RunRyder RC
 21  Topic Subscribe
WATCH
 5 pages [ <<    <     2      3     ( 4 )     5     NEXT    >> ] 7998 views
Scorpion Power Scorpion Power
HelicopterMain Discussion › Actual Servo Speed Under Load
09-14-2007 09:50 PM  10 years agoPost 61
AceBird

rrElite Veteran

Utica, NY USA

My Posts: All  Forum  Topic

It's confusing, because to look at the specs, you would think that it can operate at the rated speed at the rated torque. Life would be easy then picking out a servo!
What is so hard about picking out a servo? Pick what someone else uses that worked for them. On a 120 degree swash do a climb out and a roll. If it goes clean then you have enough torque. If it goes fast enough then you have enough speed. If it last long enough then you have a good choice.

The system of no load speed and stall torque is the only fair system of comparison for servos. It is easily to replicate and prove. Actual run torque and speed values would be so subjective. It's so simple, if you don't have enough torque go higher if you don't have enough speed go faster, if it doesn't last long enough go for the higher torque rating.

Ace
What could be more fun?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-14-2007 10:07 PM  10 years agoPost 62
Rockohaulic

rrElite Veteran

Canyon Country, CA,​USA, 3rd Rock from​the Sun

My Posts: All  Forum  Topic

The system of no load speed and stall torque is the only fair system of comparison for servos.
I disagree. They should have some idea or some ball-park figure for what actual loads we see on our 50 size helicopters or 90 size helicopters. Then they should test the servos at that load. Then there wouldn't be any guessing or assumptions on our part.

Would you want to pick out a sports car based on how fast it would go if it didn't have any aero drag or any weight? Sure would put the car magazines out of a job if we did!

Or would you want to marry your blind date because she has a really really nice personality??? (OK - that one's a stretch, but I get the shudders just thinking about it!)

Saturday morning I flew my helicopter in my pajamas
How it got in my pajamas I'll never know

PM  EMAIL  GALLERY  Attn:RR  Quote
09-15-2007 01:07 PM  10 years agoPost 63
AceBird

rrElite Veteran

Utica, NY USA

My Posts: All  Forum  Topic

Or would you want to marry your blind date
wisdom comes with age ... any person you marry is a blind date that you can't walk away from.
Then there wouldn't be any guessing or assumptions on our part.
Where is the guessing? There are thousands of people with experience on this forum. Ask them or search the threads that have already answered your question. "What servo do I use?"

Ace
What could be more fun?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-15-2007 02:12 PM  10 years agoPost 64
FOX222

rrApprentice

Atlanta, Ga.​(Actually Buford)

My Posts: All  Forum  Topic

Follow

PM  EMAIL  GALLERY  Attn:RR  Quote
09-18-2007 08:06 AM  10 years agoPost 65
ChristianM

rrVeteran

Oslo, Norway

My Posts: All  Forum  Topic

Hi everyone
The formula for servo speed under load = S/(1-L/T), where

S = Rated Speed at Zero Load
T = Rated Servo Torque
L = Assumed Actual Load
As Rockohaulic mentioned I derived the equation and posted it in a earlier thread. This has been an interesting discussion here and am sorry I could not join in earlier. I just want comment on the assumptions behind the formula.

First: The only published data we have for servo performance is the no load speed and the stall torque so I made the assumption that there is a linear relation ship between these two points. While this is not accurate it is still a good enough approximation to perform some meaningful comparisons. Generally the curve between these two points will be convex, i.e. the line curves away from the center of the graph. This means that the actual servo speed under load will be somewhat higher than what the equation predicts.

Second: Since we are used to seeing the servo speed quoted as seconds per 60 deg I have derived the formula to give the result in this format. However, seconds per 60 deg is the inverse of speed and thus is not proportional to speed. I have included a graph below to show the relationship between these two.

Below I have included a graph with the "servo speed" (in seconds per 60 deg) under load based on the formula, for a few popular servos. As you can see it is not a linear relationship when displayed in this format.

As JKos mentioned earlier in this thread you can use this little inexpensive gadget from Dimension Engineering to determine the max and average current draw from one or more servos. I have one and it really provides a lot of useful info. If you have measured the max current load during flight then you can do a bench test afterwards by adding load to the servo until the current draw matches. Then you can use the equation to predict the slowest speed of the servo in flight. This will give you some good insight and allow you to compare servos against each other but at the end of the day it is how the heli feels in the air that is important

I hope that you find it useful.

Christian

Burn fuel, be happy

PM  EMAIL  GALLERY  Attn:RR  Quote
09-18-2007 05:48 PM  10 years agoPost 66
AceBird

rrElite Veteran

Utica, NY USA

My Posts: All  Forum  Topic

but at the end of the day it is how the heli feels in the air that is important
And how long the servo last under those operating conditions …

Ace
What could be more fun?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-18-2007 06:28 PM  10 years agoPost 67
AirWolfRC

rrProfessor

42½ N, 83½ W

My Posts: All  Forum  Topic

Generally the curve between these two points will be convex
I'm curious what you base that comment on. I don't know what the answer is myself but I imagine there is some efficiency cuve for the servo motor that will come into play but not available for inspection.

Whether a curve is linear, convex or concave will also be determined by how it is graphed. Plotting a linear axis on speed is not the same as a linear plotting of time of travel. Charts can also be shown with log or linear scales like linear-linear, log-linear, log-log for the X and Y axis respectively.

In the end, so long as you plot data using the same method, good comparisons are assured.

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
09-18-2007 09:48 PM  10 years agoPost 68
AceBird

rrElite Veteran

Utica, NY USA

My Posts: All  Forum  Topic

Two end points only work if the function is a straight line. If they are not then there is no comparison unless you are into comparing assumptions.

Ace
What could be more fun?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-18-2007 11:54 PM  10 years agoPost 69
Rockohaulic

rrElite Veteran

Canyon Country, CA,​USA, 3rd Rock from​the Sun

My Posts: All  Forum  Topic

And my 2 favorite cyclic servos.

As you can see, the Hitec 6975 is faster than the Futaba 9351 up to approx 80 oz-in of torque.

Saturday morning I flew my helicopter in my pajamas
How it got in my pajamas I'll never know

PM  EMAIL  GALLERY  Attn:RR  Quote
09-19-2007 12:17 AM  10 years agoPost 70
AirWolfRC

rrProfessor

42½ N, 83½ W

My Posts: All  Forum  Topic

Two end points only work if the function is a straight line. If they are not then there is no comparison unless you are into comparing assumptions.
A)Not true. Think a bout that one.
B)The assumption is that most all servos exhibit similar characteristic curves with only the scale and range being different. I consider that a safe assumption.

If you disagree, please show some evidence.

I and the rest of us could stand to be enlightened.

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
09-19-2007 03:44 PM  10 years agoPost 71
ChristianM

rrVeteran

Oslo, Norway

My Posts: All  Forum  Topic

Airwolf
I'm curious what you base that comment on. I don't know what the answer is myself but I imagine there is some efficiency curve for the servo motor that will come into play but not available for inspection.
I have not seen the speed vs. torque for these servos but the ones I have seen of other electric motors have been convex. These curves are established based on tests so they already include any effects of efficiency (or lack thereof). These plots are all done with linear scales on both axis starting at zero.

Christian

PM  EMAIL  GALLERY  Attn:RR  Quote
09-19-2007 08:02 PM  10 years agoPost 72
SSN Pru

rrElite Veteran

Taxachusetts

My Posts: All  Forum  Topic

extrapolating those numbers
No extrapolation. It's interpolation and by definition much more accurate than extrapolation. They are interpolating data, assuming a linear relationship, based up on 0 torque at the rated speed and 0 speed at the stall torque.

If you apply the linear assumption to all servos then you have a valid comparison, even if the true numbers being shown are not accurate.

Stupidity can be cured. Ignorance is for life!

PM  EMAIL  GALLERY  Attn:RR  Quote
09-19-2007 09:06 PM  10 years agoPost 73
kangarooster

rrApprentice

Orlando Fl-USA

My Posts: All  Forum  Topic

Is there a simple way to determine servo torque and speed?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-19-2007 09:21 PM  10 years agoPost 74
AirWolfRC

rrProfessor

42½ N, 83½ W

My Posts: All  Forum  Topic

If a set of curves have the same characteristics, it will not matter if the plots are linear or log, if the curves are convex, linear or concave or anything else. The same differences between servos will be evident.

That is my point.

And as far as efficiency goes,
If we are given only stall torque and no-load speed, we know nothing about efficiency.

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
09-19-2007 10:49 PM  10 years agoPost 75
AceBird

rrElite Veteran

Utica, NY USA

My Posts: All  Forum  Topic

B)The assumption is that most all servos exhibit similar characteristic curves with only the scale and range being different. I consider that a safe assumption.
Servos that have the exact same motor and drive will exhibit similar characteristics but different motor and drives could be quite different. This concrete data that you are looking for doesn’t exist in the hobby field but in the industrial field there are thousands of servo / stepper motor and drives combinations that will give a multitude of characteristics. The programming alone will drastically change the way a motor behaves. Oversimplifying by drawing straight lines between two endpoints on a relationship that is not a straight line is an absolute no, no in my book. You can do what ever you like but I would not put faith in that approach.

Ace
What could be more fun?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-19-2007 11:00 PM  10 years agoPost 76
Rockohaulic

rrElite Veteran

Canyon Country, CA,​USA, 3rd Rock from​the Sun

My Posts: All  Forum  Topic

The 3 charts I see above look kinda curvy.

Saturday morning I flew my helicopter in my pajamas
How it got in my pajamas I'll never know

PM  EMAIL  GALLERY  Attn:RR  Quote
09-21-2007 04:43 AM  10 years agoPost 77
AirWolfRC

rrProfessor

42½ N, 83½ W

My Posts: All  Forum  Topic

Well, AceBird, the way I see it, you can keep putting down soem one else's way of tryint to come to come conclusions or you can try to help clarify the problem and ambiguities.

I'm not seeing where you're helping the quandry.

The charastic responce curves of motors used in the typical servo WILL be typical among themselves and a straight line responce curve will be meaningfull.

Now, do you want to help the clarification or do you want to keep perpetuating quandry ?

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
09-21-2007 01:38 PM  10 years agoPost 78
AceBird

rrElite Veteran

Utica, NY USA

My Posts: All  Forum  Topic

a straight line responce curve will be meaningfull.
A straight line response curve is no more meaningful then taking the published speed / torque values and making a decision from them. What I have said and now others have said is "history" is the best validation for what to use. And until you get some history on any new servo produced you are essentially shooting craps.

If you feel good about drawing straight lines then go for it, but scientific it is not.

Ace
What could be more fun?

PM  EMAIL  GALLERY  Attn:RR  Quote
09-21-2007 02:11 PM  10 years agoPost 79
SSN Pru

rrElite Veteran

Taxachusetts

My Posts: All  Forum  Topic

You can do what ever you like but I would not put faith in that approach.
Scientific and Engineering it isnt, thats for sure.

But, BIG BUT...

We are not choosing the servo that will control the ailerons on an F-22.

Our rough approximations for servo performance are as good as we are going to bet without expensive testing and evaluation.

History is a good indicator of performance but of what kind of performance? History certainly wont indicate to you or anyone else what kind of servo speed at "X" loading you are going to have. History will only give us a pretty good indicator of servo life expectancy and even then it is HIGHLY dependant on the flying style and airframe the servo is used in.

As I said before and as AirwolfRC said, using a straight line approximation is perfectly suitable for our needs. If you used a straight line approximation for one servo and actually had good speed/torque data as measured for another servo and then you attempted to compare them then that is not gonna work.

Stupidity can be cured. Ignorance is for life!

PM  EMAIL  GALLERY  Attn:RR  Quote
09-21-2007 05:19 PM  10 years agoPost 80
AirWolfRC

rrProfessor

42½ N, 83½ W

My Posts: All  Forum  Topic

AceBird, I really don't see what you have against using what information is available, imprecise though it may be, but containing pertinent information just the same.

And then you admit that you have nothing to fill the void.

Is this what you call helpfull ? ?

Is this your idea of helping others come to some conclusions ? ?

Using what information is available to make choices and decisions is called engineering. Using "history" is just following the the rest of the flock. If everybody followed the rest of the flock, we'd still have ships with steam engines. . . . Oh, I forgot, steam was the answer to "historic" sail.

The point is that people will go ahead and try things even if they don't have all the precise information.

I have been quantifying this kind of information for about 40 years, successfully. And, yes, I am VERY familiar with motor controls of all kinds, phase controled, pulse width controled, voltage controled, DC, open loop, closed loop and others - - - and their performance curves. - - - - But I suppose you think you know better.

PM  EMAIL  HOMEPAGE  GALLERY  Attn:RR  Quote
WATCH
 5 pages [ <<    <     2      3     ( 4 )     5     NEXT    >> ] 7998 views
Scorpion Power Scorpion Power
HelicopterMain Discussion › Actual Servo Speed Under Load
 Print TOPIC  Make Suggestion 

 21  Topic Subscribe

Saturday, November 18 - 4:17 am - Copyright © 2000-2017 RunRyder   EMAILEnable Cookies

Login Here
 New Subscriptions 
 Buddies Online