Selecting Electric Retracts

[leccyflyer]

9 hours ago, Geoff S said:

Perhaps it's better to use retracts only for take-off and belly land?  I've very little experience with retracts and most of the models I've built have fixed undercarriages (as their full--size counterparts). However, The WR Hurricane I'm currently building is destined to use electric retracts supplied by Richard - I'm just hoping they're of similar quality to the kits.

It's certainly a very good and sensible option that I've long used on several models with small wheels, which had a tendency to tip themselves up on landing. However for models with lots of scale detail on their undersides, expensive three, four or five bladed props, I do tend to persevere with landing on the retracts. There is also the fact that a well executed full flap, retracts down landing is one of the best moments in a flight IMO - hugely enjoyable.

 

D-Day anniversary tomorrow and I'm all loaded up with invasion-striped D-Day participants with a variety of fixed gear (Dakota), retracts (Tempest), hand launch and belly land (Spitfire and P-51) and dolly launch-belly land (WR FW190). The odd man out is that Dakota with fixed gear and that fixed gear definitely puts me off flying it - a warbird with doughnuts down and dangling is not a good look IMO.

[Gee Bee]

Thanks Brian for you quick overview of the various types of retracts.

 

What you say is very much inline with what I might have expected. This leads me to think that I'm most likely to experiment with the traditional mechanical type retracts when I decide to go down that route. 

 

Thanks also to the other contributors that have included some enlightening comments. 

 

I should say that I started flying RC in December 1976, and mainly flew slope soarers. I did also fly power models, but back then if they had undercarriages they were 'fixed'.

I also had some mini-warbirds that were hand-launched & belly landed.  I first joined a 'power model' club in 1990, where I did my BMFA 'A' & 'B' tests on the same day with a Club 20 pylon racer. Since then I've continued to fly a mix of slope soarers, other gliders & power models. 

Whilst I would say that I can do proper landings, it is refreshing to see others comment on the reality that sometimes, no matter how long you have been flying, you can get caught out and do a less than perfect landing. Been there, done that !

 

I currently only fly 'sport' power models & fully understand that warbirds don't look right with spindly undercarriages dangling for the whole duration of the flight.

I do have a scale ARTF model in my to build pile, but its a PT-19 which has fixed undercarriage as per the full-sized version. 

 

Ironically my only 'hands-on' experience with retracts on an RC model is on a 4m span ASW-28 glider. I found it easier to self launch the model with the single wheel retract down, where I could hold the wheel yoke with one hand & stabilize the model with the other hand on the LE of one wing. I would then retract the wheel after launch, & unless it was a short grass site I would belly land on open moorland with the landing gear in the retracted position. 

 

Yes, dedicated retract servos are unique in their actuation in that they only hold two positions & do not move in a proportionate manner like a 'normal' servo. They move from one set position to another once they receive a signal. Their motion is slower than standard servos but the torque is better. I would certainly use on with mechanical retracts. 

 

One thing mentioned by others is the use of 'Oleo' legs, & I think these have become more available to the mass market in more recent years. In my early RC years such Items were very specialized & from my recollection were only really seen on competition scale models. Oleos should help provide some resilience to operationally imposed loadings - providing the design parameters are not exceeded !

 

Happy flying to all ...... & safe landings .... however you choose to get back to terra firma. 

 

[Cuban8]

13 hours ago, Paul De Tourtoulon said:

As said servo slow won’t work on 'servo less retracts, my Seafire has two on different channels with a max servo delay on one, It's just enough to sequence them so that one starts to retract later on giving a scale effect.

A Seafire!..........the ideal scale subject for undercarriage collapses.........😁

[Nigel R]

22 hours ago, GrumpyGnome said:

I don't use a retract servo, just a decent 'normal' one...

 

retract servos run 180 deg for good reason, this puts zero load on the servo when locked either up or down; it also means the servo produces a lot more force at the extreme of travel, which all aids getting the lock in place; a normal servo does not.

 

To put it another way, the geometry works for the retract servo and works against a normal servo.

 

Of course you can modify some normal servos to travel 180 degrees, but far simpler and better to buy one designed for the job.

 

 

Whilst a well adjusted mechanical retract is very reliable, it is well worth checking the operation before every flight. Legs can bend and foul the wheel wells. Find out before committing aviation.

 

 

[Cuban8]

12 hours ago, Gee Bee said:

One thing mentioned by others is the use of 'Oleo' legs, & I think these have become more available to the mass market in more recent years. In my early RC years such Items were very specialized & from my recollection were only really seen on competition scale models. Oleos should help provide some resilience to operationally imposed loadings - providing the design parameters are not exceeded 

 

Decent sprung oleos can be had for not a lot of money (around twenty quid a pair for ones suitable for smaller/medium sized  models) from Ali Express. Worth browsing what they have on offer. I use Aliexpress quite a bit now and have had not one issue with their service BTW.

You could of course make a set yourself - after all they are essentially only a rod sliding within a tube. For a simple set, only basic tools would be needed - enough to drill and tap a few holes really. Other than that  a vice, hacksaw and files should be more than enough. Aluminium, steel bar and tubing of various sizes is all cheaply available in small quantities from ebay suppliers. Larger shouldered bolts and screws can be  adapted to make axles.  It's surprising what you can come up with only basic tools, a little care and ingenuity. 

I haven't needed to make a set from scratch so far, but I have upgraded commercial ones.........bigger axle, stronger fixing pin, larger screws etc.

 

[FlyinFlynn]

How about printing your own?  At least when you break them you can just replace the broken bit! 🙂  https://www.youtube.com/watch?v=LWZSr9QIn4o

 

There are, however, a couple of problems with the way KendinYap goes about his verrsion. He seems to connect 3 retracts up to a single xiao, with only one set of limit switches, this cannot properly protect against one leg getting hung up and drawing excessive current, for that, you would need one xiao for each retract unit. I also think the use of a esp32 is a bit OTT for this application, the feedback is not really needed as, with a couple of diodes you could just drive the retract motors one way until the limit switch is activated when the current will stop and when the Tx retract switch is thrown the retract can be driven to the other end via a reverse connected diode. Each retract is controlled from its own set of limit switches. All this should run fine on a ATTiny13.... 

 

The sketch itself is tiny!

 

//For ATTiny13

int RCSIGNAL; 
#define ButonA PB1
#define ButonB PB2

void setup(){
pinMode(ButonA, INPUT);
pinMode(ButonB, INPUT);

pinMode(PB0, INPUT);      // > Receiver Aux Channel | Alıcı Aux Kanal
pinMode(PB3, OUTPUT);     // > AT8238
pinMode(PB4, OUTPUT);    // > AT8238

}
void loop(){

RCSIGNAL=pulseIn(PB0,HIGH); 

if  (digitalRead(ButonA) != 1 and RCSIGNAL <1400)
 {
  digitalWrite(PB3, LOW);
  }
  else
  {
  digitalWrite(PB3, HIGH);
  }

   if (digitalRead(ButonB) != 1 and RCSIGNAL > 1700)
{
digitalWrite(PB4, LOW);
  }
  else
  {
  digitalWrite(PB4, HIGH);
  }

delay(10); 

}

 

and a pcb - Gerber_ElectricRetracts_no-feedback_PCB_ElectricRetracts_no-feedback_2_2025-05-28.zip

 

Happy tinkering!

[GrumpyGnome]

2 hours ago, Nigel R said:

 

retract servos run 180 deg for good reason, this puts zero load on the servo when locked either up or down; it also means the servo produces a lot more force at the extreme of travel, which all aids getting the lock in place; a normal servo does not.

 

To put it another way, the geometry works for the retract servo and works against a normal servo.

 

The retract mechanism means there is no load on the servo when fully up or down - I have left it plugged in and switched on and the battery capacity is undeminished after seversl hours. The servo throw has been adjusted to make sure of a positive lock without draining the battery.

 

The weight of the leg and wheel when retracting are well within the power of the servos, and do not seem to put much load on the battery pack.  

 

I'm not saying this is better than a dedicated retract servo,  just reinforcing that there are a number of options open to us

 

[Cuban8]

Always makes me laugh when I see a  - 'now you see 'em, now you don't' - installation where the wheels are whipped away at lightning speed......usually air operated retracts that are operating at full whack to the cylinders.😁

[GrumpyGnome]

Yeah, it's pretty cool when they go up, and down, slowly, and not quite simultaneously.....

[Paul De Tourtoulon]

3 hours ago, Cuban8 said:

Always makes me laugh when I see a  - 'now you see 'em, now you don't' - installation where the wheels are whipped away at lightning speed......usually air operated retracts that are operating at full whack to the cylinders.😁

I have the air up spring down ones in my Rafale, take off, vertical climb, shut the turbine off invert and retract when no one can see or hear them.

[Bruce Collinson]

As I feared, one pulled out on first landing (it was windy) and that's it, I've had it with retracts on grass and they are coming out.   Denied a chance to come last in Pick 5 too.   They are not returning to the long-hangered Saphir either, when I get round to it.   Bought a c/f u/c from LMS this morning, just have to adapt the pipe channel to accept it (and mend the wing)

 

UKCAA fly-in with two clubmates nevertheless a very pleasant day out. 

 

BTC

[Stuart Z]

Bruce

Sorry to hear that.  My Calypso and Joker both had retracts, back in the day, and over many flights they stayed where there were put even used on grass.  They were mechanical retracts if that makes a difference, but surely the only real difference is the servo pushrod?

Be interesting to see what others have experienced. 
S

[Nigel R]

In my (slightly) limited experience of retracts: Solid install of any mount plate or bearers is key to long life. Get them keyed into the wing structure and spread the load, just like any regular U/C.

 

That's it really. "With a long enough lever" etc. Get the retract on a nice solid mount and make sure it's firmly attached. Easier on a built up wing - foam wings (in my opinion) need extra structure embedded to achieve a decent mount for the retract unit.

 

I would note, the average retract unit has about an inch gap between fore and aft fixing points. This is not so dissimilar to the length of "arm" on a torque rod setup. Nearly every fixed gear sport model I've built has that kind of U/C. To my mind this means the retract unit is putting a similar strain onto the wing structure as a fixed U/C would - so it's not impossible to handle. Fuselage mounted aluminium or carbon gears are similar - often quite narrow where they attach. These also need solid structure to bolt to.

 

YMMV.

[Brian Cooper]

15 hours ago, Stuart Z said:

Be interesting to see what others have experienced. 

I have a some models which are about 15 years old and fitted with retracts. Two have mechanical retracts and two have electric retracts. 

They are mostly flown from a grass strip,  and occasionally from a hard runway. . . The wheels go up, and the wheels come down... and have never ripped out on landing. 

No problems so far. 

[leccyflyer]

Like Brian I have models which are more than 15 years old, fitted with retracts, or which have been fitted with retracts in the past, flown almost exclusively from short, well maintained grass fields. These are predominantly moderately sized foamie warbirds, predominantly electric servoless retracts and the quick stats that I have just calculated from my model spreadsheet indicate that I have had a retract problem -either a malfunction, or, more commonly ripping the retract out of the model on landing -in something like 43% of those models. That figure might be a bit off, as I haven't specifically excluded a small number of models which are retract equipped, but currently awaiting a maiden flight. Approximately 50% of retract equipped models having had a retract problem in the past is probably a fair estimate.

 

Most pertinently four models are currently u/s due to needing non-functioning retracts fixing. "Got to do better than that Flight,,,,"

[Bruce Collinson]

My understanding of torque rod fixed UCs is that they (can) have a long torque spring which is in compression under load, trying to push the fixing block out of the top of the wing; easier in theory to make robust than the opposite force, trying to rip the retract out by tension with a twisting load from front to back.   As Nigel R observes, the length of the fixing plate is relatively short.

 

I observe experienced peers straightening legs after every flight.

 

Must improve my landings.

 

BTC

[Nigel R]

That's the rub Bruce, the torque rod setup is somewhat more straightforward to key into the main wing structure (especially in a foam wing). The underside block is mainly pushed upward into the wing, and backward; both of which it can easily resist in an average wing structure. The top block - holding the torque arm - is pushed forward; again, easily restrained by your average surrounding wing structure.

 

With retracts, if you can attach the mount plate securely to a couple of ply false ribs, then you have a similar structure in place - this will also distribute load to the upper and lower sides of the wing structure.

 

Coil springs on piano wire retract legs help absorb the loads. If I'm making my own legs I wind several coils on, rather than using a straight leg or a single coil. All grist for the mill. Sprung oleos are clear winners in terms of properly absorbing the bumps and bounces. All of which takes a bit of stress and strain away from the mounting points.

 

Unless your landings are the butteriest (?) of butter smooth and you fly from the perfect tarmac strip, the legs themselves will often need a tweak. C'est la vie.

[Bruce Collinson]

Yep.

 

The Supra Fly remaining hors de combat with wing repairs, and a modification to the elevator servos to remove suspected asymmetry in the single pushrod setup, I dug out the Curare after a couple of years when I saw a reasonable weather forecast (a lie, as usual, cross strip bumpy wind).   The only trike setup I have, old retracts demoted to fixed.

 

The Clevis popped off the nose wheel arm on every landing, including the one by a much better pilot than me.   All this after I discovered Mr. Mouse had chewed through the ESC to rx feed.   Refixing said clevis was keyhole surgery but we gave up after two operations and it’s now in a queue in the hanger for modification.

 

Did fly adequately, straight, bunts better than the Supra Fly but slightly underpowered so might be a new motor too.

 

I always sensed that the ply mount plate for the retracts was a bit poorly supported and I suspect that Far Eastern factories are skimpy with the glue in places.   The front of the plate needs a lot of stress distribution, easy in a kit build, one of several compromises in ARTFs.

 

I did try to replace the shiny Chinesium legs with proper piano wire but my bender scarcely put 2 deg in the 6mm wire, hence the straight legs which I think were my downfall at Buckminster.   I was compromising my geo-political tablets of stone by surfing for sprung Oleos but the oft-cited website was sufficiently impenetrable that I lost the will to live just as I decided to put a fixed u/c onto the fus, which is now sourced (LMS) and in the repair queue.

 

on top of all this, it seems J. Perkins are launching a Kwik-Fli ARTF at Weston, so here we go again ….

BTC