Installing, set up and testing of servos.

[toto]

What is your typical procedure for installing, setting up and testing of servos. Is there a certain order / sequence or plan of attack to get the best, most accurate setup.

 

This is something that I'd like to standardise my approach to.

 

Any comments ?

 

Toto

Edited July 16, 2023 by toto

[john stones 1 - Moderator]

Never use a tester myself, always horn off, tx/rx on n off we go.

If I want to test a servo, I'll just use a Y lead off a current model, I don't care what a tester says, It's my radio gears opinion I'm after.

[Martin Harris - Moderator]

To a certain extent it depends whether it’s a build or an ARTF assembly.  In the case of a build, I’m planning servo installation at an early stage but ARTFs usually just require servos installing into preset mounting positions. 
 

Setting up - I do it with the transmitter rather than a servo tester. I aim to set the linkages to give maximum anticipated movement at full stick deflection - don’t fall into the trap of dialling down movement using travel values which will lose both holding power and finesse. 
 

 

[john stones 1 - Moderator]

Yep, same here, I set amount of movement up mechanically via horns, only fine tuning done with TX

[Paul De Tourtoulon]

I set it up while building with a servo tester, then when it's finished I start again with my radio, and none of that dual rate down to 50% all done mechanically on the servo and command horns, with my memory more than once I have left my tranny on and ruined the battery, before all those stick alarms came in.

[GrumpyGnome]

I prefer to test my servos using the set up that'll be used when the model flies - tx, rx, esc/battery pack, servos. 

 

I make sure they work in the correct direction

Switch it all off

Then install the servos

Then put the servo arms on in the 'best' orientation

Then connect the surfaces

Then centre all controls (incl throttle for ic)

Then adjust the control rods to get mechanical neutral

Then select the control and servo holes that give desired deflection

 

Typically, for my maiden, I'll have three flight modes: recommended throws; a higher level of throw; a lower level of throw (and I generally also add expo - 25%, 30%, 20%)

Following the maiden, I'll delete two of the modes and 'fine tune' the favoured one - mechanically for throws as much as possible, expo on the tx

Then I'll add any mixes, e.g. throttle/elevator  

Then I may add flight modes for take off/normal flight/landing.

 

Then just fly fly fly!

 

I do use a servo tester in three instances: if I'm just replacing one servo (I don't know why........); to make sure electric retracts work properly; if I want to 'cycle' a suspect servo (although now I can make my tx do this, I'll probably not bother with the tester for this).

 

You'll soon find a way that works for you......

 

GG

[Engine Doctor]

A tester is very good to set up a model while building / assembling saving you having to have all you radio gear get covered in sawdust and glue etc. You stil have to set them up to your radio for centering and throws when modrl is finished . For a few £s they are a handy tool to have. Besides setting throws and centering they are capable of doing a soak test yo check out servo reliability .

[Brian Cooper]

^^ All of the above. ^^

 

Plus, personally, I "soak test" any new servos for about 90 mins on a tester.  They are just left running back and forth. 

In times gone by, if something electronic was going to fail, it generally failed in the first hour or so. If it didn't fail, it was probably good for a long life. 

 

It is a system which found a couple of dodgy servos in the dim and distant past but I have to say I haven't uncovered any naughty ones for about 20 years. 

With today's modern electronics, the soak test is probably no longer necessary but old habits die hard. 

[toto]

Thanks chaps for all the responses, I think GrumpyGnome has given me the sort of run down sequence I was after. I just need to familiarise myself with how to do it and everything is good.

 

One of my club mates set up the servos etc on my Arising Star but 5o be honest, he was working that fast that I didn't appreciate exactly the principles of what he was doing and AI didn't want to interrupt him to ask. 

 

I get the impression that my club mates are always willing to step in and help so I'll maybe eventually get a bad weather day when we all wit down to copious amounts of coffee and discuss the principles of how to do this.

 

There is wo much to learn and this is obviously one of the finer points you need to get your head around in order to have your mod3l set up correctly for successful flying ...... it's just going to take time and experience.

 

Thanks again

 

Toto

[Jon H]

I just screw them in and go these days but generally i follow GG's general procedure. I completely ignore the throws in the instructions and just eyeball it. Same story with c/g if im honest. 

 

That said, i have been at this quite a while so i use experience of previous models to inform my choices. 

 

it always causes a row but i never use expo on an initial setup. Until the rates are right you shouldnt use it at all. Once the rates are as you want them add 10% or something if you need it. If i add up the number of primary control channels i have across my fleet i end up with 39 channels. Of those only 2 have expo on them.

 

I never use a servo tester or burn in new servos. In over 30 years i have never had a servo die on me out of the packet so i dont see the need. That said it does no harm...in theory. Its doubtful servos are designed to run constantly from full travel to full travel so you may do them more harm than good if some high power digi overheats.     

 

All that to one side the biggest recommendation i can make is to get all of the mechanical aspects of the radio tied up before you start messing about with sub trim and other nonsense on the radio. If the core mechanical setup is a mess no amount of computer tricks will save it. 

 

 

[Andy48]

I find a digital servo tester invaluable. The centre and end points are accurate compared to the transmitter, and I set up the control surfaces using the tester, to get the throws as described in the build data. Once the receiver is fitted, I check them again, but rarely need any adjustment. Almost always the maiden flight requires no more than a couple of clicks to achieve hands free flight.

Edited July 17, 2023 by Andy48

[GrumpyGnome]

Out of curiosity, how do you do that Andy48?

[Jon H]

One thing i did forget is to make sure you start with a clean model memory. If you are taking over a slot in the tx that was used for a now dead model dont try and rewrite it to suit the new one. Wipe the whole memory and start again. I have seen people have problems with weirdness as they forgot some mix or random end point adjustment buried deep in the system. Wipe it, start clean. 

[John Stainforth]

I begin with a blank memory with a new model, like Jon, with everything set at the defaults (e.g. 100% throws). I set the servos and horns for maximum throws and mechanical advantage, and the control surface maximum throws  in degrees (to say 30 or 45 degrees). One thing I am quite careful with is setting the servo horns on the splined shafts of the servos in the optimum positions (usually about 90 degrees to the pushrod). Those with 4 or 6 horns are cleverly designed so that they fit onto the splines subtly differently, so one can move the horns around to find the one that is in the optimum position on the shaft.

[Geoff S]

2 hours ago, Jon - Laser Engines said:

One thing i did forget is to make sure you start with a clean model memory. If you are taking over a slot in the tx that was used for a now dead model dont try and rewrite it to suit the new one. Wipe the whole memory and start again. I have seen people have problems with weirdness as they forgot some mix or random end point adjustment buried deep in the system. Wipe it, start clean. 

 

I almost always copy a model over, rename it and add a suitable image (often one of the actual model).   I clear all the channel settings but I keep all the motor enable/disable settings and the timers etc, which just saves a lot of time and means all my models use the same auxiliary features. 

[leccyflyer]

2 hours ago, Jon - Laser Engines said:

One thing i did forget is to make sure you start with a clean model memory. If you are taking over a slot in the tx that was used for a now dead model dont try and rewrite it to suit the new one. Wipe the whole memory and start again. I have seen people have problems with weirdness as they forgot some mix or random end point adjustment buried deep in the system. Wipe it, start clean. 

That's great advice. 😎 Always start with a clean model memory and methodically work through the menus to make sure that you have a clean installed, unencumbered by any inapplicable or unwanted mix or setting. If you have all of your switches assigned to the same functions for each model it only takes a few minutes to ensure that they are set up the same, with the same voice alerts if you use them.

[leccyflyer]

15 hours ago, toto said:

What is your typical procedure for installing, setting up and testing of servos. Is there a certain order / sequence or plan of attack to get the best, most accurate setup.

 

This is something that I'd like to standardise my approach to.

 

Any comments ?

 

Toto

I usually just fit them and see if there are any issues when it comes to setting up the model. That has worked fine in the vast majority of my models and in almost all cases where the servos had not been fitted to the airframe on arrival.

 

Having had some previous issues with Durafly servos as supplied in the PnP models I did adopt a procedure of leaving them on test with a servo tester for between 30 minutes up to an hour to reveal any issues. Unfortunately that didn't help, I still suffered an in flight servo failure, so I've since adopted a different protocol, in two stages.  If the model is a Durafly model, simply take out the servos for the main flight controls and replace them, The second stage came after it transpired that even a servo which is  not a main flight control - such as one flap servo  for example - if it goes immediately to the end of it's travel and then stops working, can still spoil your day. The new protocol there is to ground the model until you get a chance to replace the servos.

[Martin Harris - Moderator]

1 hour ago, Geoff S said:

 

I almost always copy a model over, rename it and add a suitable image (often one of the actual model).   I clear all the channel settings but I keep all the motor enable/disable settings and the timers etc, which just saves a lot of time and means all my models use the same auxiliary features. 

Having “liked” Jon’s advice not to try modifying an existing model which is very valid, I do have a different system for creating a new model. I have “templates” stored which contain base set-ups for various types of model such as IC, EP, EDF, gliders (timers, motor isolation switch, cross trims, auto-trim switch etc.) and make a copy to a spare memory slot which i immediately rename to the new model. 

[Paul De Tourtoulon]

I don't !, starting with a clean slate is more fun, and you won't forget how to use your radio,

[Simon Chaddock]

90 minute soak test?

I wonder how much of the servo life is use up doing this.

Is a 'used' servo more likely to fail?  

[leccyflyer]

The logic with modern electronics is that, apparently, they are most likely to fail immediately and be Dead on Arrival, or fail early on, hence the recommendation sometimes seen to soak test them. If they survive that initial testing the assumption is that they are going to be okay long term. Given that many of us have many servos, amounting to many thousands, which just keep on working for years, or even decades, suggests that servo longevity isn't a common problem. 

[Paul De Tourtoulon]

I have NEVER had a new Futaba servo fail on me, even after 10 years of use.

 

So why the soak test ?.

Edited July 17, 2023 by Paul De Tourtoulon

[Brian Cooper]

4 hours ago, Simon Chaddock said:

90 minute soak test?

I wonder how much of the servo life is use up doing this.

Is a 'used' servo more likely to fail?  

At a rough guess, probably 90 minutes' life is used. .LOL. . . Actually, it is probably less than that in real terms as the servos aren't under any stress. 

 

The test is purely to establish that the electronics are reliable.  But as said earlier, I haven't had any "moody" new servos or receivers for at least 20 years. 

[Martin Harris - Moderator]

I had two SD200s fail out of the box some years ago but I've also known a few cheapos pre-fitted to ARTFs fail.

[leccyflyer]

I had four Topaz servos - similar to those SD200s - bought at a show which were dead on arrival. None of them worked. I opened one up and it was packed with white grease.

Edited July 17, 2023 by leccyflyer
clarification