Servo interactions

[Steve]

Hmm, i've always used end points. I've done the level swash but it doesn't always work when giving full neg or pos so i just adjust at the field till it's ok. Duncan Osbourn in his Idle Up tutorials said use end points to dial out interactions. Going into mixes is something i've never dabbled in. It's another of those things that sound confusing but it's simple once you know what your doing.

Exactly. You know it makes sense.

[Nick B]

Thanks for the info Steve,my 700N had not been nice to fly since upgrading to alloy servo wheels.went out to the hangar last night and tried different positions and have now got just 2 points of subtrim on elevator and 3 on the aileron where as before i had 15 and 17 respectively.

going to leave the endpoints at where they were,not all at 100% but will see how it goes.

[scallybert]

The sub-trim and end-points are used to compensate for deficiencies/properties of the servo.

eg the servo horn doesn't actually centre when the Tx sends 'zero'; and the servos don't all have identical travels. [Part of this is that there isn't necessarilly a 'correct' position for the servo horn.]

[There are other reasons you may want to use sub-trim & end-points.]

Unless you have programmable servos, I can see no sense in suggesting sub-trim & end-points shouldn't be adjusted.

[the doc]

I still dont get it.

Adjusting endpoints shouldnt bring in interactions - adding trim will. Endpoints are just that - where the servos are at the end of your travel. If you have adjusted them to get the swash level at bottom, mid and top stick then surely that is ideal? This guys seems to make sense to me...

Trueblood Engineering - Content

I just used this method to sort out my head without mucking about with spirit levels. Hopefully fly it tonight and see if it stays straight

Just out of interest.... people with flybarred choppers - how much trim do you add to your chopper when it is in the air? Do people tend to have trim or do most of you manage to avoid it altogether?

I am hoping that by keeping the swash level through the range I won't need to use trim at all -but I am not sure how prevalent it is.

[lankey]

Just out of interest.... people with flybarred choppers - how much trim do you add to your chopper when it is in the air? Do people tend to have trim or do most of you manage to avoid it altogether?

I am hoping that by keeping the swash level through the range I won't need to use trim at all -but I am not sure how prevalent it is.

Don't use regular trim at all. IMO if you have the swash level at low, middle and high and it's pretty much spot on in-between those points, and your COG is spot on, just go fly.... As Tony said, there are far too many forces acting on the heli, just learn to fly with them.

[LaurenceGough]

I never, ever use normal trim. Level swashplate, good COG, correctly set-up head = neutral flight tendencies. A flybar heli will never be perfect without compromising agility, F3C set-up or not . Hence my love for V-bar.

[Mark_T]

Adjusting endpoints shouldnt bring in interactions

Just how much of an issue this is remains up for debate, but yes, you will get unwanted interactions because altering the end point changes how far the servo has to travel to reach the end which in turn means it will get there faster or slower than the other servos.

[Steve]

My system assumes that the 3 servos all have exactly the same amount of travel, and given modern mass production techniques, I think this is a fair assumption, especially when you're paying £100 or more for each of 3 identical digital servos. Now, assuming this to be the case and all 3 servos are set to 100% travel with 90 deg servo arms, dead level washout arms, level mixing arms, zero degrees pitch and all that stuff at exactly mid stick, then the only remaining thing that can throw the paddles out at top and bottom stick is tiny inconsistencies in push rod lengths. There are several push rods between the swash plate and the blade grip, and even one rod being half a turn out compared to it's corresponding partner, can introduce and be magnified at the blade grips and the paddle.

The important thing here is not so much the swash plate being level at the 3 positions, but more importantly the paddles staying level at the 3 positions, and then again at the two 90 degree positions of full elevator control and full aileron control, so 6 positions in total. It's the paddles and blades that control the direction of the helicopter, not the swash plate, so the focus of our attention should really be on their final output positions rather than the mechanics below.

Most helis require a little bit of right swash tilt to compensate for the sideways force imposed by the tail rotor, and also a little bit of elevator tilt to account for any desired or undesired descrepency with the centre of gravity and the effect of the horizontal fin - big flat fin gets pushed down in the rotor downwash which pulls the heli backwards, so you have to introduce a little bit of forward trim somewhere to compensate. There's a lot of things going on here, over and above just the head setup.

Personally, I adjust the servo to swash plate push rods to get a steady hover with zero trim, and if the swash ends up being ever so slightly tilted right and forwards in the hover, then so be it - that's where my heli (not yours by the way as it might be different) wants to sit steady in a hover. What I then want to do is perfectly replicate that paddle position at full top and full bottom stick which is where the lollypop stick, blue tac and bubble come in.

Let's say I adjust the servo endpoint on the aileron servo to 90% and leave the collective servo at 100% (the two servos that are side by side across the swash plate) to get my swash plate dead level at full stick. Great, the heli now climbs at full stick dead straight, but then when I roll right (say), one servo is going to 100% travel and the other is only going to 90% travel, so I have inadvertently added a tiny bit of elevator into the roll, which is going to skew the helicopter off line.

If I leave the endpoints at 100% all round and use a mix to tell both the servos to give a tiny touch of aileron at full collective stick (thereby half increasing the travel on one servo and half reducing it on the other) then again it will climb dead straight as I have a nice level swash, but when I roll, and the mix isn't now deployed, I'll get 100% servo travel on each servo, so no elevator input. Hey presto, nice roll.

The same argument applies the other way with pitch to elevator servo travels.

Then, after all that, we can start to look at deliberately introducing extra aileron to elevator and elevator to aileron mixes to compensate for gyroscopic procession and phasing errors, which I also do.

As has been said already, it's easy when you know how, and my setup routine is mostly done on the bench and doesn't take very long at all, with just a few flights to fine trim bits in a certain order.

[Nick B]

Fancy coming flying down wiltshire this weekend Steve,let you fly my 700 lol

[scallybert]

My system assumes that the 3 servos all have exactly the same amount of travel,

Ah...

and given modern mass production techniques, I think this is a fair assumption, especially when you're paying £100 or more for each of 3 identical digital servos.

Based on my experience with HS65MGs (etc), it's not a fair assumption.

It may, or may not, be with £100+ digital servos - though I guess you can then use a servo programmer to correct them.

Let's say I adjust the servo endpoint on the aileron servo to 90% and leave the collective servo at 100% (the two servos that are side by side across the swash plate) to get my swash plate dead level at full stick. Great, the heli now climbs at full stick dead straight, but then when I roll right (say), one servo is going to 100% travel and the other is only going to 90% travel, so I have inadvertently added a tiny bit of elevator into the roll, which is going to skew the helicopter off line.

No.

The endpoints effectively 'scale' the movement, and we're adjusting it so they move the same amount.

What you're describing is what would happen if you *didn't* adjust the endpoints.

The endpoint and sub-trim adjustment is compensating for the servo shortcomings, so they approximate the perfect ideal that your servos apparently embody.

I agree that it's the paddle & blade pitch that you actually care about.

Of course, if the servos are perfect, then the set-up would result in no messing with sub-trim or end-points, anyway. (Well, except to compensate for the link pitch not allowing the links to be exactly the right length.)

[the doc]

That does sound iunteresting Steve.

Any good resources on how to do this simply without buying that guys DVDs/ books though?

[Steve]

Yes the endpoints scale the travel, but when you adjust them, you use them to adjust just one control function (i.e. collective pitch), but the same 3 servos are used to also control the aileron and elevator, and now you have skewed those controls by limiting the travel of one or two of the servos. I don't know about 450 size Hitec servos, but expensive JR and Futaba full size digital servos are pretty consistent and accurate - that's what you pay all that dosh for.

The slight interactions are a function of the head design, build accuracy and just how close to 90 deg you can get that servo horn. Can you tell the difference between 89.5 deg and 90.5 deg at the servo output arm? I use a mini steel set-square to set mine up and still get very small interactions. I've tried moving all 3 servos around the swash plate by one ball, and guess what, the interactions remain exactly the same, so it's not the servos, well not at least the expensive JR ones I like to use.

If you're even a tiny bit out at the servo arm, or your wash out arms are just a fraction of a mm out at mid stick, or the mixing arms a squidge out or any combination of these and other things, then you will get a lovely mid stick position and then one or more of the controls will be compromised at full top and/or bottom stick. The errors get amplified through the head as all the moving parts reach the limits of their travel..

[Steve]

Bottom line for me is:

1. Go for an uber-accurate setup on the bench and use mixes to dial out elevator and aileron interactions at top and bottom stick.
2. Test hover, trim for stable hands off hover, then land and translate those trims into transfer arm-swash plate adjustments to get the hover bang on with zero trim.
3. Engage digital phasing adjustment and go fly - lots of fast forwards rolls and stationary flips and adjust the phasing until they are straight - maybe one or two flights.

Now I have a heli that rolls straight, flips straight, climbs straight and tracks beautifully in fast forwards flight, all with the exception of my naff thumbs of course which introduce further unknown interactions by default!

[cjcj1949]

"Most helis require a little bit of right swash tilt to compensate for the sideways force imposed by the tail rotor." from Steve

Are you sure about that? The heli needs to be tilted so that the main rotor can provide a force to equal that produced by the tail rotor. If what you say is true then swash levellers must be tilted.

[Steve]

It's not a lot, but defo just a tiny bit.

The point about swash levellers is that they are a point of reference, somewhere to start from which you can be confident is always the same on every heli you build and setup.

I use one to setup my initial zero degree position and then work from there by testing the paddles and flight testing.

At the end of the day, it's how it flies that really matters, not what the swash leveller or pitch gauge says.