Issue using specific servo (Futaba S3305) with Atmega 2560 & N-channel mosfet

Other Hardware Motors, Mechanics, Power and CNC
1

Hey! Thanks in advance for helping out-

I'm having strange issues using a Futaba S3305 servo. The strange thing is I've been able to work other servos (a Z270 race servo & an MG995) with this setup but I haven't been able to get it to work with any of my Futaba S3305 servos!

I'm powering the servo from a 5V supply that is separate from the arduino's 5V line. I have made sure that the power supply & the arduino's grounds are tied together. The power supply is not limiting the current because it is set to 5V, 3.5A max & is only drawing ~.04A. I have this circuit setup shown on bildr (http://bildr.org/2012/03/rfp30n06le-arduino/) in combination with the servo mentioned for a project (instead of a motor or light bulb). I am not using a diode. The mosfet I'm currently using is an IRF630- an N-channel mosfet (datasheet here: http://www.jameco.com/Jameco/Products/ProdDS/210534.pdf). Basically I use pin 13 (and a couple of resistors so that the gate only sees ~3.4V) to signal the mosfet to turn on and connect the servo's ground line to turn the servo on and off. This is also (ideally) supposed to prevent the servo from jumping around upon arduino startup/code installation. Pin 12 is outputting the PWM signal for the servo.

When I connect the S3305 to the arduino directly (GND, 5V, pin 12), it runs without any issues, and doesn't even hum. As soon as I add in my mosfet circuit, it just hums and I have no control of it from the serial line. I also noticed that it's fairly easy to turn the servo by hand under these circumstances (maybe even slightly easier than it is to turn the gears when the servo is off).

Any insight would be GREATLY APPRECIATED! Thanks again!

EDIT::
I put an LED onto the circuit and the switch works fine... The LED switches on and off as it should. But when I add the servo onto it in parallel, I noticed that even when pin 13 is low & the transistor is supposed to be off, the LED stays lit (but grows dimmer). This is not the case for other servos- when I plug them in parallel to the LED, when pin 13 is low the LED & servo are both off & can be switched on as expected.

It appears that there is something significantly different causing this behavior on the inside of the Futaba servo.... Maybe my next step will be to take apart the servo..

2

Well, the MOSFET you are using appears to not be a "logic level" one as used in the bildr circuit. The bldr comment "... but you can use any N-Channel MOSFET exactly the same way" is BS. Also if you are switching power on the ground leg of the servo, you might erratic operation or possibly damage the servo.

3

Firstly never power a servo from the Arduino 5V rail, you risk damaging the Arduino.
Servos contain a motor, motors put out inductive spikes, if you're unlucky you could
fry the Arduino.

A nice powerful 6V supply for the servo would be better.

If you use low-side switch for the servo's power then you should add a 5--10k resistor in
series with the signal line to prevent damage to the servo's electronics. You should also
set the signal line to high-impedance (INPUT mode) before powering down servo.

Non-logic-level MOSFET is useless in this context. Using a logic level n-channel MOSFET.

For high-side switching a logic-level p-channel MOSFET could be used, and if the
servo supply is higher than 5V a level-shifting NPN transistor is also needed to
guarantee proper switch-off.

4

Thanks for the feedback!

I ordered the same transistors from Sparkfun & are waiting for them to come in the mail... I'll post an update on how those do afterwards.

Do you guys mind explaining why the non-logic level mosfets don't work? Just for future reference. And why would switching power on the ground leg cause damage? Do you think high-side switching is the way to go? Maybe I shouldn't switch it at all... I was trying to prevent erratic jumping upon startup, but maybe I should just add a mechanical switch in or something instead?