Choosing a Good Servo

[Nigel Heather]

I get what you need, you want a servo that has the torque needed for the model, one that centres well, does not have any slop, has bearings and gears that don't wear easily/quickly.

 

But how do you choose that - is it really the case that budget servos are all poor and you have to buy expensive branded servos.

 

Also what are the thoughts on metal vs plastic gears.  You'd naturally think that metal were better but I've seen comments saying that metal wears quicker than plastic.

 

Cheers,

 

Nigel

[john stones 1 - Moderator]

"Facts" get used to suit viewpoints, I use plenty MG servos, not replaced gears once, plastics have been replaced, but only off crashed models.

Minefield choosing servos because there are so many now, choose one then google it for reviews or ask on a forum, other than that it's choose according to specs and trusting the seller.

I've usually gone for popular brands and rarely been let down, never done a gear via catching a surface (hanger rash) and I'm clumsy.

Lastly, someone who's only ever used standards isn't the best person to ask, which servo ? answers usually "Standards".

[john stones 1 - Moderator]

Have a look at the ready to fly stuff available to buy, some feisty stuff on offer, how many have standard servos ?

 

[Peter Jenkins]

Nigel, check these out.  They are listed at 46 g so a bit lighter than the standard but almost twice as heavy as a mini.  They are currently on sale from Bondaero.  

[Nigel R]

14 hours ago, Nigel Heather said:

Also what are the thoughts on metal vs plastic gears

 

Metal gear servos tend to cost a bit more and make more noise. Not sure I've noticed much other difference...

 

14 hours ago, Nigel Heather said:

you want a servo that has the torque needed for the model, one that centres well, does not have any slop, has bearings and gears that don't wear easily/quickly

 

Don't forget speed of movement and servo weight...

 

[Jon H]

When i was in my teens i couldnt resist a bargain and bought 20 no name mini servos off ebay for about 12 quid direct from china. My word they were rubbish and are the only servo failures i have ever had on fixed wing. Fortunately they only controlled things like retracts but still, we wont be doing that again. 

 

If i had the choice between a 6kg non digital plastic gear futaba or savox servo at £20 or £30, and a 10 or even 15kg metal gear digital turnigy or unknown brand servo at £10 i would absolutely take the savox/futaba. To use the example Peter posted, i would take a savox 252 over that unknown random brand any day of the week. The specs are near identical and so is the price. I just wouldnt risk the unknown brand. There is also a tendency to choose overkill servo specs for no reason, and then buy cheap brand servos as high spec ones from better brands are too expensive. A better bet is to buy the higher quality servo, even if the spec on paper is lower for the same price. 

 

The other thing to consider with some of the cheaper servos is you may find they lied about the spec anyway (a bit like high c lipo's), and/or you might encounter another issue. last time i bought cheap servos from ripmax i sent them back as it was possible to send them into a spasm if you let the stick go and it twanged back to centre. They also had ghastly resolution and very poor precision, to the point where you could almost work out how many teeth were on the output gear by watching the servo move. I only needed them for a wot4xl to test an engine, but they sucked so badly i just sent them back and fitted 148's i had kicking around in a drawer. 

 

So when it comes to servos i would be reluctant to skimp on quality. The life of the whole model may depend on one servo and so i would always favour a trusted brand with a known track record. More importantly, the life of a club mate or innocent bystander may also depend on the reliability of one servo. a failure can lead to the model becoming totally uncontrollable and going more or less wherever it wants. while its not impossible for a quality servo to fail, you are stacking the deck in your favour by using the higher quality gear in the first place. 

 

With model crashes now more heavily scrutinised by people like CAA and AAIB i would also not want to run the risk of being accused of using sub standard equipment in the event of an incident, so i will be sticking to my known brands of futaba, savox and hitec. 

 

 

[Brian Cooper]

Choosing servos is similar to choosing tyres for your car. . A tyre is round and black but that is where the game begins. . It is the only contact you have with the road so buying cheap may give poor handling and might cost you dearly. 

 

Similarly, servos come in a plastic box and have an output arm on the top. . But, again, this is where the game begins.

A flying surface is the only contact we have with the air, and if the servo isn't up to the task, the control response will be poor. 

 

If you are seeking crisp responses in a high performance application, the servos won't be found in the bargain basement. 

High performance, in whatever sphere of life, always costs more. 

 

There are some who will probably maintain it's perfectly acceptable to use a dull Futaba S148 in a 8ft span warbird. . . It isn't.  

 

Pay particular attention to the surfaces downwind of the propeller (rudder and elevators). Those  servos come in for some heavy duty abuse to overcome the force of air being blasted over them.  

 

It's horses for courses. 

A basic trainers doesn't require too much servo power, but models with higher expectations need better servos. . Those who disagree will have probably never experienced flying with better servos. 

 

Powerful, fast and accurate servos can transform the handling of a model. 

Don't be timid.... give them a try. 

[John Lee]

15 hours ago, Nigel Heather said:

Also what are the thoughts on metal vs plastic gears. 

 

One of the Members of my Club was UK F3A (Aerobatics) Champion & represented the Country at the World Champs a few years ago. I was a bit surprised that his model had servos with plastic gears (high end JR stuff, about £80+ a pop) as I assumed that metal would be better. He explained that metal gears inevitably had some backlash then wore & developed more slop which he could detect within a month (he flew at least 5 flights a day, 5 days a week for practice), whereas the flexibility of the plastic gears meant that they took up the backlash and stayed accurate for much longer, though he still periodically replaced the gear train.

 

As a mere sports flying mortal this was beyond my needs & largely irrelevant to my flying until I bought a Sebart Su29. That has big elevators with a large throw for 3D flying and much reduced rates for conventional aerobatics. I found I could not fly it dead level when fitted with a new digital Hitec 5625MG servo, it was constantly hunting very slightly due to the normally undetectable slop in the metal gears. A change to a digital JR8231 with plastic gears cured the problem.

 

However this is rather the exception. I generally default to metal geared servos for robustness on any decent model & stick to the main brands from reputable suppliers as mentioned by others.

[Nigel Heather]

I get that there are super duper servos cost £50 or more each bit I'm not looking to spend that sort of money.

 

What I'm trying to establish is if I look at a no-name servo for £12 and compare it with the entry-level Futaba for £20 how much difference is there.

 

The no-name has digital electronics, metal gears and ball-raced shaft.

 

The entry-level Futaba has analog electronics, plastic gears and bushed shaft.

 

Which is more accurate, which is more robust, which will wear better, how much does the Futaba sticker cost.

Edited October 3, 2023 by Nigel Heather

[Peter Jenkins]

9 minutes ago, John Lee said:

 

One of the Members of my Club was UK F3A (Aerobatics) Champion & represented the Country at the World Champs a few years ago. I was a bit surprised that his model had servos with plastic gears (high end JR stuff, about £80+ a pop) as I assumed that metal would be better. He explained that metal gears inevitably had some backlash then wore & developed more slop which he could detect within a month (he flew at least 5 flights a day, 5 days a week for practice), whereas the flexibility of the plastic gears meant that they took up the backlash and stayed accurate for much longer, though he still periodically replaced the gear train.

 

As a mere sports flying mortal this was beyond my needs & largely irrelevant to my flying until I bought a Sebart Su29. That has big elevators with a large throw for 3D flying and much reduced rates for conventional aerobatics. I found I could not fly it dead level when fitted with a new digital Hitec 5625MG servo, it was constantly hunting very slightly due to the normally undetectable slop in the metal gears. A change to a digital JR8231 with plastic gears cured the problem.

 

However this is rather the exception. I generally default to metal geared servos for robustness on any decent model & stick to the main brands from reputable suppliers as mentioned by others.

Funnily enough, it was Hitec 5625 servos that caused the problem on my 2 m aerobatic aircraft!

 

These days, there are an increasing number of metal geared servos that maintain their resolution over time.  Perhaps they are using a different alloy for the gears?  There are a number of reasons that most club pilots don't notice the poor centreing of their servos.  Apart from the obvious one that they are still relatively inexperienced and are just struggling to fly the aircraft, the other major reason seems to be a failure to trim the aircraft properly. 

 

I have been handed models to try out that have been trimmed and immediately have to re-trim them to the amazement of the owner.  I think it comes down to the ability to see what the aircraft is actually doing.  The other reason linked to this issue is that they very rarely fly in a straight line for more than the briefest of moments.  If you try and fly in a straight line without climbing or diving for the length of the patch you will usually see an out of trim situation.  Even when aerobatics are flown, they are usually very small figures and there is a lot going on such as wings not level when loops are started to there are other issues beyond trim that occupy the pilots mind. 

 

Flying with the wings level is not so easy!  It may look like wings level but to me the wing is banked.  Colleagues say to me, how do you know the wings are level?  The answer is that if you pull up into a loop and the aircraft goes to the left you had left bank on and vice versa.  If you pull up and the aircraft goes straight, always assuming you have set the correct side thrust, then you have the wings level.

 

Servos that do not centre accurately are also responsible for these characteristics.  So, sometimes its difficult to apportion blame correctly and fix the problem.   It could also be a combination of things - and usually is.

[Peter Jenkins]

12 minutes ago, Nigel Heather said:

I get that there are super duper servos cost £50 or more each bit I'm not looking to spend that sort of money.

 

What I'm trying to establish is if I look at a no-name servo for £12 and compare it with the entry-level Futaba for £20 how much difference is there.

 

The no-name has digital electronics, metal gears and ball-raced shaft.

 

The entry-level Futaba has analog electronics, plastic gears and bushed shaft.

 

Which is more accurate, which is more robust, which will wear better, how much does the Futaba sticker cost.

All good questions Nigel, but there will be some cheapo servos where all the tolerances are at the tight end and it's a wonderfully performing servo.  Then you get the other one and it's rubbish.  At least with a "name" servo, if you get a problem you can send it back and get a replacement for the dog you had in the first place.  

 

Plastic gears are fine but when you have a large petrol engine with lots of vibration it may cause problems.  However, metal geared servos may develop more slop with a lot of vibration.  There is, I'm afraid, no dead cert way of knowing how good the servo that is a cheapo one will operate and last.  On the issue of longevity, you need to fly aircraft a lot more than the average club pilot does to come up to the amount of flying, with the same aircraft, that competition aerobatic pilots do.  The option remains to replace gear trains if you detect wear.  

 

If I ever had a hard arrival(!), I would replace the plastic gear trains of the servos affected to be on the safe side.  It only takes one broken gear tooth to jam the servo and cause mayhem.  

 

Buy cheap and you could be lucky most of the time.  Buy expensive and you are paying for better QA throughout the manufacturing process.  It isn't always a case that a named brand is more expensive for no reason.

[Jon H]

23 minutes ago, Nigel Heather said:

What I'm trying to establish is if I look at a no-name servo for £12 and compare it with the entry-level Futaba for £20

 

Or just buy a branded servo for £12 and be done with it. https://www.modelshopleeds.co.uk/catalog/product_info.php?products_id=66633&gad=1&gclid=CjwKCAjw9-6oBhBaEiwAHv1QvBIdEOxHlRkMy3nV5A7nh5buCqH8-vwqfXEpzGJ2-8jialP1Dop0kRoCsukQAvD_BwE

 

These are fine for sport models (wot 4, acrowot, spacewalker etc) up to about 80 inch and 20-30cc class. If its a dedicated aerobatic model like an edge or extra, then 65-70 inch. WWII fighters 70-90 inch depending on the surface its used on and the specific model. 

 

With the demise of the 148 and hitec 311's being so expensive this savox 351 will be my go to servo for more or less everything with the 352 for a little extra elevator pull, and 252 in mind for things like flaps on my big warbirds as they need a bunch of power to move them. 

[Nigel Heather]

Yes but I take these guarantees with a pinch of salt.

 

Servo fails (branded or no-name), plane crashes, £300 written off.

 

No-Name -possibly tough luck

Branded - £20 servo replaced

 

I doubt any of the big names will accept liability and refund the damage that their faulty servo caused.

[Nigel Heather]

5 minutes ago, Jon - Laser Engines said:

 

Or just buy a branded servo for £12 and be done with it. https://www.modelshopleeds.co.uk/catalog/product_info.php?products_id=66633&gad=1&gclid=CjwKCAjw9-6oBhBaEiwAHv1QvBIdEOxHlRkMy3nV5A7nh5buCqH8-vwqfXEpzGJ2-8jialP1Dop0kRoCsukQAvD_BwE

 

These are fine for sport models (wot 4, acrowot, spacewalker etc) up to about 80 inch and 20-30cc class. If its a dedicated aerobatic model like an edge or extra, then 65-70 inch. WWII fighters 70-90 inch depending on the surface its used on and the specific model. 

 

With the demise of the 148 and hitec 311's being so expensive this savox 351 will be my go to servo for more or less everything with the 352 for a little extra elevator pull, and 252 in mind for things like flaps on my big warbirds as they need a bunch of power to move them. 

Thanks, these look good.

[GrumpyGnome]

The description of those servos says "ideal for throttle servo or parkflyers"........... I'd suggest the Gangster isn't a parkflyer.....

[leccyflyer]

I'd suggest that is a mistake on MSL's webpage, possibly a copy and paste from a much smaller servo's details.

[GrumpyGnome]

You're probably right...... although others (nexus for example) quote the same........

 

So, as Jon said earlier, if an accident is scrutinised carefully, and you fitted a servo advertised as unsuitable for your purpose, where would you stand?

 

(Best get the servo from somewhere whose description is less specific 😀)

Edited October 3, 2023 by GrumpyGnome

[GrumpyGnome]

My point really is that descriptions are often misleading......

[Nigel Heather]

I have looked at the Savox 351 - initially thought that was the answer but as I dig a little deeper I’m not sure.

 

The spec looked good, torque of 4.1, a little slow at 0.17 secs but fine for my ability.

 

But then I realised that this is at 6.0V

 

It took a bit of time finding the spec for 4.8V, the manufacturer site doesn’t even list it but found it on the Servo Store website but only in imperial units so took a bit of conversion.

 

At 4.8V it is quite unimpressive, toque = 3.2 and speed = 0.22

 

So not so sure about that.

 

For the Gangster, in my shortlist I have

 

Corona 339, size = mini, weight = 32g, torque = 4.4, speed = 0.15

Savox 351, size = standard, weight = 41g, torque = 3.2, speed = 0.22

 

Running at 4.8V, Corona has metal gears, Savox has plastic, both digital, both ball-raced

 

Think I would need to look at the Savox 352

 

Cheers,

 

Nigel

Edited October 3, 2023 by Nigel Heather

[Peter Jenkins]

Don't worry about servo speed for precision aerobatics.  It's the heli boys who need it for their applications.  They are going to react faster than you can and rapid movement is only needed for snap, or flick, rolls.  Even then, the Savox speed would be fine.

 

The key question for the Savox 351 is will it run on 6 v.  JR used to be strictly 4.8 v only and 6 v, 5 cell NiMH, would fry them.  If they are not rated for 6 v you will be limited to a 4 cell NiMH or turn down the supply from the BEC if you go that route.

 

Torque, again, at 3.2 kg.cm is perfectly adequate for the Gangster.  On my 2 m I've used mjni servos rated at 4.2 kg.cm and they were fine.  So for the Gangster this shouldn't pose a problem.  Futaba S148s were rated at 3 kg.cm IIRC and they would have been fitted, and probably still are, back in the day.

 

I prefer a separate battery supply from the main flight pack and use a Powerbox Digiswitch to provide a controlled voltage.  I use a 2S LiPo and set the switch for 7.4 v.  I have also used the 6 v settkng whdn using JR servos as a fully charged 4 cell NiMH provides almost 7v when fresh off charge.  I also use 2S LiFe packs for 4.8 v servos as the LiFe voltage drops quickly from the max 7.2 to 6.9 v pretty quickly and again have proved not to damage JR 4.8 v rated servos.

[Jon H]

1 hour ago, Nigel Heather said:

I have looked at the Savox 351 - initially thought that was the answer but as I dig a little deeper I’m not sure.

 

The spec looked good, torque of 4.1, a little slow at 0.17 secs but fine for my ability.

 

But then I realised that this is at 6.0V

 

It took a bit of time finding the spec for 4.8V, the manufacturer site doesn’t even list it but found it on the Servo Store website but only in imperial units so took a bit of conversion.

 

At 4.8V it is quite unimpressive, toque = 3.2 and speed = 0.22

 

So not so sure about that.

 

For the Gangster, in my shortlist I have

 

Corona 339, size = mini, weight = 32g, torque = 4.4, speed = 0.15

Savox 351, size = standard, weight = 41g, torque = 3.2, speed = 0.22

 

Running at 4.8V, Corona has metal gears, Savox has plastic, both digital, both ball-raced

 

Think I would need to look at the Savox 352

 

Cheers,

 

Nigel

 

You are overthinking this mate. Its a gangster, just fit the 351's and fly it like you stole it. It will be more than adequate. 

 

 

 

 

[Nigel Heather]

1 hour ago, Peter Jenkins said:

 

 

The key question for the Savox 351 is will it run on 6 v.  JR used to be strictly 4.8 v only and 6 v, 5 cell NiMH, would fry them.  If they are not rated for 6 v you will be limited to a 4 cell NiMH or turn down the supply from the BEC if you go that route.

 

 

Yes they will - in fact that is how Savox quote them.  The thing is that BECs on the simpler ESCs only do 4.8V, I could run a separate UBEC at 6V but would prefer not to do that.

[Bruce Collinson]

I stick to Savox on the basis that they seem to be good quality, voltage isn’t an issue and I reckon they will be re-cyclable for some time to come.   And they’re not Chinese.   I suspect equivalent Hitec are probably as good as.   Buy once, buy well.

 

BTC

[GrumpyGnome]

Nope. Manufactured in Taiwan.

[Bruce Collinson]

Yes, Taiwan.   I have an adequate and robust response to your implication (or my inference) but it’ll get deleted for encroaching on geo-politics.