As I progress with my build I am doing research into building my own brushless gimbal frame using spare carbon fibre and control based on the Alex Mos board http://www.simplebgc.com/eng/. I have spent the last week looking at suitable motors.
Looking at Zenmuse and the AM gimbal the idea of flat dc motors was appealing aesthetically yet expensive on the basis of purchasing very good Maxon motors that would not require me rewinding the copper. Most third-party manufactures of complete AM brushless gimbals use these motors. From a DIY perspective 40 GBP (60 USD) is quite expensive.
I also knew that there were potentially better motors out there that have higher resolutions. Ie piezo motors used in autofocus SLR units, engraving machines, medial machinery. The advantage is the higher resolution, response speed and power-saving features. A supplier in the UK of the rotary motor manufactured in Sweden http://www.piezomotor.se/products/rotary/ quoted 1,200 GBP for the 50mNm and 600 GBP for the 80mNm.
However I have now ruled this out as the distributor has informed me that I would require a 4 phase motor driver to control the motors.
Other than using a sub $10 Turnigy low kv/ low profile motors and rewinding the copper has anyone out there found a better alternative high resolution motor for their brushless gimbals?
Comments
Many people are re- winding their own motors for use in gimbals.
Instructions and lots of other good info here.
http://www.fpvhub.com/index.php/topic,11021.0.html?PHPSESSID=9d6e94...
iFlight has also released a series of iPower Brushless Gimbal motors:
http://www.iflight-rc.com/Gimbal-BL-Motor.htm
Stepper motors do have some advantages, for one thing you can actually move them a specific number of steps without having to have closed loop sensing of position. (So long as there is no slip).
Also they can be microstepped up to 64 microsteps per step, though the truth is that it gets a bit assymetrical and sloppy and is largely dependent on how precisely the motors are built so 32 or 16 microsteps are a more practical upper limit.
For this application you would probably want to either gear down or, possibly better use a toothed timing type drive belt to increase resolution and reduce step ratcheting.
That said, 200 step per revolution stepping motors are very cheap and microstepping ICs that can handle 2 or 4 of the motors are available inexpensively.
You can use a switch on one end to establish a start point to count from and have it auto initialize by turning till it hits that switch.
The CNC guys already know this.
Some advantages are high holding torque without power (At least at full step increments), built to last a really long time, no brushes, pretty cheap and full solution driver ICs.
Disadvantages are tend to be a bit heavy and you need to do something to convert PWM to step control (if you really want to use PWM anyway).
In any case, you need some different and specialized control.
None of this is a really big deal and steppers could easily be used very well as camera gimbal movers.
Our RC servos have lots of problems for this because of gear lash from really large gear trains, mechanical brushed motors and sensor potentiometers and mismatch between servo resolution and speed from actual flight controller PWM update speed.
Interestingly a servo that is faster (read digital) than the actual PWM update rate actually compounds the twitchy problem by letting the servo completely start and stop between PWM updates.
Steppers may well be the best solution for this for now.
http://www.applimotion.com/ult.php
these look quite interesting
I looked the T-Motors gimbal motors look nice
How much did the GB2208 Cost and do you have a link to the supplier
Interesting - what's the max accuracy of such gimbals ? in mrad ?
Indeed after doing my research into regular servo, v stepper v brushless the key is resolution of movement/ backlash but have to be rewound. Obviously there are gimbals that use gears or levers to reduce this but reduce responsiveness and increase gimbal size and weight. No matter how good the electronics driving the motor stepper or not the motor has to have the accuracy and resolution hence why I initially thought of piezo motors. For a dc motor to get there they are either rewound or have 4 to 6 phases controlled by a specialised controller.
Steppers, step from one magnetic pole position to the next.. And guess what brushless does pretty much the exact same thing, but they can be rewound so that you do standing wave positioning (static positioning in between magnets poles). The hard part is the controller doing proper micro stepping so that the motor smoothly and consistently move from one position to the next.
My t-motors guy just came through. Ordering me 4 GB2208. Made my day.
We are actually having a batch of gimbal motors being manufactured. These are custom wound motors ranging from gopro size to SLR gimbal size motors so keep an eye out for them soon!
-
1
-
2
-
3
of 3 Next