I am not sure what the "DC" actually stands for in the official marketing media article of the new DA 18-135 DC. But typically DC motor just refers to Direct Current Motor after all in technical sense.
But then the above press release also mentions about its "SDM autofocus operation" (but the lens now has no SDM designation btw), so now I take and (re-)understand SDM just only means Silent Driving Motor then, in contrast to previously told long form of Supersonic Driven Motor or whatsoever / whatever you or the Pentax marketing guys like to call it!
Below maybe the patent and detailed design that used in the new lens for that "DC" "ring-type" AF motor, have a look, it is not too difficult to be understood if you really want to study:-
http://www.freepatentsonline.com/20090323208.pdf
In short, it is just a "conventional" DC Motor with magnets arranged in ring-form, continuously and in cyclic, which is yet somehow unconventional then, as those permanent magnetics are now arranged "outside" the core/coil of the motor (in certain sense), but not inside in a self-contained unit and then the self-contained motor is used to drive anything "outside" as desired by/thro./via gears. As such, this new DC motor is simply nothing Ultrasonic and the working principle is totally different.
I would say it is a very clever design with low and old technology used. Let's see if it will perform and if it would be more durable than the old SDM motor units, which are proven to be rather vulnerable and slow which are just helpless, frankly speaking.
Putting aside the speed/performance issue, the other major concern is about its AF accuracy. My judgement is that if the motor has minimal inertia, then its focusing accuracy will be good. Otherwise the servo control and algorithms of the AF must be designed and written to be very good for the control model for a targeted/desired stop-down (curve) . The old SDM is proven to be not accurate for the stopped position and thus focusing accuracy is affected and sacrificed (which is actually worse than the old body-driven screw-type AF). Now, there is hope in the Pentaxland again! Let's see if this cleverly re-designed motor using the most primitive and old technology will work, or not! :-o :-)
Anyway, last but not least, I feel happy to see Pentax/Hoya really have done some R&D of their own. And most importantly, to realise their R&D results and put them into their new products! :-))
Logging you in...
Ronny · 755 weeks ago
Zebooka · 755 weeks ago
Old 11 points safox sensor (and don't lie me Pentax - it is not new, it is OLD).
Farewell Pentax, I need more professional gear.
RiceHigh 110p · 755 weeks ago
Michael A. · 755 weeks ago
Zebooka · 755 weeks ago
D7000 is more professional in terms of 3D AF tracking and far more advanced flash system. I need first, my wife needs both. No money and wish for two systems. That's all. I like Pentax, but it does not produce gear I need. That's why I'm leaving, not because I dislike it (well I hate AF and flash work on my K10D, but I believe that K-5 has at least good AF).
The other most important thing is price. D7000 will be 1.5x (or even more) cheaper than K-5 in Russia. And there are lots of friends using Nikon gear (that I can take and use if I need). And far more cheap old D lenses (50/1.4 D is almost 2x times cheaper, there is 85/1.5, there is FX camera and so on....)
Thank you :)
I love Pentax. I love FA limiteds (but I don't have them).
Michael A. · 755 weeks ago
There are already reports the AF of K-5 is GREAT even with 0.30 firmware, so that "little fellow" (I'm getting for sure) should be a treat; indeed.
It was VERY frustrating when I lost a several nice shots of Steeplechase in EU just because I had istD what couldn't keep up even at jpg setting.... Not the case anymore with this new incoming "devil"!
Watch out though, good N glass is very expensive, unless you go with 3rd party. Tell me about oldies, I still play on occasion with 8x10 LF gear. Just too much fun to let go and that look of bypassing people is PRICELESS.. :o)
RiceHigh 110p · 755 weeks ago
Zebooka · 755 weeks ago
I'm gonna to buy Canon s95 (as soon as it arrives to local shop).
And I have Mamiya C220 6x6 and couple of old Zenit cameras, aaand oldy *ist DS still rests on the shelf.
Hehe :)
John · 755 weeks ago
RiceHigh 110p · 755 weeks ago
R.E. · 754 weeks ago
Its basically a stepping motor.
They can choose to drive it with DC level signals and micro-stepping techniques which are all digital and high efficiency or AC drive.
The torque and hold capacity is very high. It is a non contact drive mechanism so will be far more reliable than ultrasonic drive and far less critical of machining to high tolerances.
The technology may be old, in fact its as old as multi-pole AC motors since that is pretty much what it is but it is also one of the most reliable and efficient electro-mechanical mechanisms available since the advent of low Rds FET's and rare earth magnets.
It is also very precise if they micro-step the drive mechanism.
Advantages:
High torque.
Reliability
Inherent manual override
Efficiency in converting power to motion
No need for position sensing as it can be driven openloop with one reference point
Accurate positioning
Simple construction using a well known technology
High speed
Can be used as an encoder when in manual override so it won't lose it position which would require a resync to the reference, just needs to have the electronics designed to have this ability (not hard).
Disadvantages
A bit on the heavy side but has high power to weight ratio so this would be design dependent.
Requires power to hold a microstepped position, though this can be overcome to hold on primary poles and have high speed microstep repositioning when needed with very little power consumption by good use of electronics design and software. We'll see if it gets implemented however.
Noisy and vibrates if not microstepped. This is eliminated if microstepped.
Basically all the negatives can be overcome with good design and for thought. Plenty of advantages to make a fantastic lens focus system if they choose to implement it.
They can even have calibrated distance measurement for each lens type allowing focus to be set at an exact distance via software if that is known because it can be used as an absolute encoding system once sync'ed and powered up or give a focus point in meters/feet in the file information.
Lots of potential.
I fully approve especially if they implement it to its full possibilities.
R.E. · 754 weeks ago
"Putting aside the speed/performance issue, the other major concern is about its AF accuracy. My judgement is that if the motor has minimal inertia, then its focusing accuracy will be good. Otherwise the servo control and algorithms of the AF must be designed and written to be very good for the control model for a targeted/desired stop-down (curve) ."
The control system is almost eliminated by use of this type of system. Note I'm using the term "control system" in a very technical sense.
It can remain open loop so long as its dv/dt remains less than a certain rate.
DC motors need sophisticated feedback loops and control systems to be accurate. This doesn't.
You just tell it where to go and it does. End of story.
The only times this will not apply is if the torque to move the drive exceeds the torque the motor has at any given RPM. Then it will slip and lose position. This is a better response than a DC motor which will start to pull higher current under loads and then stall drawing full current and stressing the motor/mechanicals/electronics. This system will simply slip once the torque is exceeded.
Since the drive has a lot of torque its not hard to stay well below this limit and still have useful response times and require no closed loop feedback.
Of course they can add encoding to it as well and have no resync issues under slip conditions.
RiceHigh 110p · 754 weeks ago
Thanks for your detailed inputs btw, which are all very interesting! :-)
R.E. · 754 weeks ago
Quick summary:
Microstepping allows you to slice the distance between each individual step to about 256-1024 depending on the system.
Lets look at what that means for a really basic stepper of four poles and 8bit microstepping.
Thats 1024 discrete steps per revolution for one of the most basic types of motor.
Now lets look at this design, admittedly very quickly.
It has ten poles and uses a hex encoding scheme to give 60 (6states x10poles) effective positions without any microstepping involved.
Now multiply that by 256 which is our rule of thumb bottom line figure for a good microstepping system.
We now have: 15360 positions per revolution.
Depending on how many revolutions it takes to go full focus travel will determine how many effective positions this system has as a minimum.
My guess would be that they will either microstep to about 512-1024 levels or if the gearing is done right they may be able to get all the fine movement they need using the base 60 positions.
Whats the point of making a stepping motor as the step size approaches zero?
Well, that is the ultimate goal for one.
If you have a step size approaching zero then you also have near infinite position capability and all with high torque and positional repeatability in open loop control.
The closer you can get the better.
BTW, a DC motor itself has discrete steps. Look at the commutator and brushes. Its steps are in the form of extra torque depending on angle. It has a very low step count per revolution. Certainly lower than this system has. The biggest difference is that a DC motor has no stable position so will continue to rotate by design. There is also no way to stop a DC motor reliably at any given point in its rotation. That error is always there and is usually corrected by gearing and positional feedback. The DC motor in this application is one that is inherently unstable and requires advanced feedback and control systems to get what is required here: Small repeatable incremental movement and large repeatable positional accuracy. Its the worst of the lot in terms of what it has going for it inherently in its design. Everything must be forced out of it.
So if you're assuming that the DC motor has better positional accuracy you are incorrect. It is all highly dependent on the gearing. Gearing creates backlash problems. DC control causes positional accuracy issues due to unstable step angle and no real control of where the motor will end up unless a complicated feedback network is used and you can never stop the motor at a particular angle, it will always be stopped with feedback after the motor via the gearing. Then you have error in the feedback path, its either encoded digital or analog which must be converted to digital with an ADC. This causes other issues.
Either way, you default back to less accuracy than with this system.
The stepping motor is inherently stable, accurate and repeatable with no backlash problems if not geared. If geared it will have less backlash issues because of the higher torque not requiring as many gears.
It really is the a near ideal system for focusing a camera lens.
All the features one needs or wants inherit in design and all the negatives are easily solvable.
BTW, ultrasonic drive is a form of stepping motor. The primary difference is that it relies on friction so it loses the most important criterion associate with stepping motors, their positional accuracy and repeatability. It is overcome in these systems by use of the focus circuitry driving it in an error feedback loop.
It cannot ever be designed to go from one point to another and be consistent in the number of steps it takes. It will always slip some small amount.
Its also has high manufacturing precision which means its also easy to damage.
Look at all the systems from first principles and tell me which is better for this application.
I have and am finally glad someone at Pentax came to the same conclusion as I did.
The stepping system has far more advantages and less disadvantages than any other.
I just hope they implement it to its full potential.
RiceHigh 110p · 754 weeks ago
R.E. · 754 weeks ago
I wouldn't call myself an expert, (expert is a term given by others to someone who knows more than they do but generally is meaningless and far more often overrated).
But, I do know about the different systems and what they can/can't do and decide based on their inherent properties and can work out which is the most suitable base for a design for any given application. Usually its not a one horse race, this one is. This is the best by inherent design for the application. If they implement it to its full potential it will be an amazing product having the potential to do things that the other systems simply cannot. e.g. presentable focus points. Fast and does not use the focus system. Perfect for video work. You can define multiple focus points and then fade from one to another smoothly and with variable rate. This also is fantastic for security applications etc.
If they miss the opportunities that the changing to this design gives them then it will be a better than average product. It will however be reliable unless they totally mess it up, which I doubt.
Fred · 754 weeks ago
Fascinating! This is information I have not read anywhere else. Hopefully the DC system lives up to its potential.
Thanks for bringing up the topic, Michael.
RiceHigh 110p · 754 weeks ago