BE_BXL

I got hold on a 1/3.5 wilga kit. I will be using it as a tug plane. As I need some strong servos in it, I considered the F9206 (futaba), C4821 or C 4421 (both Graupner), HS945MG (hitec) or a multiplex power. All these have strength around 9Ncm.

What I do not know is how much power they use. Is speed related to power? As this is a tug, servo speed will not be an issue. The less power the servo uses, the longer the batteries last, and there would be no need to use an opto isolator (DPSI) I think.

Lynx

A standard servo will use roughly 100ma idle, (no load) and between 1 and 2 amps in operation, probably averaging around the half amp area (these are just approximations you should really test YOUR servo's with a multimeter (very easy to do)) Keep in mind if you use a properly leveraged control arm you don't need a servo that has a lot of torque, leverage can do most of the work for you.

cliftonbel

A four servo setup would draw 400 ma total with idle current being 100 ma per servo.I dont think so.Maybe 15ma per servo.

dirtybird

My Feedback: (5)

I have recently been doing some servo load testing. My findings are as cliftonbel states, the idle current of the standard Futaba servo is less than 15ma. Perhaps you are using digital servos?
Of the servos BE_BXL mentioned the only one I have tested is the Futaba 9206. This is the best servo I have tested but it produces about 30% less torque than Futaba advertises. Its still a strong servo and very economical on power.

BE_BXL

The speed of the different servos:

0,23s/40° Multiplex Power
0,14s/40° Graupner C4421
0,14s/45° futaba 9206
0,16s/60° Hitec 945mg
So If speed would be an indication of the power a servo uses, the multiplex would be the most economical on power. And the hitec would use the most power.
Is this a correct deduction?
Can anyone tell me how I should proceed to check the power with a multimeter?

Geistware

My Feedback: (16)

A standard servo draws about 10ma.
Stalled it draws about 200-250ma.
Normal flying is about 75-150ma.
Why do you ask.

Rodney

My finding are the same as Geistwares accept for the stalled current. I have had some servos that will draw close to 2 amperes if stalled at full load but even those idle at about 10 milliamperes. Digital servos may draw more but I have not checked any of them.

dirtybird

My Feedback: (5)

Slower speed indicates a higher gear ratio. Given the same motor the higher gear ratio would require less power to move the servo with a constant load depending on the efficiency of the gears. Just the opposite of your assumption.
If your multimeter is a true RMS unit just measure the current thru the motor and the voltage across it and multiply them together. Otherwise you will not be able to do it. If it is a true RMS multimeter it will say so on it somewhere.

Lynx

RMS is term for measuring AC voltage, not DC currents...

BE_BXL

The aim is to get a 3.20m wilga equipped. The wilga will be used as a tug plane. On towing events, 50 tows a day is a normal number. So powermanagement is an important issue. I do like to use a DPSI or easy-patch. I will be using two receivers, so all the current will need to go trough these receivers. For this reason I do not want to use digitals - they consume too much. So I am looking for the servo which is most economical on power.

dirtybird

My Feedback: (5)

Root Means Square (RMS) is the effective value of a waveform. It can be used to decribe any waveform. It requires a special meter to measure it. The oldest is the iron vane meter. You don't see much of them anymore but there are more modern versions of it. To get an accurate measure of the power consumed you need to get the RMS value. Sinewave AC current is only a special case where RMS value is used. When you measure AC current or voltage on a meter the meter has a special conversion built in to give you the RMS value of a sine wave and it only works for a sine wave. The true RMS meter will give you the effective (RMS) value of any waveform.

Geistware

My Feedback: (16)

I have never seen that much current on a stalled standard servo.
At 4.8 volts on a 42ozin servo, 2 amps is really high

mr.rc-cam

I just grabbed four servos and measured them using a Fluke 8040A TRUE RMS DMM. Here are the results at 4.8V operation:

JR 517, No load , moderate speed end-to-end transit: 150mA to 320mA was observed. Mostly seen was the low to low 200's.
JR 517, Full Stall: 645mA

Futaba S3003, No load , moderate speed end-to-end transit: 160mA to 330mA was observed. Mostly seen was the low 200's.
Futaba S3003, Full Stall: 660mA

Hitec HS605, No load , moderate speed end-to-end transit: 170mA to 1005mA was observed. Mostly seen was the mid 400's.
Hitec HS605, Near Stall: 970mA

Futaba S9404, No load , moderate speed end-to-end transit: 210mA to 440mA was observed. Mostly seen was the mid 200's.
Futaba S9404, Full Stall: 1.2A


I have seen some high torque and digital servos experience 2A peak currents. Some pack a punch, and can strip gears if you stall them. So do not test them for stall currents!

RC-CAM

rajul

My Feedback: (58)

Rc-cam, interesting results. Can you share the electrical setup you used for the test ? How did you simulate the stall ? Thanks

Rodney

In a typical servo, the only thing limiting the current at stalled conditions is the resistance of the wires to the servo plus the saturation resistance of the power transistor plus the dc resistance of the motor windings (usually less than one ohm stalled). You will get a drop of about .7 volts accross the power transistor.

mr.rc-cam

Fluke 8040A DMM, 3A Bench Supply, RC-CAM Servo Exerciser. Stall was simulated by brute force -- I held the servo arm while commanding servo movement. I could not stop the arm on the HS605 because the round wheel was too slippery. Hence, the "Near Stall" notation on that one (I got close to holding it still).

Some servo designs use MOSFET bridge amps (they tend to be used on modern Hi-Torque designs). The voltage drop is considerable less on them. These are usually the servos that can see huge stall currents.

RC-CAM

dirtybird

My Feedback: (5)

Note that the Futaba servos that use the MOSFET output transistors limit the applied voltage to 4.8V. Besides broken gears you can destroy the servo by overheating the armeture.

rmh

For the" lectricly challanged "--
The Whattmeter used by lots of guys who are into electric powered models, is a real eye opener.
basically it does this:
plug battery into input --attach rx and servo (or any load ) into load.
The Whattmeter will read and display voltage , amperage and watts -all at the same time.
I got mine to setup my parkflyers and it is absolutely one of the handiest tools in my shop.
One of the surprises was the fact that all batteries are not created equal.
Some batts simply fold up under sudden loads and effectively drop the power to the RX below operating threshold.
On the electric models -it shows why the controllers quit if too large a prop is used .
a sudden application of power will drop some types of cells to less than half their rated voltages.

Lynx

NiCad's are supperior for sourcing large loads like that . NiMH's will fold, so will Lipoly's only faster. Suffice to say, if you have a lot of shock loads to worry about NiCad's are best. If you have a steadier current base Nimh's are better. If you have very low peak current draws but sustained requirements Lipoly's are best.

Lynx

A good general rule is 10 5 2 as far as current sourcing of Nicad/Nimh/lipoly

freestyle

"Slower speed indicates a higher gear ratio. Given the same motor the higher gear ratio would require less power to move the servo with a constant load depending on the efficiency of the gears. Just the opposite of your assumption."

But even that rule-of-thumb assumes that all motors are the same, and they aren't, so while what you say might be true in theory, it's not worth anything in practice (that sounds harsh, but really I mean no offense). You could try multiplying speed and torque to get a figure that means nothing by itself but that would be semi-useful in comparing servos to guess at the power consumption. However, given variations in motor efficiency and control circuit design (analog vs digital, different transistor types, etc) even that would be of limited usefulness.

Lynx's post has me wondering - has anyone tried running large capacitors in parallel with NiMH and LiPoly batteries to reduce the voltage drop during brief current peaks?

Lynx

Even 'brief' current peaks in RC applications are really too high to practically use capacitors to buffer. You run into serveral problems. One being size, the number of caps needed take up a lot of space and decrease performance more by the added weight than they increase it through smoothing voltage dips. Also using supercaps or other high value low voltage capacitors and you end up with a device that has a fairly large equivilant series resistance, so it's constantly wasting at least some power. Further yet in RC applications the useable voltage of a capacitor is only about 1 volt of it's total capacity, after bout that much and you're essentially drawing current directly from the battery. An easier and less complex (IE failure prone) method of decreasing voltage drop and increasing current draw ability is simply using smallter packs (1000mah instead of 2000mah) and simply putting two of them in parallel. It's slightly heavier than an equivilant 2000mah pack but because they're run in parrallel the current drawing capacity is doubled, and logically speaking of course, the equivilant voltage drop is halved (or more) I'm surprised more hardcore racers don't do this with their ground vehicles because of the high loads they put on their motors, it's not so usefull for aircraft because the torque applied to a prop is virtually constant compared to a wheeled vehicle.

Flypaper 2

BE BXL:
You might consider booster tabs on the control surfaces. They can lower the servo load, current, by as much as 80%.

mr.rc-cam

The concept has been bantered about by the eflight modelers. Here is one such discussion:
http://www.rcgroups.com/forums/showt...8&page=1&pp=15

RC-CAM

BE_BXL

Flypaper 2,
What are booster tabs?