rrTV-PHOTO   New HD TV
HOME   rrTV-PHOTO   GALLERIES   MY GALLERY   HELP-FAQ
myHOME PM pmRR MEMBERS 783 ONLINE 21 EVENTS SEARCH REGISTER  START HERE
 
3 pages [ <<    <    ( 1 )     2      3     NEXT    >> ]2707 viewsPOST REPLY
Next D . Fast Lad Performance . Ace Hobby

.
.
Main Discussion > SERVOS: An in depth look - Part 2
 
 
playfairKey Veteran - Location: Rochester, NY -
Continuation from SERVOS: An in depth look - Part 1

In order for a speed vs. torque graph to be really significant, we need to understand what the actual servo loads are during flight.

This is made possible in part by using a data logger to monitor the current draw of a particular servo. The exact servo used isn't important, as long as its current draw vs. output torque is calibrated.

For the field test, I asked my buddy Kyle Stacy to fly his 90 eccpm heli with the logger installed, which uses BLS451 servos on the swash.

The following graphs were created to find the relationship beween current and torque at 6 volts:

The resulting polynomial equation is used to convert current to torque for each reading through a spreadsheet.

The flight data looks similar to that from my R90, where I put the logger on the S9351 collective servo, calibrated for 5.3 volts:

This shows that 60 ounce-inches was easily exceeded during flight in this dedicated position. The eCCPM machine, which monitored the "elevator" servo, peaked at about half of this (the "center" servo on a 120deg eCCPM swash is arguably the hardest worked).

Either way, the graph by itself doesn't tell the whole story, as we have no idea what happened with what input.

To visualize this, I wrote a program to display it in bar-graph form, and synchronized it with the flight video.
The following was the ultimate goal in this "investigation":

In flight Servo Torque Video (43mb wmv file)


The sky is our canvas
04-14-2008 08:10 PM
HOMEPAGE  
 
 
tauscnc
Veteran
Location: IL : www.cuttingedgecnc.c om

Very impressive data! The video version really helps me understand it.... a little I have always asked and wondered about this question...what amount of power is really needed to fly our airplanes and helis. I remember back in the day we would never of dreamed of putting a sub micro (futaba S-133) on an aileron, etc and now a days we put HS-55 on EDF jets, etc.

So looking at the video one could conlude that a 42 oz/in servo could fly that 90 heli???? The avg. oz/in seemed to hover around 20 with a spike toward 35 every now and than.

Am I looking at that correctly?

Thanks for all of your hard work! Great stuff

taus
www.cuttingedgecnc.com
04-14-2008 08:30 PM
HOMEPAGE  
 
 
playfair
Key Veteran
Location: Rochester, NY

Quote 
one could conlude that a 42 oz/in servo could fly that 90 heli?
Not exactly, taus.

If I had to sum up the data from this and the other thread, I would conclude that you never want to load a servo past its halfway to "stall" spec.


The sky is our canvas
04-14-2008 08:44 PM
HOMEPAGE  
 
 
jschenck
Elite Veteran
Location: La Vista, NE.

What I'm seeing so far is a 9351 is a really nice servo under load with a 6V power supply. Be interesting to have an 8717 put through these tests.

Thanks - this is a lot of work and really informative!

...yep...
04-14-2008 10:03 PM
 
 
Morris
Senior Heliman
Location: Hong Kong

Hi playfair, very impressive analysis. Can you help to do the same test on a 9257 (at 4.8V), 9650( 4.8V and 6.0V) and then 9650 with 9257 gear (4.8V and 6.0V)?

I found this modification in the following thread when I found speed is amazing fast (theoretically 0.06 second/60 degree)but not know how we trade off the torque.

[url=http://www.runryder.com/helicopter/t386123p1/]

Many thanks since you have everything set up.
04-14-2008 11:26 PM
 
 
playfair
Key Veteran
Location: Rochester, NY

Morris- if you want to send the servos to me, I'll check them out.

The 8717 is at the top of the list, as soon as I get my buddy to pull it out of his heli!


The sky is our canvas
04-15-2008 01:51 AM
HOMEPAGE  
 
 
Stick Twanger
Senior Heliman
Location: Honeoye Falls, NY

Quote 
If I had to sum up the data from this and the other thread, I would conclude that you never want to load a servo past its halfway to "stall" spec.


Playfair
Is it safe to say that a servo that does well in in a CCPM environment may not do as well in a non-CCPM helicopter?
04-15-2008 01:14 PM
 
 
rob10000
Key Veteran
Location: Western Massachusettes

Interesting data..
When I bought my Concept 60 about 15 years ago, I put on some of the best servos available at the time, the coreless Futaba S9401's with .14/60 degrees and 44oz/in of torque!
04-15-2008 02:04 PM
HOMEPAGE  
 
 
Topher
Veteran
Location: Grosse Ile, Michigan

I havent watched your video since I really cant while I am at work but instead of correlating video chronologically to data why not just record the servos pulse train? If you already measuring current you logger should be able to easily be configured to record a PWM. BTW, nice job.

Get a life? How the hell do you fly a life?
04-15-2008 02:13 PM
HOMEPAGE  
 
 
Chuckie
Key Veteran
Location: Crofton Maryland, MHA member

playfair,

Another excellent post. I'm still trying to soak all this in especially the video. Total Amps looks interesting, a couple of times the amps shoot up but the elevator torque stays low. The rudder must be drawing a good amount of power.

Charles

Please stand by for faster service!
04-15-2008 05:51 PM
 
 
bcm
Senior Heliman
Location: Tuolumne,CA

actually chuckie's comment was right in line with a question that I had. Is the amperage graph shown on the video representing single servo amperage, or total system amperage?
04-15-2008 07:24 PM
HOMEPAGE  
 
 
Topher
Veteran
Location: Grosse Ile, Michigan

Given the amount of current thats being drawn, I would think its only a single servo.

Get a life? How the hell do you fly a life?
04-15-2008 08:05 PM
HOMEPAGE  
 
 
bcm
Senior Heliman
Location: Tuolumne,CA

that was my initial thought, but if the update speed of the data logger was slower, I was thinking that it may show more of an "average" load.
04-15-2008 08:31 PM
HOMEPAGE  
 
 
AirWolfRC
rrProfessor
Location: 42½ N, 83½ W

You calculate the torque from the current for the purpose of the flight video and therefore the torque and current should be in lock-step but they're not.

Or am I missing something ?

As far as the overlay in the video goes . . . nice job . . . how did you do it ?

Also, what did you use to derive the curve fit polynomial ?
04-15-2008 08:42 PM
HOMEPAGE  
 
 
playfair
Key Veteran
Location: Rochester, NY

You're right, AirWolf- servo current and torque would follow each other. The right side green bar is the TOTAL current coming out of the battery. It's a gimme with the Eagle Tree data logger, along with battery voltage. I included it to show what's being drawn from a 7.4v lipo through the linear regulator (I believe he uses a SG Reactor). When the total current spikes up, it's either the tail servo, or either of the other two cyclic servos (used heavier during aileron movements).

The data plots were done in Excel, which allows you to fit a curve and show its equation.

Topher- The logger rate was set to 8 samples per second, with a max of 10. Still way to slow to see the servo pulse train.


The sky is our canvas
04-15-2008 08:51 PM
HOMEPAGE  
 
 
GimbalFan
Elite Veteran
Location: Copter County, Nv

Good job!

op-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-thwõp-t
04-15-2008 08:55 PM
 
 
Bad Karma
Veteran
Location: UK

That polynomial for everyday use could be simplified by ignoring the X^3 and X^2 terms since their coefficients are so small they would be near enough zero in all cases.

I would be interested to see a comparison of the 8900G aswell if you can?
04-15-2008 09:14 PM
 
 
JKos
Elite Veteran
Location: City of California in the state of Maryland

> ignoring the X^3 and X^2 terms

Given the magnitude of x being fed into those terms, the small coefficients don't make them insignificant. The curve is obviously not straight.

- John

MSH Protos
04-15-2008 09:17 PM
 
 
steve9534
Key Veteran
Location: yakima, wa.

Beautiful

I had stated several times here on RR that I thought a graph of current consumption vs torque output could be used to obtain useful info regarding the actual in flight servo loads, but I didn't have the electronics skills to put it all together. I've been hoping that someone would pick up on the idea and make it work. You're other thread on servo torque values and centering capabilities was equally as informative. Thanks much for the info. steve.
04-15-2008 11:18 PM
 
 
Bad Karma
Veteran
Location: UK

1E-9 and 2E-5 is tiny, most times you would round that to zero, especially the E-9, its straight enough for most uses.
04-15-2008 11:58 PM
 
 
3 pages [ <<    <    ( 1 )     2      3     NEXT    >> ]2707 viewsPOST REPLY
Esprit Model . Thunder Power RC . Mikado Modellhubschrauber

.
.
Main Discussion > SERVOS: An in depth look - Part 2
 PRINT TOPIC Advertisers 

Subscribe to This Topic

Friday, December 5 - 1:11 am - Copyright © 2000 - 2008 runryder.com | email | link to rr | runryder needs cookie