HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Sep 27, 2011, 09:35 AM
Al Ducharme
photronix's Avatar
Orlando, FL
Joined Jan 2005
734 Posts
Well let me ask this. Considering the transfer function curve there is no response past about 50hz. Since many of us are using turnigy do you believe changing pwm faster than 100hz has any effect on flight? Because according to the data it wouldn't be possible. I am really interested to know your thoughts.
photronix is offline Find More Posts by photronix
Reply With Quote
Sign up now
to remove ads between posts
Old Sep 27, 2011, 10:59 AM
Tri-Quad-Hexa-Octo-copters!!
United States, TX, San Antonio
Joined Feb 2007
14,408 Posts
PT,
Sure. The first thing to recognize is that sometimes engineering data does not reflect operational reality, especially if it is mis-applied Also smaller multicopters under say about 30cm in diameter require faster updates of 400Hz+ to perform well. Conversely larger multicopter diameter frames can perform OK/well with slower ~100Hz-200Hz updates
I suppose if one compares the ~50Hz-100Hz poor multicopter performing Castle ESCs with the ~200Hz-300Hz non-modified Turnigy Plush series and re-brands, that would tell the tale. Based on most folks experiences, this has already been proved.

KapteinKuk ESC update frequency testing: http://www.rcgroups.com/forums/showt...date+frequency

Old Man Mike's Update and motor testing: http://www.rcgroups.com/forums/showt...e#post11657593

Here is a simple video showing the positive effects of increasing ESC PWM update frequency by Lesnewell http://www.rcgroups.com/forums/showp...&postcount=156:
Upgrading the firmware in a Hobbyking ESC (1 min 46 sec)


Cheers,
Jim
Quadrocopter and Tricopter Mega Link Index


Quote:
Originally Posted by photronix View Post
Well let me ask this. Considering the transfer function curve there is no response past about 50hz. Since many of us are using turnigy do you believe changing pwm faster than 100hz has any effect on flight? Because according to the data it wouldn't be possible. I am really interested to know your thoughts.
jesolins is offline Find More Posts by jesolins
Last edited by jesolins; Sep 27, 2011 at 11:18 AM.
Reply With Quote
Old Sep 27, 2011, 11:12 AM
Registered User
Hoofddorp, Netherlands
Joined Sep 2006
137 Posts
http://www.howstuffworks.com/analog-digital3.htm

substitute sound for ESC THR output, with ESC update rate = sampling rate.
Long story short, beside inicreased response it increases the throttle resolution.

Groeten,

Peter
Djago is offline Find More Posts by Djago
Reply With Quote
Old Sep 27, 2011, 11:34 AM
Al Ducharme
photronix's Avatar
Orlando, FL
Joined Jan 2005
734 Posts
Quote:
Originally Posted by Djago View Post
http://www.howstuffworks.com/analog-digital3.htm

substitute sound for ESC THR output, with ESC update rate = sampling rate.
Long story short, beside inicreased response it increases the throttle resolution.

Groeten,

Peter
Yes the aliasing is caused by the sampling rate used in the Castle.
photronix is offline Find More Posts by photronix
Reply With Quote
Old Sep 27, 2011, 11:41 AM
Registered User
VA
Joined Nov 2009
2,628 Posts
Quote:
Originally Posted by Flyboone View Post
Well for starters the eRC has more programming options and above the 35A ESC they all come have switch mode BEC's with adjustable output voltage. Not saying these are way better or the end all be all of ESC's, just that they are proven to work well at 400Hz, have a good price, are available in the US readily, and are backed up with our support.

Jason
Has anyone run your eRC 120A opto esc on 6s pulling a constant 100A? I had looked at them but questioned the update rate. I have had plenty of trouble with heatsinks on the big ones.
pug398 is offline Find More Posts by pug398
Reply With Quote
Old Sep 27, 2011, 11:49 AM
Al Ducharme
photronix's Avatar
Orlando, FL
Joined Jan 2005
734 Posts
Quote:
Originally Posted by jesolins View Post
PT,
Sure. The first thing to recognize is that sometimes engineering data does not reflect operational reality, especially if it is mis-applied Also smaller multicopters under say about 30cm in diameter require faster updates of 400Hz+ to perform well. Conversely larger multicopter diameter frames can perform OK/well with slower ~100Hz-200Hz updates
I suppose if one compares the ~50Hz-100Hz poor multicopter performing Castle ESCs with the ~200Hz-300Hz non-modified Turnigy Plush series and re-brands, that would tell the tale. Based on most folks experiences, this has already been proved.
It scares me when people say this kind of thing "sometimes engineering data does not reflect operational reality" You don't work at NASA do you.

Jesolins can you expand on "mis-applied".

Now back to my original question. If most people are using Turnigy with good results on most flight controllers, and considering the fact that the motor does not change speed above 50Hz (which would be a 100Hz throttle input), then why is it assumed that you have to have 200Hz, 400Hz, or 1000Hz.

I would really like your experienced thoughts on this past "it feels better" Remember that you can't simply change PWM update rate and see if the craft flies better. There are many other variables.

Oh and I have read the other threads on the subject. Honestly, the engineering figure-of-merit used for any system is Transfer Function. You can sweep and step and wobble all you want but in the end Sine in Sine out is what you have to measure.
photronix is offline Find More Posts by photronix
Reply With Quote
Old Sep 27, 2011, 12:21 PM
Tri-Quad-Hexa-Octo-copters!!
United States, TX, San Antonio
Joined Feb 2007
14,408 Posts
PT,
No I have never directly worked for NASA, only USAF Flight test Yep I certainly would not have wanted to be the Aux rocket engineer at NASA...or the engineer that designed the pitot tube system on the B2 bomber, so you should be always be a bit scared as an engineer to be at your best...

Mis-applied refers to the lack of transfer of operational performance between the data you are posing and the way multicopters that use higher ESC update rates perform. Maybe I am preaching to the choir but do note that there are several frequencies used in the path from correctional inputs from the properly calibrated and PID applied gyros/accelerometers to FCB's, to ESCs, i.e. code loops, FCB PWM output frequency, ESC update frequency, Motor timing, ESC processor code loop timing and steps/smoothing and other "enhancement coding", all play a big part in the "real" operational performance off the multicopter.

This all applies to the fact that ESCs designed for multicopter use will ultimately be the best performers by at least a small margin in stability as the top multicopters like the Mikrokopter and X-BL have demonstrated. That said many folks are very satisfied with an extra oscillation or two in windy conditions and happy not to pay the premium for that extra performance

Again, we engineers should not ignore operational reality despite how we "read" the data, especially when they conflict.

Cheers,
Jim
Quadrocopter and Tricopter Mega Link Index




Quote:
Originally Posted by photronix View Post
It scares me when people say this kind of thing "sometimes engineering data does not reflect operational reality" You don't work at NASA do you.

Jesolins can you expand on "mis-applied".

Now back to my original question. If most people are using Turnigy with good results on most flight controllers, and considering the fact that the motor does not change speed above 50Hz (which would be a 100Hz throttle input), then why is it assumed that you have to have 200Hz, 400Hz, or 1000Hz.

I would really like your experienced thoughts on this past "it feels better" Remember that you can't simply change PWM update rate and see if the craft flies better. There are many other variables.

Oh and I have read the other threads on the subject. Honestly, the engineering figure-of-merit used for any system is Transfer Function. You can sweep and step and wobble all you want but in the end Sine in Sine out is what you have to measure.
jesolins is offline Find More Posts by jesolins
Reply With Quote
Old Sep 27, 2011, 12:31 PM
Al Ducharme
photronix's Avatar
Orlando, FL
Joined Jan 2005
734 Posts
Well said.

Now back to the question. Any idea why Turnigy work well? They are essentially a low pass filter. It's a bottle neck that people seem to ignore and I can't figure out why.

Thanx for the info.
photronix is offline Find More Posts by photronix
Reply With Quote
Old Sep 27, 2011, 12:55 PM
Registered User
Joined Feb 2006
725 Posts
I donīt understand many of the parameters and therms used on this threat or some of the links but... if the question is to get as high Hz value as possible from a ESC, some heads in Castle Creations co. must roll away. The can update the firmware to run at 100 hz and with the same effort in programing hours to do it for 200 hz or above. The will do it to run at 100 hz only because they canīt see any difference in the final airscrew reaction when using values above 100 hz and donīt want to waste energy in heat form if there is no advantage. That is so far what I was told from this history. CC has lost a piece of the market selling ESCs that are not usable in multirrotors and they will only adapt them to be 1/4 of the chinese speed at x8 times expensive? Are they crazy or their test are really representative and there is no need to increase speed beyond 100 hz but to work on other parameters as well?
elossam is offline Find More Posts by elossam
Reply With Quote
Old Sep 27, 2011, 01:02 PM
RCG Staff Member
Jason Cole's Avatar
United States, TN, Nashville
Joined Oct 2003
5,008 Posts
Quote:
Originally Posted by pug398 View Post
Has anyone run your eRC 120A opto esc on 6s pulling a constant 100A? I had looked at them but questioned the update rate. I have had plenty of trouble with heatsinks on the big ones.
Not that I've seen yet. If you want, I can cut you a deal to try them out. Let me know. I'll be out the rest of this week at E-Week in SC, but will be back in the office next week.

Thanks,
Jason Cole
Jason Cole is offline Find More Posts by Jason Cole
Administrator
Reply With Quote
Old Sep 27, 2011, 01:28 PM
Registered User
Joined Feb 2006
725 Posts
Jesolins. My tricopter is much more stable than the one of the last video (modified pwm) and mine is based on std Futaba gyros with std output and not with the DS output enabled. May be I'm messing something but if my gyros work on std servo pwm and it's 50hz I can not provide faster changes than 50hz to the ESCs so as long as the ESC can follow this changes (under 50 hz) any speed of them above that value has no sense. What have I missed?
elossam is offline Find More Posts by elossam
Reply With Quote
Old Sep 27, 2011, 01:43 PM
Registered User
VA
Joined Nov 2009
2,628 Posts
Quote:
Originally Posted by Flyboone View Post
Not that I've seen yet. If you want, I can cut you a deal to try them out. Let me know. I'll be out the rest of this week at E-Week in SC, but will be back in the office next week.

Thanks,
Jason Cole
I will pm you later then. Thanks.
pug398 is offline Find More Posts by pug398
Reply With Quote
Old Sep 27, 2011, 02:20 PM
Tri-Quad-Hexa-Octo-copters!!
United States, TX, San Antonio
Joined Feb 2007
14,408 Posts
Hi El,
More info please. Do you have a similar indoor video showing stability? What ESCs are you using and what version of Futaba gyros and Tx/Rx are you using? How did you measure your Gyro and ESC PWM rates?
Cheers,
Jim
Quadrocopter and Tricopter Mega Link Index

Quote:
Originally Posted by elossam View Post
Jesolins. My tricopter is much more stable than the one of the last video (modified pwm) and mine is based on std Futaba gyros with std output and not with the DS output enabled. May be I'm messing something but if my gyros work on std servo pwm and it's 50hz I can not provide faster changes than 50hz to the ESCs so as long as the ESC can follow this changes (under 50 hz) any speed of them above that value has no sense. What have I missed?
jesolins is offline Find More Posts by jesolins
Last edited by jesolins; Sep 27, 2011 at 02:25 PM.
Reply With Quote
Old Sep 27, 2011, 02:51 PM
Registered User
United Kingdom, England, Weston-super-Mare
Joined Jul 2010
477 Posts
I believe (and please don't jump all over me if I get this all twisted) that the answer to this conundrum lies not in whether the motor can meaningfully respond to signalling above 100Hz but in the rate at which the FC is calculating new commands. By blasting the ESC with commands at 400Hz (or 1000Hz, or whatever) you ensure that when the ESC/motor does react, it's acting on the very latest command, not one that was issued "eons" ago. In other words, rather than discarding all the intermediate calculations that the FC does in 20ms (which is effectively what you're doing at 50Hz update rate), send as many of them as you can to the ESC and let it "do its best" with up-to-date commands.
jes1111 is offline Find More Posts by jes1111
Reply With Quote
Old Sep 27, 2011, 03:45 PM
Registered User
Joined Feb 2006
725 Posts
Quote:
Originally Posted by jesolins View Post
Hi El,
More info please. Do you have a similar indoor video showing stability? What ESCs are you using and what version of Futaba gyros and Tx/Rx are you using? How did you measure your Gyro and ESC PWM rates?
Cheers,
Jim
Quadrocopter and Tricopter Mega Link Index
Hi Jesolins. this video was taken one year ago, just before I discovered how useful was to set the ATV from the TX at max and reduce the ammount of stick imput on the AP2000i mixer instead of reducing the ATV on the TX. That in therms of linear response to the stick orders that helps alot taming the machine while hovering. Tricoptero last flight. (1 min 38 sec)
Have no indoor videos but I wish to have an opportunity to fly one indoor. No wind breezes to disturb a relaxing flight. The high walls makes the slightly breeze to run rotors everywhere.

TX Futaba FC28 V1 (about 18 years old technology)
RX Futaba FP-R118F (ppm 35 mhz)
Engines BM2410-9 http://www.hobbyking.com/hobbyking/s...idProduct=6549
ESC Turnigy plush 10A http://www.hobbyking.com/hobbyking/s...ontroller.html
Gyro: 1 Futaba GY352 + 1 Futaba GY401 both running in normal pwm output (50 hz) not in DS (270 hz) and set in HH mode. Tail servo was the first piezoelectric gyro from GWS I bought 20 years ago, has no label to see the model, itīs a a simple blue box whithout HH capabilities or even remote gain.
Airscrew Epp 10x4.5
Mixer: Ap2000i

As you can see nothing but old stuffs. As far as I know the standar servo output of those gyros is 50 hz (20m/s) so if the ESC takes care of changes at that speed from the Gyros it will not send to the engine any different information even having a 400 hz comunication between ESC and engine, simply there is no new information to send. Have no idea about the PWM of the ESC but that doesnīt mind me in this case so long as the Gyro canīt provide more info or do it faster that what it was designed for , 20 m/s to work with std servos.

I built this tricopter as itīs only to demonstrate me and others time ago that a tricopter using std gyros and a mixer was perfectly doable and I was satisfied with the results. Please take it in mind when comparing it with recent hi-tech units.
elossam is offline Find More Posts by elossam
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question MotorFly ESC PWM Frequency Setting ZackJones Power Systems 0 Jul 24, 2011 10:47 AM
Question Hobbywing ESC PWM frequency? abenn Power Systems 4 Jul 03, 2011 04:13 PM
Discussion ESC timing and frequency Uno Electric Ducted Fan Jet Talk 13 Jun 08, 2011 02:31 PM
Discussion Sensor update frequency skogsvargen UAV - Unmanned Aerial Vehicles 2 Sep 15, 2006 01:00 PM
Frequency List Update Don Leitz VOA Flyers of West Chester Ohio 5 Mar 15, 2003 03:39 PM