I have tried several different brand of servos and although some were better than others, all have have poor resolution when setup with large control horns for 3D rates. I've tried Airtronics, GWS (very bad servos!) and Hitec. Although with high rates set mechanically at around 30° on max throw things aren't that bad but if I go to 45° or more then each 'step' of the servo is a very noticable amount - as a gues I'd say at least a degree or two or a millimeter at the tip of the elevator (Sig Cap 231ex).

Is this a common problem? Are some servos better than others? I have noticed my 5645 digital on the rudder even with it's large throws isn't as bad but still noticable. It's not killing me since I can barley fly this thing now but eventually I'd like to start trying 3d (harriers, elevators for starters) and I'm not sure that the 30° I have is enough and I am not daring enough to try and fly it with the 3D setup since the elevator is very sensitive and I'm sure that 1mm of travel is quite a bit of elevator to have as a minimum resolution. Anyhow, how do you guys deal with this issue? $150 servos?

in what way are you putting in "1 step" to see this difference.....if it is using the trim tabs on the transmitter then depending on which transmitter you are using it could be up to 6-10 steps per click...i am flying with the JR 8103 and the "trim steps" are programmable....i find it hard to notice each "step" by only going off stick movement...

First of all, make sure you are running your ATV or rates or whatever high enough that you are using most of the throw the servo can provide. Use the linkage geometry to get the actual surface throw where you need it. On my JR I set the throws at least 125%
If you are using pull/pull you will need a bellcrank setup to accomplish this.
Digitals will always be much better resolution, but good coreless non-digitals are good too.
It looks like many affordable digitals are coming our way!
I think digitals will be the only thing available in a few years.

If I am not mistaken most receivers are capable of up to 1024 steps. With this said, the frequency tramsmitted is from 1.5ms to 1.9 ms (512 steps) and 1.5ms to 1.1ms (512 steps). Ann Marie or Mike would be better able to answer this issue, but this is what I have been told and think is true.

dial a BUNCH of expo in so you can get in to the 3D rates and when your done with the manuver turn them off. Im flying profiles right now and i have about -50% expo on the elevator. I fly the whole time on 3D rates. The expo feels like it has better res untill you really get in to the stick than all he** breaks lose. hehehe.
Travel on elevator is atleast 50 deg. hope this helps.

No, I am dialing elevator with the stick. I have the mechanical linkages set so that at full servo deflection I get approximately 30° of movement (this is on my elevator). I can move the elevator stick ever-so-slolwly and see each 'step' of the elevator. Some servos worse than others (GWS servos were probably 3° per step!). This was worthless as there was no 'center' by which to go by. As far as the digital trims, yes, they are set right now to 4-steps but to be honest when I click it one step at a time I can not see movement on all steps. maybe every 3 clicks it will deflect a little. I have tried this on my Airtronics and Futaba radios. Currently, its a Futaba 9C with a R148DF receiver (PPM).

Perhaps I need to use PCM to get the 1024 steps? I have 605BB's installed now and they aren't as bad as some others but if I try to crank them to get 45° or more then the steps become more noticable. Definately not 512-steps per side, more like 40 steps or so.

- Joe

Joe B, I believe that the steps you are seeing are due to what is called dead band by most servo people. and it is definitely related to the price of the servo. It is specified in microseconds.
A good servo might have a dead band of 1 to 2 microseconds whereas a poor servo might have a deadband of 10 or more microseconds.

The deadband is the amount of change in the servo input pulse which will cause an error signal in the servo that will cause it to move to a new position. Most of our systems now are sending to the servo a 1.5 millisecond pulse (approximately) for a neutral position. This increases by 500 microseconds for full throw in one direction or decreases by 500 microseconds for full throw in the other direction. If the deadband of the servo is 10 microseconds the servo will move approximately 50 steps from neutral to full throw. If the deadband is 1 microsecond the servo would move approximately 500 steps from neutral to full throw.

The only manufacturer that I have seen publish figures for deadband is Airtronics. I have an old 1993 Airtronics catalog which calls out a deadband of 1.3 microseconds for their high bucks coreless ball bearing servo and a deadband of 6 microseconds for their standard 3-pole motor no bearing servo.
Believe me there are lots of servos out there which are no where as good as that Airt. standard.

I guess it all boils down to the fact that you very seldom get more than you pay for (of course you can certainly get a whole lot less) Bob C.

I started flying my 1/4 scale Cap with Hitec 605's all around (two on elevator) and a 615MG on rudder. I had the large arms and big throws on high rates. The plane always seemed to 'wander' never truly returning to neutral after high deflection manuevers.

Watching the elevators on the ground, the centering (or lack thereof) was very noticable.

I put Futaba Coreless (9202) on elevator halves and ailerons and it is a different airplane. When I release the sticks it locks right into neutral trim. Also the sensitivity to very small sitck movements is much better. I put one 9202 in and left one 605 on the other elevator half during the changeover. The 605 didn't even respond to very small stick movements around center that the 9202 reacted very quickly to. In the end, the increased response gives the plane a lot more elevator 'feel' on low rates and made landing actually much easier.

Yes the 9202 is a $65 servo vs a $30 Hitec 605. But there is no comparison in this type of application.

You mention $150 servos. Not neccesary in the 1/4 scale range. There are several good coreless high torques in the $60-$90 range that will work fine. I now use both JR and Futaba coreless models with good results. Save the $150 mega torque digitals for the 35% Gasser with the barn door control surfaces ;-).

Yeah Jim,
I've seen almost the exact same thing on an 80" Sukhoi.
I started out with one coreless servo on the elevator. Not enough torque and very spongy feeling. I went to two JR 517's and it felt the same, but with tracking problems in loops.
I put in two Futaba 9304's and it was very acceptable. Now I have two of the low $ Hitec 5645 digitals and the increased torque at low deflections was very noticeable. It felt like I turned my expo off. Unfortunately they don't center as well as the coreless Futaba's. They find a different center from each direction of deflection.
Pull an up elevator manuvuer and it needs a click of down trim afterwards. Vise-Versa for down elevator.

On rudder I have went from 2 517's (not enough torque or centering accuracy) to 2 9304 Futabas (not enough torque) to 2 JR 4721's (closer) to the current one JR 8411 digital with a bellcrank dividing the 125% throw into +/- 45 degrees.
It still feels like it may need just a touch more torque in some knife manuveurs.
I am thinking about trying one of the Hitec 1/4 scale digitals, but I don't know if it will center as well as the JR 8411 does.

I have always like Airtronics equipment and now I see why. Perhaps deadband is the issue, the incriment at which the elevator travels for each step is noticable and I have noticed it was more noticable on different brand servos. I'll crank down the throws (mechanically) even more to get this to a smaller amount but will definately think ahead to other hardware. Does anyone have a listing of servos and their measured deadbands?

- Joe

I just installed 8 HS605BB servos in a PW Extra. Could you tell me exactly what you did to notice the reduced performance. I want to try these out and don't want to go through the trouble of dialing in the plane and then find that I need to change the servos.

Thanks

Has anyone else noticed that the digital servos are MUCH better for centering than standard servos. I won't use anything but digitals in aircraft where I care about response and centering. I guess that makes me a digital convert, huh?

Dale

This is what I did to notice the 'problem' (I say that loosely as the plane seems to fly fine).

Setup throws to 45°± while TX is set at 100% travel. Then setup low rates to a normal amount, for me it was something like 20% of the full 3D rates. Now while moving the stick VERY SLOWLY you can see the elevator move up in small incriments. Think of it as a Rolex vs. a fake. The rolex ticks 5 times for every second whereas the fake probably ticks once per second. This is the best analogy I can think of. Anyhow, with the large elevator of the Cap I can see at the tip each step, not a constant motion. Sure if I yank it I can not see the steps but if you go slowly you'll see it. If you have a 'small' elevator you may not notice it since I can not notice it on my ailerons which are only 1-1/2" or so, but on the elevator which is 8" or so from hinge-line to tip it is noticable.

As far as the digitals, I have the 5645 digital on the rudder which is perhaps 12" from hinge-line to tip and it is much smoother. I guess I need to spend more $$$ to get it any more percise.

Thanks Joe. I guess I have always experienced this but I have always had in-expensive servos. I have a couple of Futaba and JR coreless servos that I may try on another plane to see how it work!

Thanks again

Set up your linkages so that you use full deflection of the servo and less deflection on the surface. Sometimes when you setup a plane for 50-60 degrees of throw your also throwing out the resolution because your getting less movement at the servo in order to move the surface so far. In other words do you want the plane to fly well, center well and be well tamed or do you want 70 degrees of throw and have a plane that you constantly fight to fly level? Most planes really don't need that much throw to 3D so settle for something in the middle and enjoy it.

Does this make any sence?

It makes perfect sense. I got some resolution back by cutting my throws down to 30°. I'm not good at 3D but I would like to start trying elevators and harriers so hopefully what I have now will be enough.

- Joe

What radio/servos are you using? Do you have the ATV/EPA set above 100%? My 9C will go to 140% EPA which is enough to set up 3D rates on a switch and still give the the ability to set them lower for precision flying. It's still a matter of finding a good median for both on the linkages but in time and practice flights it can be done...

Good article here on how setting ATV and D/R affects servo motion. Geared to Super 7 but should apply to most computer radios in general

http://www.wtp.net/DBEST/Sup7388tips.html

This is exactly what I saw with the 605's. If you have plane set up for dual elevator servos, it's astonishing to see the difference. Like I said, I installed a 9202 on one side and left the 605 on the other to compare. Night and day. Slow stick movements produced jerky spurts of deflection on the 605 vs a much smoother tracking of the stick movement by the 9202 coreless.

The other big difference is the in flight constant trim fiddling Jason mentioned above. Gone with the 9202's. But with the 605's I'd find myself adding one blip of up trim and taking it out 2 minutes later, with the plane never really feeling 'locked in'.

Funny thing, the Cap is my first 3D type plane and It would make me very nervous to fly due to it's very hot elevator and shady reputation for tip stalling. After changing the servos, I immediately felt more relaxed with the planae. I realized that some of my previous discomfort was caused by the vauge handling of the plane with the 605's.

The difference here is the Hitec's use a 3 pole motor where as the Futaba is a coreless motor. HUGE difference but in the end you get what you pay for. I'd take a Futaba 9202 over a Hitec anyday of the week...

It's a big difference and the 9202 is considered a bit of a bum servo.
Since using JR digitals I've had a totalrethink... all I have left is some 6x futaba 9303's which aren't to shabby. But boy Im impressed with these JR digitals...nothing can compare.
I run all my servos almost maxed out. I say almost because I've have had some Futabas and a 7000 super servo lock out a max deflect (not fun) it's never happened with the dig's.

Servo choice is difficult so I work on a simple priciple. Simple aeroplanes have the simplest/cheapest available. Expensive/complicated aeroplanes the opposite.

Yeah, I hear you on the JR digitals. Its getting to the point where the servos cost as much as the rest of the plane. Once you've gone digital you will never go back.
I actually had a dream about Santa bringing a case of them for Christmas. Like a lot of good dreams, my wife woke me up before it was over. She thought it looked like I was having too much fun. Imagine if I told her what I was really dreaming of?