</li>
</ul>
</li>
<li><strong>Example 1</strong>
<ul>
<li>Moving the aperture one two motors</li>
<li>Main motor STM</li>
<li>Auxiliary motor DCM</li>
<li>And then rotate the first auxiliary motor, to assist the acceleration of the main motor</li>
</ul>
</li>
</ul>
<p>Source: [<a href="http://egami.blog.so-net.ne.jp/2014-09-13" target="_blank">EG</a>]</p>
<p><strong><span style="color: #ff0000;">c</span>r</strong></p>
 

jrista

EOL
Dec 3, 2011
5,348
36
jonrista.com
Rocky said:
The pattern is for aperture control (closing and opening). Not for auto focusing.

Agreed, it looks like it has something to do with improving the response time of the aperture blades. From what I understand, there is an apparatus the blades sit inside, and that apparatus rotates (or maybe the blades rotate) to control how far stopped down you are. It sounds like this patent is aimed at increasing the speed at which the aperture stops down. I would guess that's an attempt to minimize the overall lag time between pressing the shutter button, and actually exposing the image.
 
Upvote 0
BozillaNZ said:
It's for step-less aperture operation, not AF

One motor (left) for normal stepping aperture in 1/3 stop increments, one motor (right) for smooth step-less aperture change.

More of a video recording gadget than still image.

Are we generally to dismiss things you have to say as pure conjecture disguised as certainty or do you know something contrary to the above information? Such a trait in life is only a hindrance to yourself and everyone around you who doesn't know to be more dismissive of you, consider doing that less, perhaps? ;)
The reason I say this is that the last bullet point clearly states one motor is to assist another with acceleration... and I apologise for being an ass to you but people have buttons, just like cameras do, and this kind of insistent tripe disguised as informed fact pushes mine in the context of people trying to gain an education and understanding.

But yeah, I'm a little surprised the CR folk didn't clock the word "Aperture" or recognise anything from the diagramme as aperture control hardware.

jrista and pin008 sound like they're barking up the right tree! Now I've to obsessively research these terms I don't know already; Canon EMD aperture? DCM motors?...

Also, if anyone's curious and isn't familiar with the design it appears that 7a are pins which the aperture blades pivot around and slot into 3b. 7b are little pins attached to the blades, these rest in 6a. Parts 3 & 7 rotate, 6 remains static and the blades open and close as the pins 7b travel along slots 6a.

I've aaalso just noticed the motors operate different cogs, 3a & 4A, hence their varying axel length (is 'axel' correct here?)
I'm wondering if 4 is perhaps relatively loose, offering enough friction to add to the acceleration but not so much as to be in conflict with the precision offered by the other motor?
 
Upvote 0

jrista

EOL
Dec 3, 2011
5,348
36
jonrista.com
Here is a translation from the linked source:

The continuous shooting speed reduction due to narrowing

Canon EOS system of several, including the adopted electromagnetic diaphragm, continuous shooting speed when you've narrowed down I will be lowered. Even without continuous shooting, release time lag of when you've narrowed down so lowered, It is a specification that anxious. Canon It is a thing of the patent applications that accelerate the narrowing and opening recovery time by two motorized. Reduction of continuous shooting speed by narrowing has been pointed out as disadvantages of EOS, but such concerns I might not.

Pretty bad translation, but it seems to make it clear that this patent has to do with increasing the speed at which the aperture stops down, reducing shutter lag, thereby increasing the maximum potential speed for continuous shooting.

I do not believe this particular patent is related to providing smooth continuous aperture control (which would be valuable for video)...it does indeed sound like a shutter lag related thing.
 
Upvote 0
jrista said:
Here is a translation from the linked source:

The continuous shooting speed reduction due to narrowing

Canon EOS system of several, including the adopted electromagnetic diaphragm, continuous shooting speed when you've narrowed down I will be lowered. Even without continuous shooting, release time lag of when you've narrowed down so lowered, It is a specification that anxious. Canon It is a thing of the patent applications that accelerate the narrowing and opening recovery time by two motorized. Reduction of continuous shooting speed by narrowing has been pointed out as disadvantages of EOS, but such concerns I might not.

Pretty bad translation, but it seems to make it clear that this patent has to do with increasing the speed at which the aperture stops down, reducing shutter lag, thereby increasing the maximum potential speed for continuous shooting.

I do not believe this particular patent is related to providing smooth continuous aperture control (which would be valuable for video)...it does indeed sound like a shutter lag related thing.

Thanks for the info, jrista! It doesn't appear to be concerned with video, with the emphasis being on acceleration, but I'm curious about past/current designs utilising stepping motors. Anyone know if this dual motor design is a first for STM incorporation in aperture control or is that element old news?
 
Upvote 0
Jul 21, 2010
31,099
12,863
pin008 said:
canon EMD aperture needs 55ms for a full narrow down.

I'm not so sure that's the case. Most lenses can stop down between 6-8 stops from wide open. My 1D X manual states, "The normal shutter release time is approx. 0.055 sec when the aperture is stopped down by no more than 4 stops. That suggests that more than 55 ms may be needed for the 6-8 stops to fully stop down the aperture.


jrista said:
Pretty bad translation, but it seems to make it clear that this patent has to do with increasing the speed at which the aperture stops down, reducing shutter lag, thereby increasing the maximum potential speed for continuous shooting.

Exactly. The 12 fps of the 1D X is specified for ISO 25600 or lower, 1/1000 s shutter or faster, and aperture wide open. Higher ISO, slower shutter, or stopping down can reduce the max frame rate. Looks like Canon may address one of those limitations in future lenses.
 
Upvote 0
Khufu said:
I've aaalso just noticed the motors operate different cogs, 3a & 4A, hence their varying axel length (is 'axel' correct here?)
I'm wondering if 4 is perhaps relatively loose, offering enough friction to add to the acceleration but not so much as to be in conflict with the precision offered by the other motor?

That aspect is making my scratch my head as well. I would not think a friction engagement since environmental conditions would change the behavior of the shutter and you could end up with the body running out of sync with the aperture. Ring 4 seems to fit into the space between 3a and 3d. There isn't any obvious indexing apparent in the diagram for ring 4. I am somewhat surprised to see 4a run the entire circumference of ring 4, but that might just be for ease of assembly rather than indexing. The small holes at the base of tabs 3d might be the indexing points for ring 4, which would be at the inside and below, and therefor not visible on the diagram.

From :
http://www4.ipdl.inpit.go.jp/Tokujitu/tjsogodben.ipdl?N0000=115
Problem to be solved by the invention

In recent years, the demand of the imaging devices which can perform both static image photographing and animation photography by high definition is increasing. Therefore, the imaging device provided with the light volume adjustment which combines the smooth low-speed-driving performance at the time of animation photography and the high speed drive performance at the time of static image photographing is called for strongly. On the other hand, in the technology described in the above-mentioned Patent document 1, the microstep drive of a stepping motor performs smooth low speed driving in animation photography, securing the high speed drive performance for a still picture seriography as above-mentioned by 1-binary phrase drive of a stepping motor.

However, since the electromagnetic drive collimator which is a light volume adjustment described in the above-mentioned Patent document 1 is driven with one stepping motor, high speed drive performance and low-speed-driving performance will be rate-limitted by the performance of a stepping motor. That is, if the magnetic flux density of a rotor magnet is improved so that the performance of the high speed drive demanded by one side may be satisfied, smoothness will be impaired by increase of cogging torque in the smooth low speed driving demanded on the other hand. Therefore, in the light volume adjustment driven with one stepping motor, it is forced the design which balances high speed drive performance and low-speed-driving performance.

When switching animation photography and static image photographing and performing them for example, smooth low-speed-driving performance and high speed drive performance are required substantially simultaneous from a light volume adjustment. It is difficult to be able to say that this demand is one of the severest driving conditions to a stepping motor, and to attain simultaneously the smooth low speed driving which is the commercial-scene request in recent years to a light volume adjustment, and the further high speed drive in one stepping motor.

An object of the present invention is to provide the light volume adjustment which improved the high speed drive performance with the low-speed-driving performance maintained.

So this really does seem to be an attempt to improve the low-speed smoothness of aperture control while shooting video rather than improving the response time (or at least, not hindering it) during high-speed still-shooting.

Taking a look at the velocity response curve of this design (top, a) versus the standard design (bottom, b)
http://www4.ipdl.inpit.go.jp/TD2/web014/IMAGE/20140914014552003688.gif
What gets me is the reversal that occurs, which would seem to be less smooth. The aperture expands sooner, stops sooner, and then contracts and stops again. It is listed as an example, so maybe they just haven't tuned the response just yet. However, high-speed shutter response for continuous still shooting would be a concurrent advantage given this response curve.
 
Upvote 0
Feb 12, 2014
166
1
I havent read any of the documentation, but the drawing is fairly clear and i think it makes sense... One motor drives the ring (4) that operates the aperture via pins at the pivot point (7a). One motor drives the short rack (3a?) that operates the aperture via the standard helical motion of ring (6) , the slots in it, and pins 7b. I would assume the short throw-> ring operation is for smooth video changes, and the rotating ring on pins method would be for stills where you want the fastest operation. the gearing from the ring (4) to the pins would give the shutter a very fast action, the motor would only need to step maybe 3 or 4 teeth on the ring.

could it point to a new STM tele lens to go with the 7dII? all new aps-c lenses have been stm, but perhaps stm on a tele is too slow for 10fps?
 
Upvote 0
(Slightly modified) translation from the Patent.

This patent concerns a light flux adjustment device (i.e. an aperture) for use in digital video cameras and digital still cameras. In single reflex cameras, the aperture must first stop down and then open as fast as possible after the release button is pressed. In recent years, cameras that can take both moving and still pictures have become popular, and there is a need to actuate the aperture slowly and smoothly in the former case and quickly in the latter case. In the prior art involving the use of a single stepping motor, the fast and slow speed requirements interfered with each other; when the magnetic flux density of the rotor was increased to satisfy the fast-speed requirement, the smooth movement at slow speeds was hampered due to the increase in the cogging torque. Thus the designer was forced to make a trade-off. The present invention maintains performance at both fast and slow speed regimes.

The abovementioned first and second motors can rotate the rotational mechanism independently at an arbitrary speed and direction, irrespective of each other. The two motors can also be used in unison to achieve an acceleration and maximum rotational speed above the capability of a single motor. One of the motors can rotate in the reverse direction w.r.t. the other, thereby reducing the time needed for the rotation mechanism to slow down and come to a halt.

The main stepping motor 1 is fixed to the frame 5, and its actuation gear 1a engages with gear 3a, thereby transferring torque from motor 1 to the rotational mechanism 3. The blade feathers 7 have tabs 7a and 7b; 7a is connected to hole 3b, whereas 7b is connected to cam slots 6a on the lower cover 4. When motor 1 is actuated, mechanism 3 rotates and tab 7b is guided along cam slots 6a. The feather blades open and close in this way.

Secondary motor 2 is fixed to cover 5, and actuated by a control system in the camera (not shown). Its gear 2a meshes with the teeth of rotational mechanism 4, which in turn rotates around the optical axis along a rib on the main rotational mechanism 3. As shown in Fig. 2, the rotational mechanisms 3 and 4 are in contact at least at one point along the outer and inner rims 3d and 4d, thereby transferring frictional force. Since 3 and 4 do not mesh by teeth etc., they can move freely irrespective of each other. For example, even if motor 2 rotates mechanism 4, as long as motor 1 produces enough opposite torque to counteract the frictional force, mechanism 3 can be prevented from rotating.

As shown in Fig 3(a), prior to rotating mechanism 3, the auxillary motor 2 rotates mechanism 4 (note at this point, mechanism 3 does not actually begin to rotate since 3 and 4 are not locked). When next motor 1 is actuated, less torque is necessary to begin the rotation of 3 since the static frictional force is reduced due to the presence of motor 2. In this way, a rapid acceleration is ensured.
 
Upvote 0
Wizardly said:
That aspect is making my scratch my head as well. I would not think a friction engagement since environmental conditions would change the behavior of the shutter and you could end up with the body running out of sync with the aperture. Ring 4 seems to fit into the space between 3a and 3d. There isn't any obvious indexing apparent in the diagram for ring 4. I am somewhat surprised to see 4a run the entire circumference of ring 4, but that might just be for ease of assembly rather than indexing. The small holes at the base of tabs 3d might be the indexing points for ring 4, which would be at the inside and below, and therefor not visible on the diagram.

The patent notes:
The positional precision of the blades is established by motor 1 which is a stepping motor; the secondary motor 2 only provides auxiliary speedup and slowdown, and hence an economical DC servo motor will suffice. The degree of friction between the rotational mechanisms 3 and 4 must be adjusted to correspond to the torque of motor 2, for example by applying a suitable treatment on the inner and outer surfaces of the moving parts (i.e., roughing it up?).

The patent doesn't say, but I suppose if the main motor has some rotary encoder feedback (e.g., the shutter does not open until the aperture blades stop moving and locks into the correct position), then there is no danger of the camera running out of sync even in various environments.
 
Upvote 0
lambda100 said:
Wizardly said:
That aspect is making my scratch my head as well. I would not think a friction engagement since environmental conditions would change the behavior of the shutter and you could end up with the body running out of sync with the aperture. Ring 4 seems to fit into the space between 3a and 3d. There isn't any obvious indexing apparent in the diagram for ring 4. I am somewhat surprised to see 4a run the entire circumference of ring 4, but that might just be for ease of assembly rather than indexing. The small holes at the base of tabs 3d might be the indexing points for ring 4, which would be at the inside and below, and therefor not visible on the diagram.

The patent notes:
The positional precision of the blades is established by motor 1 which is a stepping motor; the secondary motor 2 only provides auxiliary speedup and slowdown, and hence an economical DC servo motor will suffice. The degree of friction between the rotational mechanisms 3 and 4 must be adjusted to correspond to the torque of motor 2, for example by applying a suitable treatment on the inner and outer surfaces of the moving parts (i.e., roughing it up?).

The patent doesn't say, but I suppose if the main motor has some rotary encoder feedback (e.g., the shutter does not open until the aperture blades stop moving and locks into the correct position), then there is no danger of the camera running out of sync even in various environments.

That would probably explain the negative velocity - reversing the inertia of the friction-drive auxiliary ring. I can't help but think of the havoc this is going to play on the poor fool who left his lens in his car on a sunny day - not only does he have oil on his aperture blades but oil on his auxiliary ring.
 
Upvote 0
Feb 12, 2014
166
1
"As shown in Fig 3(a), prior to rotating mechanism 3, the auxillary motor 2 rotates mechanism 4 (note at this point, mechanism 3 does not actually begin to rotate since 3 and 4 are not locked). When next motor 1 is actuated, less torque is necessary to begin the rotation of 3 since the static frictional force is reduced due to the presence of motor 2. In this way, a rapid acceleration is ensured."

theres the kicker. motor 2 is preloading the aperture so the stepper doesn't have to work as hard, and making it so the stepper is not overcoming friction, just controlling the throw.
 
Upvote 0

dr croubie

Too many photos, too little time.
Jun 1, 2011
1,383
0
Canon Rumors said:
“This is not a focus drive patent, it’s a SPRINGLESS LEAF SHUTER mechanism. EF Leaf shutter lens that will allow a true 1/1000s are on the drawing board, they just can’t make the Cmos do a clear scan and dump faster than 1/250s no matter what they have tried , and HSS make you lose way too much power.”

Well, it might not work on my 7D if the sensor limit is 1/250s, but it would make it nice to get real flash sync using Portra on my EOS3 without adapting MF Leaf-shutter lenses (at which point, I'd just use an MF body and be done with it).
 
Upvote 0
Olympus will introduce tomorrow the first lens with dual VCM for autofocus.

http://www.43rumors.com/precision-camera-discloses-full-olympus-40-150mm-lens-details/

TWO IS BETTER THAN ONE
Smooth, rapid autofocus is a crucial factor in image quality. The 40-150mm f2.8 PRO’s innovative, industry-first Dual VCM (voice coil motor) AF delivers. Separating the focusing mechanism into two lens groups controlled by individual linear drive motors makes AF performance extremely fast and, with no gears involved, nearly friction-free and incredibly quiet.
 
Upvote 0
hoodlum said:
Olympus will introduce tomorrow the first lens with dual VCM for autofocus.

http://www.43rumors.com/precision-camera-discloses-full-olympus-40-150mm-lens-details/

TWO IS BETTER THAN ONE
Smooth, rapid autofocus is a crucial factor in image quality. The 40-150mm f2.8 PRO’s innovative, industry-first Dual VCM (voice coil motor) AF delivers. Separating the focusing mechanism into two lens groups controlled by individual linear drive motors makes AF performance extremely fast and, with no gears involved, nearly friction-free and incredibly quiet.

That is something entirely different than this patent. The motor described here drives the diaphragm to stop down the aperture, which is quite different than the focusing groups in a lens.
 
Upvote 0