I wasn't convinced brushless servos are better - so I bought one and made a video

[LMF5000 54]

In one of my other threads, someone had suggested investing in a good servo to complement a high-end build. I had already been eyeing the entry-level brushless servos on tamico. They're 3x the price of a good (waterproof, metal gear) digital servo so I was hesitant, but that discussion convinced me to take the plunge.

When I was doing my research, I couldn't find any really good quality side-by-side comparisons online, so after buying I decided to set up a side-by-side comparison video to help others make a more informed decision. As a mechanical engineer I tried to set up the experiment to eliminate as many variables as possible, so both servos are being driven from a Y-harness on one channel so they're getting exactly the same signal at all times. I used a receiver instead of a servo tester to reflect real-world conditions. If you keep the sound on, at the end you can also compare the noise made by the two servos individually.

Here's the video. First 15 seconds compare the speed, next 20 seconds compare power consumption:

 

[njmlondon 805]

That’s awesome to see. I am quite impressed with the speed of response of the brushless. 
 

[Wooders28 5120]

1 hour ago, LMF5000 said:

I couldn't find any really good quality side-by-side comparisons online

I may have set my youtube vid to, link only and not public...🤦‍♂️

Pretty much the same though. I bought an Altrun race servo for a Novafox build, and the specs said it was pretty much the same as my Futaba Brushless servo, but for a 1/5 of the price!! 

Urm..

 

Although, Futaba say its metal geared, but with mine being a low profile servo, the gears are thinner, and managed to rip some gears off, in a crash on the straight (other car crashed on another part of the track, and cartwheeled across the straight, I was flat-out, so not a tiny incident).

 

[Sgt.Speirs 1624]

@LMF5000: Only a question from my side: Does it make sense to compare servos with different specs/speeds?

In the YT video description you tested a "Absima Digital Brushless Servo LP15DBF HV" and a "Amewi Digital Servo DC5821LV":

Absima Digital Brushless Servo LP15DBF HV   VS.   Amewi Digital Servo DC5821LV (specs from Tamico's homepage)

Speed @ 4.8V:  0.082 sec/60°    VS.    0.180 sec/60°
Speed @ 6.0V:  0.073 sec/60°    VS.    0.160 sec/60°

Don't get me wrong, but for me you have tested a "fast" brushless servo against a "slow" brushed servos...with a result which is not surprising  

[LMF5000 54]

2 hours ago, Sgt.Speirs said:

@LMF5000: Only a question from my side: Does it make sense to compare servos with different specs/speeds?

In the YT video description you tested a "Absima Digital Brushless Servo LP15DBF HV" and a "Amewi Digital Servo DC5821LV":

Absima Digital Brushless Servo LP15DBF HV   VS.   Amewi Digital Servo DC5821LV (specs from Tamico's homepage)

Speed @ 4.8V:  0.082 sec/60°    VS.    0.180 sec/60°
Speed @ 6.0V:  0.073 sec/60°    VS.    0.160 sec/60°

Don't get me wrong, but for me you have tested a "fast" brushless servo against a "slow" brushed servos...with a result which is not surprising  

That is a very good point, and one that I was wondering about myself. It appears some (not all) coreless servos can also achieve similar speed (example) for as low as €28. But I couldn't find a conventional (brushed) servo that could manage 0.08s/60°, at least on tamico. So I'm thinking maybe it's an innate advantage of brushless that they can achieve that kind of speed more easily than conventional ones (maybe less rotating mass since a brushless just has to spin the inrunner motor magnet, a coreless just has to spin the copper wire, but a conventional servo has to spin the entire armature).

I'm not discounting that I might have missed something - have you found any similarly fast conventional servos in your own searching?

[Wooders28 5120]

3 hours ago, Sgt.Speirs said:

Does it make sense to compare servos with different specs/speeds?

Mine was an Altrun Race servo -

Speed @ 6V - 0.09s/60°

Vs 

Futaba BLS551-

Speed @ 6V - 0.10 sec/60°

 

There should be visibly no difference, with the Altrun specs saying quicker by 0.01 secs...

 

[ad456 551]

I’d be interested in a full shoot out on servos of all 3 types with speeds, weights, amp draw and stall amp draw. The last of those could be very surprising. 

[Wooders28 5120]

1 hour ago, ad456 said:

The last of those could be very surprising. 

Think that's why 'brown out' is a thing with Savox servos, they're quite power hungry,  and people forget to check the BEC current rating of their esc before purchasing one (plus not setting end points). Most are only 3Amp.

Not sure if HV servos require less amps, given they're higher voltage running (need a suitable esc to supply HV BEC right enough).

[ad456 551]

@Wooders28, I suspected the same with Savox as the endless issues and jittering folks had seemed that kind of thing. 
 

Is this a possibility? I’m no math wizz with electrics so please tell me if/why I’m wrong. 
 

Say I’ve got a HV 8.4v servo direct wired to a 2s along with a 60a ESC. I manage to stall the servo somehow while using all of the esc as well(Yeah, not a common thing I assume but this is a “for fun” scenario). So that would add x amps to the already 60a, say x is 10a then wouldn’t that be 70a on the battery? If I had tightly specced for the esc only would I maybe damage my battery cells from over stressing them. I’m not crying Lipo fire here, just weakening which lowers the punch and maybe the life of them. And this is also just for fun with theory. 
 

[LMF5000 54]

2 minutes ago, ad456 said:

@Wooders28, I suspected the same with Savox as the endless issues and jittering folks had seemed that kind of thing. 
 

Is this a possibility? I’m no math wizz with electrics so please tell me if/why I’m wrong. 
 

Say I’ve got a HV 8.4v servo direct wired to a 2s along with a 60a ESC. I manage to stall the servo somehow while using all of the esc as well(Yeah, not a common thing I assume but this is a “for fun” scenario). So that would add x amps to the already 60a, say x is 10a then wouldn’t that be 70a on the battery? If I had tightly specced for the esc only would I maybe damage my battery cells from over stressing them. I’m not crying Lipo fire here, just weakening which lowers the punch and maybe the life of them. And this is also just for fun with theory. 
 

My servos, not stalled but moving at full speed, were drawing about 0.5A. If most BECs can only supply 3A, then I'm going to guess most servos won't draw anywhere near 3A when stalled.

Let's take a pessmistic case where you only had a tiny 1000mAh lipo with a conservative 30C rating - that's 30A of max current, so worst case a stalled servo will use 10% of the battery's capability. If you have the more typical case of a 3000mAh or larger battery with a 40C rating, that's 120A max current, so the worst-case servo is only 2.5% of the battery's current capability.

In short, in my opinion (and speaking as a real-world engineer), it's not worth thinking about. Plus if you're accelerating at full throttle your steering will likely be mostly straight so servo load will be very very little.

[ad456 551]

Cool, kinda seemed like I was overthinking but sometimes it worth an ask👍

[alvinlwh 8001]

@ad456 that is kind of what I do with my TEU-101/4, plugging the red batt plug directly into the RX. These ESCs do not have a BEC and feed full fat battery down the batt feed, in the past requiring a BEC RX. Of course a 8.4v-able servo and RX will be required. 

The thing about stall/higher amp is often misunderstood, many thinking it only happens when the servo is turning and draws little when going straight. This is absolutely not true as when going straight, the steering will be subjected to all kind of force, and the servo will have to keep the steering in the exact straight position (say 1500, TX range 1000 - 2000). Forces (bumps, acceleration, deceleration) will try to push it out of 1500 and the servo will have to use power to keep it there. So a 30kg servo will be using up to that much force to hold the car straight. Just try to force a powered on servo, see how hard it is to overpower. Electricity will be needed to resist your force. This can draw as much as the stall amps even when going straight. In fact I noticed most brown outs on the TBLEs when accelating straight. 

[ad456 551]

Yeah a servo is always going to small adjustments indeed, nearly always trying to be where it should be.

I never considered it would go to full stall current unless it was actually bound up. 

[alvinlwh 8001]

12 minutes ago, ad456 said:

I never considered it would go to full stall current unless it was actually bound up. 

It does not need to go into full stall to cause a brown out. Using the old calculation of 3kg/Amp (I know there are more energy efficient ones now), a 15kg servo will cause a brown out on a TBLE at just 5kg of force. 

[ad456 551]

I’m was thinking the max that the servo could draw in a complete jam.
 

For fun I went to the shed and stuck an eBay special MG995 in the vice with a solid horn on it then hooked it up to a power supply and a servo tester. It took 2.8a at 6v when it topped out so that would totally track with your equation👍. It didn’t rise to that again and I might have felt the cheeky scent of brushes being burned👌. I did a couple of others from the same bag and they came out just under at 2.72 and 2.69. None were run in as such.

I rattled one around in my Cc-01 and bound up the steering(it does have a saver on it) and it would only draw 1.08a when fighting me a bit and around 0.8a when randomly changing direction quickly under no stress.  

Im gonna say what I did was hardly scientific by any stretch but I found it a bit interesting.

[alvinlwh 8001]

11 minutes ago, ad456 said:

Im gonna say what I did was hardly scientific by any stretch but I found it a bit interesting.

What you did is more scientific than my simple test on a SkyRC ESC programmer and Mk. 1 Eyeball observations.