Hobby Servo Current Limiting

[skeel]

Hi all,

Trying to figure out a cost effective solution to limiting the current into a number of hobby grade servos.
The servos are controlled via a mini Maestro and powered by 7.5V-15A DC supply, which allows for 1.5A draw for each of the 8 servos if required, plus an additional 3A headroom. After one of said servos burnt out, I came to the realisation that I probably needed some way to limit the current draw of my servos. I'm looking at simply adding a linear/switching voltage regulator for each servo, but looking any guidance as to alternative ways to limit current and maintain a constant voltage?

Cheers

[bson]

Why not a resistor?

[skeel]

Why not a resistor?

Any chance you could elaborate?

[Ian.M]

1.6A polyfuse per servo in the individual power feeds?   The Mini Maestro already has a 220R protection resistor in each servo's signal line so that should be all you need.   

[Kilrah]

1.6A polyfuse per servo in the individual power feeds? 

+1.

And check your servos, you might be simply asking them too much. Or they're crap Chinese ones and a few defective ones aren't surprising.

[bson]

Any chance you could elaborate?

Well, I asked because an inline resistor on the supply wire is the obvious answer if there isn't one already.  If there is one and the servo is overloading, then the polyfuse suggested is the next obvious reply.  The problem with a polyfuse in a model aircraft of course is if it trips the aircraft will crash; crashing the aircraft to save a servo seems... less than ideal.

Presumably these are 3-wire PX4 or other PWM servos, with 4-5V or so supply, GND, and a PWM control signal that runs ~3-3.5V, at 30-50Hz.  If so, check the supply voltage; if it's over 5V throw in a series diode, like a 1N4148 (or maybe beefier like 1N400x) to see what happens if you drop it a bit.  The supply might just be a little too hot for these particular servos.

[skeel]

I'll probably go with some polyfuses; i'm not using them for model aircraft so it should be fine (Its a mechatronic chordophone). I'm surprised I haven't been taught anything about them! Thanks for the help!

[Psi]

After one of said servos burnt out, I came to the realisation that I probably needed some way to limit the current draw of my servos.

If your servos are burning out in normal operation then limiting the current is probably not the correct way to fix the problem. The servo should be able to have unlimited current available and not burn out.

Keep in mind that a lot of cheap RC servos are not designed for continuous operation at full load. They're for plane control surfaces where they spend most of their time static and only having to counteract the wind trying to force the surface flat.

[skeel]

They are Protek 170T (160USD) so not cheap servos. Although manufacturers do not provide detail on stall current I've found that ball park figure for some similar servos is around 4A+. I suspect that drawing 4A for an extended period of time (I'm not using batteries) is the reason for the circuitry in the servo to have literally melted. Please correct me if this wouldn't be the case?

[Mechatrommer]

or just maybe, your servo is rated for 5V use which is too much if you supply from 7.5V...

[Kilrah]

They spec it at 7.4V so it should be OK.

But you should probably review your operation mode... RC servos are indeed not intended to be used stalled. If a polyfuse is needed then while your servo will be protected your [thing] simply won't work as it will trip the fuse and stop...

You'd want to set your end points properly so the servo doesn't try to go further than where the minimal force needed to drive whatever mechanical device you're actuating and avoid forcing more than necessary.

Hard to say more without extra info about the application.

[Ian.M]

Protek RC appears to be a retail brand name owned by AMain Performance Hobbies.  There are no datasheets for their servo range.  See  http://www.amainhobbies.com/rc-cars-trucks/protek-rc-170t-chad-bradley-team-edition-high-torque-servo-high-voltage-ptk-170t/p265048 for what little data is available.

The reviews are quite interesting - a number of users are reporting premature failures.
As is usual with OEM rebranded RC servos, the specs suck, as there is no data for current draw vs load, continuous or peak power rating or even operating temperature range. Edit: Its rated for -10 to +60 deg C, but no info on wheter that's ambient or case temperature.  Nor do they give any information on what it does if the control pulsetrain ceases.  However running it at the max voltage rating isn't very smart - I certainly wouldn't want to exceed 7V long-term.

If you are burning these servos up in a mechatronic application, IMHO you'll need to add temperature sensors to each, (they're metal case so an external sensor with some foam insulation so it reads case temperature not air temperature should do nicely), log the temperatures, and shut off all servo and motor power if either an excessive rate of temperature rise or an excessive peak temperature is detected.   I'd contact AMain and ask if they can tell you the max rated current draw and case temperature - but don't hold your breath waiting for answers!

[Psi]

They spec it at 7.4V so it should be OK.

China spec, "It seemed to move ok when i supplied 7.4V but when i tried 7.5V it smoked"

[Kilrah]

Nah, if it's rated at 7.4V it's normally that it's an "HV" type servo which is usable with a 2s Li-Po RX pack (i.e. 7.4V nominal) but obviously should also hold up to the 8.4V of a fully charged battery.

[Ian.M]

From the max torque, and unloaded speed it is possible to estimate the mechanical output power, assuming it can provide half the torque at half the speed.   I calculate that as 10.6W.  Assuming overall efficiency of 60%,  that would be 2.4A @7.4V, and the servo would be dissipating 7W as heat.     A 10cm black anodised aluminium cube has a thermal resistance to ambient in free air of 2.2 deg C/W.  Comparing surface areas, I estimate the servo will have a thermal resistance of not less than 20.5 deg C/W, and it is obvious that it will burn up if operated continuously  (143 deg C temperature rise @2.4A).

Limiting the current to 1.5A is *NOT* enough to prevent burnup.  If its stalled, that's 11W of dissipation, and its only good for about 2.4W continuous for a 50 deg C temperature rise.