Pull-Pull Control Systems

[Martin_H]

I have noticed over the years quite a few construction articles regarding pull-pull control systems that mention much care is to be taken ensuring the servo end and surface end control arms should be of similar size to eliminate differential binding.

Whilst that may be so there is an easy and effective way of setting up pull-pull systems that is fool proof and as old as powered flight itself. Most pull-pull systems as mentioned above run the control cables parallel, from the servo to the control surface, but much is to be gained by crossing the cables over mid run.

The size of the control horns no longer becomes an issue, rather, can be optimised for scale effect or to utilise full servo range (and torque) as a precise control movement. Instead of a 15 deg servo movement being required to produce a 15 deg surface deflection, by proportionally increasing the size of the surface control arm for example you could achieve a 15 degree control deflection using a 60 deg servo movement. 4 times more accuracy in servo placement and spreading the wear over a greater gear range in the servo. Another advantage is the initial rigging of the control system needs not be so extremely accurate, as any tiny differences in the angular alignment of the servo and control arms is no longer an issue.

Obviously, this means being able to cross the cables over, either internally or externally, and some true to scale applications may not allow this.

There are considerations though. You don't want bare cable rubbing at the cross over point. Use nylon covered trace wire or (for those DLG's) one of the hi tech fishing lines now on the market. Mount the servo with a slight radial twist relative to the control surface. This will set one cable lower than the other avoiding rubbing at the cross-over point. Check cables before and after each flight, if possible. Keep spare cables in your flight box.

Make more use of your servo's travel, increase the accuracy of control surface deflection, reduce control surface flutter and extend your servo gears useful life.

Oh, and don't forget to reverse your servo direction!

[John F]

Thank you for the info Martin. Installing the servo at a slight angle is a very useful tip.

With regards to the fishing wire comments, what is a DLG? (I know it will be so obvious once I know but the only daft question is the one you don't ask)

What strain of trace wire are we talking about? I take it the minimum 20lb trace wire would be more than adequate?

[Dave Bran]

DLG = Discus Launched Glider

[Martin_H]

Thanks Dave, for your answering the DLG question.

John F, you would have to be guided by the size of your model and what you are putting it through - i use a commercially available system that is available in a number of sizes, simply because they include all that is needed to make the cable runs. Suffice to say the 4-40 system I use in my larger models (including my Hangar 9 E-Flight 110 powered 1/4 scale Cub) in addition to the same manufacturers 1/4 scale turnbuckles, and the 2-56 system in my smaller models, including my Wasabi Dynamic Slope Soarer. Where I use a pull-pull system in my DLG's I have used both kevlar and Dyneema line successfully.

[Peter Miller]

Providing the actual holes for the connection of the cables are in line with the pivots/hinge line there should be no trouble with different lengths of horn to out put arms. I have never had any. Try drawing it out and you will see.

Useful tip. If you have the room it is far better to set up an idler bellcrank and connect the servo to that with a pushrod. Much easier to adjust trim and saves a lot of hassle if you want to remove the servo.

[Allan Bennett]

Posted by Martin_H on 23/02/2014 02:08:43:

I have noticed over the years quite a few construction articles regarding pull-pull control systems that mention much care is to be taken ensuring the servo end and surface end control arms should be of similar size to eliminate differential binding.

Whilst that may be so there is an easy and effective way of setting up pull-pull systems that is fool proof and as old as powered flight itself. Most pull-pull systems as mentioned above run the control cables parallel, from the servo to the control surface, but much is to be gained by crossing the cables over mid run.

The size of the control horns no longer becomes an issue, rather, can be optimised for scale effect or to utilise full servo range (and torque) as a precise control movement. Instead of a 15 deg servo movement being required to produce a 15 deg surface deflection, by proportionally increasing the size of the surface control arm for example you could achieve a 15 degree control deflection using a 60 deg servo movement. ...

I don't understand this. By "differential binding" I presume you mean the effect where the cables tighten up as the servo moves, due to different lengths of driving and driven arms. From a purely mechanical point of view, it doesn't matter whether the wires run straight or crossed, you still need the servo arms the same length as the horns on the control surface, otherwise you'll get the binding.

On a couple of models where I use pull-pull control, I use a system (can't remember who it's by) that uses a spring to tension the wires as they go round a wheel. This ensures the wires don't go slack, and compensates if there's any difference in driver and driven lengths. The wheel is then driven by a pushrod from the servo.

[cymaz]

I think that wheel compensator was sold by Flair. I use a couple in my Puppeteer.

[Tony K]

Posted by Martin_H on 23/02/2014 02:08:43:

Whilst that may be so there is an easy and effective way of setting up pull-pull systems that is fool proof and as old as powered flight itself. Most pull-pull systems as mentioned above run the control cables parallel, from the servo to the control surface, but much is to be gained by crossing the cables over mid run.

The size of the control horns no longer becomes an issue, rather, can be optimised for scale effect or to utilise full servo range (and torque) as a precise control movement. Instead of a 15 deg servo movement being required to produce a 15 deg surface deflection, by proportionally increasing the size of the surface control arm for example you could achieve a 15 degree control deflection using a 60 deg servo movement. 4 times more accuracy in servo placement and spreading the wear over a greater gear range in the servo. Another advantage is the initial rigging of the control system needs not be so extremely accurate, as any tiny differences in the angular alignment of the servo and control arms is no longer an issue.

Martin, could you expand on why it is necessary for the cables to cross.

For example, if you want 15 deg deflection with ,say, a 40 mm radius control horn, you need 10,4 mm axial "pull". To get 10,4 mm axial movement for 60 deg of servo movement, you need a 12 mm servo arm.

That means that the cables, if straight, would not be parallel. In fact 24 mm apart at one end and 80 mm apart at the other end. Is that an issue? If it is, how does crossing the cables alleviate the problem?

I am currently incorporating pull/pull into a model for the first time, hence my interest.

[cymaz]

Links

Hope these help Tony K

[Bob Cotsford]

Posted by Peter Miller on 23/02/2014 08:51:31:

Providing the actual holes for the connection of the cables are in line with the pivots/hinge line there should be no trouble with different lengths of horn to out put arms. I have never had any. Try drawing it out and you will see.

Useful tip. If you have the room it is far better to set up an idler bellcrank and connect the servo to that with a pushrod. Much easier to adjust trim and saves a lot of hassle if you want to remove the servo.

In theory the tension may vary with different size servo and surface horns, but in practice any variation is negligable. I don't think I've ever matched horn sizes and often size them to gear down the movement and never ripped a rudder off or caused a servo to fail. Sometimes I cross the cables, sometimes I run them straight, it just depends on the layout of the model.

As Peter said, an idler belcrank is a great way to do it if you have room.

[Martin_H]

Best idea, gentlemen, is to draw it out then figure what happens. Better still, get some balsa strip and pins and try it out on the building board. It's a very simple geometry solution.

[Former Member]

[This posting has been removed]

[Dave Miller]

I think the equal size thing is another of those old wives tales that gets a regular airing; for it doesn’t matter if the arms are of a different dimension to the horn spacing, for that only affects the gearing not the wire tension, just the same as using different hole spacing on pushrods systems. The important thing is that the wire length is the same both sides, the route doesn’t matter. I usually cross the wires with in a fuselage because I get a better exit angle for the wires at the back. I also usually route the wires through 1/16” bore aluminium tube at the exit and sometimes these are bent to a gentle “S” shape to avoid structural members.

One other point is that you do not need to incorporate adjustment at both ends of the wires.

[john stones 1 - Moderator]

I agree with you Dave

adjust on outside, its cheaper more convenient

I use Bowden cable outers for exit guides, buy fishing trace as well, much cheaper

[PatMc]

As already said there's no need to have the input - output at the same size. If we did we'd either have around 60 degrees deflection each side of neutral or we'd have to use a large amounts of reduction rates, which would reduce the servo effective torque & increase effective gear slop. Because of the relatively long distance between servos & control horns there's no need to use cross over or bother about servos being in line with the horn centre.

Examples of gearing down movement with acceptable, best & wrong geometry :

a) No binding as the cables will have equal tension over the whole movement. However there will be a constant tension on the cables & the actual control surface against the hinge post (e.g. fin TE) which can cause top hinges to pull away when the horns are low on the rudder.

b) & c) No binding as the cables will slacken during action, the tension on the "pull" cable will be dependant on the slipstream. If the control is left off normal when the model isn't in use there will be no tension on the cables or pressure on the hinges.

d) & e) Will have binding when off normal as the "pull" points will be progressively closer together than the inactive points when in use. Also constant tension on cables when at neutral.

[Bob Cotsford]

Nice illustration Pat!

[Rich too]

does it matter if the cables rub on the fuselage on exit? or on a former inside the fuselage? I have always avoided it, but I am assembling a model and the factory exits are not in the right place where it is possible to avoid the cable rubbing.

[Geoff S]

Not really a problem provided the friction is minimised. On my DB Tiger Moth with scale(ish) closed loop elevator controls I feed the cables through snake outers to the scale exits. In fact the top cables on the full size feed through a fairlead on the tailplane leading edge as a straight line would take them through the tailplane itself.

Geoff

Edited By Geoff Sleath on 22/07/2017 11:04:28

[Rich too]

Thanks Geof, that's what I thought. Unfortunately the cables are already made, otherwise I would have used snake outers too.

[Former Member]

[This posting has been removed]

[Rich too]

Thanks Tom. I now realise that only the one doing the pulling is tight, and I wish I had known that before as it would have made setting up easier! Hey ho, I've done it now

Edited By Rich2 on 22/07/2017 15:25:01

[kc]

I know the wire used for closed loop is some sort of fishing stuff but exactly what should I buy? It's easier to get from a fishing supplier than a specialist model supplier these days and I bet it's cheaper. ( I notice that The Range shops have all sorts of fishing stuff)

[John Bisset]

Another option is control line wire - 'Laystrate', 'Staystrate' or similar if you can find it. Plenty strength..

Edited By John Bisset on 22/07/2017 20:08:16

Edited By John Bisset on 22/07/2017 20:08:43

[PatMc]

Posted by kc on 22/07/2017 19:21:22:

I know the wire used for closed loop is some sort of fishing stuff but exactly what should I buy? It's easier to get from a fishing supplier than a specialist model supplier these days and I bet it's cheaper. ( I notice that The Range shops have all sorts of fishing stuff)

I use fishing "trace wire", either 20lb or 30lb breaking strain. It's nylon covered & it usually comes with a number of crimps, if you need more crimps they're easily made from fine bore brass tube.

PS one advantage over c/l wire is that the nylon covering eliminates the risk of electrical "noise" being generated.

Edited By PatMc on 22/07/2017 20:32:17

[kc]

Thanks for that PatMc.

What is the best way to pass the wire through the crimps. Just through twice or is it better to do some fancy loop and third pass? ( I am sure I saw a diagram once which showed a fancy method)