Originally Posted by mavlo77 --
If you can't adjust your radio, just make sure that your throws are ok by using the right servo arm length etc... Then you'll be fine.

Originally Posted by kadil

Thanks Mark, did that, just wondering if I am missing out on some precision of control by not using the whole range. Not really important.

Speaking of precision though, when I overhauled the dropbear, I made a new hole in the control horn on the elevator to be a very neat fit for the clevis I use on that control surface. Well..... what a difference that made. I suppose it stands to reason, but if you want a challenge, try a loose joint on the elevator control horn and you'll know what I mean. I have made a couple of comments about smooth flights in other peoples video's. I think this was the reason for the difference.

Most non-computer radios provide for "full" movement of the servo as the default. Your servos should be move through about 90 degrees from end to end. This is what I get with my DX5 and AR500 or AR6200 receivers.

TonyS

Originally Posted by kadil

If you do not have these controls on your transmitter (like my dx5) is there any way to actually make sure you get maximum servo throws? I am sure I can manually move some of my servo's across a greater range than it actually uses with my setup.

(some servo's can be pushed, other(read cheaper servos) just feel like they are going to break if pushed)

Generally you don't want or need maximum servo throws. There are two issues with going much beyond 100% throw. First, it slows down response (since it takes 40% longer to travel to 140% than to 100%).

Second, as you go beyond the roughly 30-35 degrees that is normal angular throw on typical servos (i.e., 100% on the transmitter), the output becomes increasingly non-linear with a normal mechanical setup. (Note that this is not so with the huge throws seen on 3D models, where the angular deflection of the control surface is similar to that of the servo. That's why indoor 3D models can use 140% throws.) The result of excessive angular displacement of the servos is adverse expo. That is, the controls are most sensitive in the middle and become progressively less so towards the ends. This is the opposite of what we normally want.

The best way to set up a model like the Polaris for regular flying is to leave the throws at 100% initially and set up the mechanical linkages to give the maximum desired throw (which is probably the throws called for in the manual). Then set up dual rate at about 70%. Go fly it and see how it feels. You still have room to increase the maximum throw (though most people won't want to) and you have D/R to allow you to adjust for smooth flying.

As for expo, on the Polaris, I like about 20-25% on elevator and 30-35% or rudder and aileron.

Originally Posted by kadil

if you want a challenge, try a loose joint on the elevator control horn and you'll know what I mean. I have made a couple of comments about smooth flights in other peoples video's. I think this was the reason for the difference.

Right on! The ingredients of smooth flying:

• No slop in the linkages
• Using the most of the normal travel of the servos
• No excessive control throws (i.e., moderate control rates)

Note that the great advantage of expo is that it allows you to have both moderate control rates where it counts for smooth flying and plenty of control when needed at the ends of the travel. But too much expo can mean that high rates come on abruptly.

Finally, some people need to be reminded that the transmitter is not a video game controller. Less stick movement is better, not constant rapid jabbing!

Originally Posted by Daedalus66

Generally you don't want or need maximum servo throws. There are two issues with going much beyond 100% throw. First, it slows down response (since it takes 40% longer to travel to 140% than to 100%).

Second, as you go beyond the roughly 30-35 degrees that is normal angular throw on typical servos (i.e., 100% on the transmitter), the output becomes increasingly non-linear with a normal mechanical setup. (Note that this is not so with the huge throws seen on 3D models, where the angular deflection of the control surface is similar to that of the servo. That's why indoor 3D models can use 140% throws.) The result of excessive angular displacement of the servos is adverse expo. That is, the controls are most sensitive in the middle and become progressively less so towards the ends. This is the opposite of what we normally want.

The best way to set up a model like the Polaris for regular flying is to leave the throws at 100% initially and set up the mechanical linkages to give the maximum desired throw (which is probably the throws called for in the manual). Then set up dual rate at about 70%. Go fly it and see how it feels. You still have room to increase the maximum throw (though most people won't want to) and you have D/R to allow you to adjust for smooth flying.

As for expo, on the Polaris, I like about 20-25% on elevator and 30-35% or rudder and aileron.

All true I suppose. Lots of ways to skin the cat. Maximum servo travel has positive benefits (maximum torque, better centering) and negative side effects (longer time to achieve max control throw). The inverse is also true, moderate servo travel has positive benefits (faster time to achieve max control throw) and negative side effects (less torque, potentially worse centering). The mechanical angle issue is relative to the amount of throw AND also length of the servo arm. With more servo movement, the arm doesn't need to be as long, so this probably balances out. Ultimately, it all comes down to "is my servo set up meeting my goals?" Since low torque could contribute to poor centering, and poor centering is enhanced by less servo movement, I just thought I'd throw this on the table in case it might be related. ;-)

Originally Posted by kadil

Thanks Mark, did that, just wondering if I am missing out on some precision of control by not using the whole range. Not really important.

you'll want to use the full servo range by using a shorter arm if you can over dumbing the servo down using programming. this will be you the maximum tourq the servo can provide.

Originally Posted by kadil

Thanks Mark, did that, just wondering if I am missing out on some precision of control by not using the whole range. Not really important.

Speaking of precision though, when I overhauled the dropbear, I made a new hole in the control horn on the elevator to be a very neat fit for the clevis I use on that control surface. Well..... what a difference that made. I suppose it stands to reason, but if you want a challenge, try a loose joint on the elevator control horn and you'll know what I mean. I have made a couple of comments about smooth flights in other peoples video's. I think this was the reason for the difference.

Good to hear Kim!!!!
That's why I use epoxy board as control horn. You know, from the printer circuit boards in electronics. Very though and durable, no wear at all! It's demonstrated by the fact that it ruins your saw! . So I always make very precize holes, so that there is zero slop. If you try to move my elevator up and down by hand, you don't feel any form of free play. And yes then you can fly smooth! Totally agree .

BTW in my older Polaris I used separate servo's for left/right aileron because the I fed the music wire through a S-shaped tube in the wing. When everything was cured, I noticed that the mechanical resistance was huge.... That created a kind of slop and hysteresis in my aileron control. It's still flying OK, but in roll you miss a bit of fine control. In my new Polaris I use the torque rods as in Steve's plan. That's definitely more accurate and lighter control. So you can do very smooth and slow rolls.

Originally Posted by fmidgett

All true I suppose. Lots of ways to skin the cat. Maximum servo travel has positive benefits (maximum torque, better centering) and negative side effects (longer time to achieve max control throw). The inverse is also true, moderate servo travel has positive benefits (faster time to achieve max control throw) and negative side effects (less torque, potentially worse centering). The mechanical angle issue is relative to the amount of throw AND also length of the servo arm. With more servo movement, the arm doesn't need to be as long, so this probably balances out. Ultimately, it all comes down to "is my servo set up meeting my goals?" Since low torque could contribute to poor centering, and poor centering is enhanced by less servo movement, I just thought I'd throw this on the table in case it might be related. ;-)

My point is that with the kind of models we are flying, 100% throw gives a good compromise of resolution, speed, linearity and torque. It isn't till you get down to 50 or 60% throw that you start to notice a significant loss of precision in control. Much more important in most cases is the quality of hinging, freedom from slop in the pushrod bearings, and lack of bowing in the pushrods. Likewise, the better linearity of control provided by not going to extreme angular displacements outweighs any possible gain in accuracy of positioning. As for torque, it simply isn't an issue if an appropriate servo is chosen in the first place. For the Polaris, a decent 9g servo will do the job nicely at 100% throw.
I've seen a number of problems of linkage geometry caused by things like misaligned aileron pushrod tubes that cause bending of the wire and consequently adverse aileron differential. Such problems are made worse by using extreme throws.

Originally Posted by mavlo77

BTW in my older Polaris I used separate servo's for left/right aileron because the I fed the music wire through a S-shaped tube in the wing. When everything was cured, I noticed that the mechanical resistance was huge.... That created a kind of slop and hysteresis in my aileron control. It's still flying OK, but in roll you miss a bit of fine control. In my new Polaris I use the torque rods as in Steve's plan. That's definitely more accurate and lighter control. So you can do very smooth and slow rolls.

That's interesting. I'm planning to use torque rods in my new Polaris to maximize precision (though I should say that the standard setup with a single servo and pushrods works very well if you take care to align the pushrod tubes properly).

One thing I find is that the plywood elevator and aileron horns are just a bit on the short side. I have set them into the foam so that at the front the ply is just on the surface of the foam. The rear end is set in and this gives a small angle which further distances the push rod hole from the hinge. By the way, it should be stressed that the hinges should not be centered in the surface but at the top -- this further increases the effective horn length (i.e., the elevator or aileron should be angled back at 45 in a single bevel, not bevelled top and bottom).

@mavlo - I used some plastic from a cd stack cover (I think it is polythene) to make control horns out of. Originally I was drilling the hole (which made it oversize). This time I heated up a needle with a cigarette lighter and pierced a hole and it was perfect. PCB sounds better though.

@daedalus66 - were you secretly looking over my shoulder? Your comment about jabbing at the transmitter like a video game controller is exactly the habit I had to break. I think I am through that phase, but I was definitely there.

I probably should have done some measurements before opening this can of worms about servo deflections, but I certainly learned a lot. I probably am getting 100% of the intended servo throw to maximise linearity and responsiveness, even though they can physically be pushed further


Cheers guys!!!

I'm off to the local S.M.A.L.L. event tomorrow. The Polaris fits the spirit of the event very nicely and I'll also take a 30-year old original design trainer powered by an equally old OS Max 15.

In case you're not familiar with it, S.M.A.L.L. is what the name suggests, a fun fly for models with engines of less than 0.28 cu. in. displacement (0.30 for 4-strokes) or roughly equivalent electric power.

Originally Posted by kadil

Thanks Mark, did that, just wondering if I am missing out on some precision of control by not using the whole range. Not really important.

Speaking of precision though, when I overhauled the dropbear, I made a new hole in the control horn on the elevator to be a very neat fit for the clevis I use on that control surface. Well..... what a difference that made. I suppose it stands to reason, but if you want a challenge, try a loose joint on the elevator control horn and you'll know what I mean. I have made a couple of comments about smooth flights in other peoples video's. I think this was the reason for the difference.

Re the loose joint. How big was the hole and how small was the pin?

Originally Posted by Flatbus

Re the loose joint. How big was the hole and how small was the pin?

we are talking 2mm hole and 1.5mm pin. I knew it was loose, but really wanted to keep the clevis I salvaged from another model. It allows you to adjust the length of the control rod by screwing the clevis in and out, but did not have a good selection of drill bit sizes.

How high or rather what was the distance from the hole to the hinge? That makes a difference too. If the distance was 2cm, that 0.5mm slop would not be as affected as a 1cm distance from the hinge (moment arm). Just curious to know, because the 1st polaris I built also had some play in the hole because like you, the size of the drill bit I had was a tad too small for the pin so I had to go the next size bigger which gave it some play. However no complaints on my part on the controllability for my level of skill, low skill level so high accuracy is not my aim at the moment.

Originally Posted by Flatbus

How high or rather what was the distance from the hole to the hinge? That makes a difference too. If the distance was 2cm, that 0.5mm slop would not be as affected as a 1cm distance from the hinge (moment arm). Just curious to know, because the 1st polaris I built also had some play in the hole because like you, the size of the drill bit I had was a tad too small for the pin so I had to go the next size bigger which gave it some play. However no complaints on my part on the controllability for my level of skill, low skill level so high accuracy is not my aim at the moment.

10mm from the hinge, maybe more than .5mm slack, not really measured, just guess by looking at it. It was quite noticeable slack though just moving the control surface by hand.

I think it is more important to have this tight precision of mechanics, the less your skills are. Otherwise you constantly have to adjust for the slack while flying.

Agreed, no slack better. But the mistakes I've made along the way have been still manageable. Live and learn. Gets better as we go along. If we take the trouble to build with accuracy.