HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Jun 27, 2010, 03:03 PM
Registered User
Joined Jan 2009
461 Posts
Steve,

I know your quadrotor brushless ESC project is probably going to be for sale, so I understand if you don't want to answer these questions, but I hope you don't mind my asking.

- Does your high efficiency come mostly from the regenerative braking, or do you have other techniques? And if so, are they hardware (e.g. more efficient switching power transistors) or software (better algorithms)?
- My understanding is that LiPOs can only take a certain recharge rate. Any concerns that you may be overcharging the batteries via the regenerative braking, and if so, how are you protecting against it?
- Most ESCs, I believe, just control the voltage going to the windings (via high frequency PWM), but what we need for quads is thrust control, for both increasing and decreasing thrust. Thrust is not easy to measure, but it should be proportional to RPM squared, and RPM is obviously measured via the electronic commutation process. Is this what you're offering? If so, do you have an inverse quadratic function on RPM command, so that the thrust is linear in the ESC command from the flight controller?
- If I recall correctly, the rotor RPM is a first order lag with a noticeable time delay following a step change in RPM command - rise time on the order of 0.1 second. I thought it might be possible to improve the overall response by providing some lead shaping on the command. Any thoughts on this?

I did have some questions on your flight controller as well, but I thought I'd post them under your "control loops in action" thread http://www.rcgroups.com/forums/showt...1229516(unless you'd prefer I post here). SOunds like we have very similar control response type strategies in mind.

Thanks
Roy
RoyLB is offline Find More Posts by RoyLB
Last edited by RoyLB; Jun 27, 2010 at 03:13 PM.
Reply With Quote
Sign up now
to remove ads between posts
Old Jun 28, 2010, 12:20 AM
Registered User
Meopham, Kent UK
Joined Jul 2008
136 Posts
Quote:
Originally Posted by RoyLB View Post
Steve,

I know your quadrotor brushless ESC project is probably going to be for sale, so I understand if you don't want to answer these questions, but I hope you don't mind my asking.

- Does your high efficiency come mostly from the regenerative braking, or do you have other techniques? And if so, are they hardware (e.g. more efficient switching power transistors) or software (better algorithms)?
- My understanding is that LiPOs can only take a certain recharge rate. Any concerns that you may be overcharging the batteries via the regenerative braking, and if so, how are you protecting against it?
- Most ESCs, I believe, just control the voltage going to the windings (via high frequency PWM), but what we need for quads is thrust control, for both increasing and decreasing thrust. Thrust is not easy to measure, but it should be proportional to RPM squared, and RPM is obviously measured via the electronic commutation process. Is this what you're offering? If so, do you have an inverse quadratic function on RPM command, so that the thrust is linear in the ESC command from the flight controller?
- If I recall correctly, the rotor RPM is a first order lag with a noticeable time delay following a step change in RPM command - rise time on the order of 0.1 second. I thought it might be possible to improve the overall response by providing some lead shaping on the command. Any thoughts on this?

I did have some questions on your flight controller as well, but I thought I'd post them under your "control loops in action" thread http://www.rcgroups.com/forums/showt...1229516(unless you'd prefer I post here). SOunds like we have very similar control response type strategies in mind.

Thanks
Roy
Efficiency from the switching techniques and mainly applies at half throttle regime.
There is little difference from a conventional ESC at full throttle.
The commutation software is a little more tricky.

The regeneration is only the short term recovery from prop inertia.
Banking probably absorbs this energy into the opposite motor more so than putting it back into the LiPo.
Certainly bench testing with a PSU does cause the main rail to overshoot
- a larger capacitor is necessary to control this.
With LiPo there is no issue.

The ESC is nominally set up for linear PWM duty cycle versus input command.
It is possible (and a reasonable thing to do) to linearise this to thrust for the Motor/Prop combo.
We now prefer to this in the flight controller algorithms.

Lead shaping sounds a very good idea - we haven't tried it.
I would actually use a gyro differential term and let the flight controller look after it.
Currently we use PI on angular velocity which works well most of the time.
Heavy props like APC14" may be better treated with a D term.
Care is needed since noise on gyro samples will inflict more jitter.

Flight control questions best put under the other thread.

Steve
airbotix is offline Find More Posts by airbotix
Reply With Quote
Old Jul 01, 2010, 09:52 PM
Registered User
Joined Jan 2009
461 Posts
Quote:
Originally Posted by airbotix View Post
Efficiency from the switching techniques and mainly applies at half throttle regime.
There is little difference from a conventional ESC at full throttle.
The commutation software is a little more tricky.

The regeneration is only the short term recovery from prop inertia.
Banking probably absorbs this energy into the opposite motor more so than putting it back into the LiPo.
Certainly bench testing with a PSU does cause the main rail to overshoot
- a larger capacitor is necessary to control this.
With LiPo there is no issue.

The ESC is nominally set up for linear PWM duty cycle versus input command.
It is possible (and a reasonable thing to do) to linearise this to thrust for the Motor/Prop combo.
We now prefer to this in the flight controller algorithms.

Lead shaping sounds a very good idea - we haven't tried it.
I would actually use a gyro differential term and let the flight controller look after it.
Currently we use PI on angular velocity which works well most of the time.
Heavy props like APC14" may be better treated with a D term.
Care is needed since noise on gyro samples will inflict more jitter.

Flight control questions best put under the other thread.

Steve
Steve,

Both thrust linearization and lead shaping can be done in the controller instead of the ESC. That's what I'm planning to do since I will be using off the shelf ESCs (at least at first).

- Roy
RoyLB is offline Find More Posts by RoyLB
Reply With Quote
Old Aug 06, 2010, 11:15 AM
Registered User
Joined Jan 2009
461 Posts
Quote:
Originally Posted by RoyLB View Post
Steve,

Both thrust linearization and lead shaping can be done in the controller instead of the ESC. That's what I'm planning to do since I will be using off the shelf ESCs (at least at first).

- Roy
Here's the impact of lead shaping:

http://www.rcgroups.com/forums/showp...&postcount=325
RoyLB is offline Find More Posts by RoyLB
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Jeti brushless controller... HELP!!! Dan Chase Power Systems 6 Oct 14, 2003 11:16 AM
Jeti 70 amp brushless controller on 18 cells-- no problem Andreas Power Systems 12 Sep 25, 2001 09:58 AM
Aveox L160 Brushless Controller Steven Horney Aircraft - Electric - Airplanes (FS/W) 4 Sep 17, 2001 06:13 PM
pic based brushless controller fisher Power Systems 1 May 28, 2001 10:39 AM
Kontronik Brushless controller modes (3 or 4)? ShakyHands Electric Heli Talk 0 May 25, 2001 07:40 PM