In another thread I mentioned that a sales clerk at the local Ace Hdwr. store suggested modifying a servo for a cheap 5-6 volt motor. He suggested moving the ‘pot’ and setting it in the center position. The result is a continuous run, polarity sensitive motor, since I had 2 on the shelf, I grabbed one and found it fairly easy to do;

Since my application is a cheap motor, it raised questions regarding the wisdom and a link to more powerful editions…and questions!

I don’t have the answers and thought a specific thread might be helpful to the Others.

I googled continuous run servos and saw thousands, I was impressed with the $2,100 beauty, but since I’m not aiming lasers I clicked on …

Thanks for starting this John. I’ll repost my questions…

I looked at that link and am having a hard time telling the, for instance, power usage for these vs a regular motor. Any idea for the watt requirements? I saw 3 amps at 8.4 volts but that seemed to be a maximum and not the normal. And they seem to be on average about 1 rotation/second, or about 50-80 rpm.

I’m wondering what the battery life would be for a servo vs a motor, and also how this fits with existing decoders. Do the current decoders (such as airwire) do pure voltage variation to control speed, or do they do pulse width modulation (which this seems to be.

I’m afraid I’m also a bit confused about why servos take both a VCC (like 5 volts) input and a separate PWM (pulse width modulation) input to control speed. Why do they need both? Anybody know?

Here’s that link …

https://core-electronics.com.au/components/motors/servos/continuous-rotation-servos.html

A couple of answers:

1. Airwire uses PWM through an H bridge for motor control - i.e. the polarity of the PWM signal reverses to reverse the motor direction.

2. A continuous rotation servo has an H bridge built in. To control it, you have a constant DC (+5 V), and a control line. A pulse of 1 ms will run the servo at 100% speed CW, 1.5 ms will stop it, and 2.0 ms will run 100% CCW. The current is drawn through the DC input, so the control line can be interfaced directly with a microcontroller or other low-current device.

3. You could gut the servo and attach leads to just the motor and have a motor with reduction gears that would operate just like a DC motor - e.g. on the output of an AirWire board. BUT, it’s a 5V motor, so you’d have to somehow reduce the PWM output to 5V.

Thanks Eric! Any idea of the power comparison vs a raw motor? Presumably if I gutted the servo and used it as a raw motor there wouldn’t be much difference?

As to #2, I guess I would have to send a pretty much continuous set of 1.5ms pulses to keep the motor from turning, eh?

Here’s a few wikipedia pages: H bridge, general servo, arduino control of continuous servo.

I’m worried that there’s a power drain even when the servo is not moving (1.5ms pulses). Any info on that?

I haven’t actually used a continuous rotation servo for anything, but I have used conventional servos quite a bit. A conventional servo is centered with a 1.5 ms pulse. If you continue sending the pulse, it hold that position with a good deal of strength - handy if it’s connected to a steering mechanism or aileron. In this scenario, it draws current continuously, since it’s constantly making little corrections. If you stop sending pulses, it stops holding position, and you can turn it by hand. It also stops drawing current.

The problem with killing the control line but not the power line is that any noise that is picked up can be misinterpreted as a control signal and the servo will go bonkers.

How exactly this translates to continuous rotation, I’m not sure. What’s happening in continuous rotation is that the servo is hunting for a position that it will never reach.

What I’ve been doing with the servo controlled Kadee couplers is putting an N-channel MOSFET inline with the ground so I can interrupt the power when not in use. When I want to open the coupler, I begin the pulse stream on the control line, and then switch on the MOSFET. The servo goes where I want it, and then I switch it back off. I suspect this approach would work fine for what you’re doing as well.

Like this:

BTW, I’m not sure of the power consumption.

Sure that makes sense (the n-channel MOSFET switch to the power).

BTW: I’m not doing this yet, just filing this away for a possible future bash like John is doing. It may make more sense to just use a regular can motor and an off-the-shelf airwire receiver with integrated H-bridge. Less fiddling. And my electronics is rusty.

Thanks!

Constant rotational servos are designed to be used with ANY regular digital proportional R/C receivers for a lowish cost 5-6 volt combined ESC, motor and gearbox, in a super compact format. I doubt they can be used with any regular PWM output controllers such as AirWire or Revolution unless they are equipped to control servos such as used by Kadee.

All you need is some form of chain drive attached to the output shaft and the axle you wish to drive.

I have tested all three of the servos listed on that link.

All three provide continuously revolving proportional control either side of neutral.

The smallest one revs quite high but is quite smooth. Sadly it would be not very long lasting if used to drive a small rail vehicle.

The middle one tested very well. Nice and smooth either way. Probably 50 -60 rpm. No idea on current drain but I doubt much more than a few hundred milliamps max.

The most expensive one was the least smooth, as it started with a bit of a jerk. It was very powerful.

Be wary of any servo that does not have metal gears as constant revolution is going to wear them out rapidly. Preferably they should have ball bearing shafts as well.

I have seen many different “servo drivers” over the years, these gadgets allow you to control any servo as designed and or a constant rotation servo with a 4.8v-6.0v receiver battery pack by turning a knob. I use servo driver to test servos and set-up and cycle aircraft flight surfaces, throttle and any linkage mechanically to prevent binding. I’ve seen the drivers go for under $10.00 on up.

Michael

Seen 'em listed as “servo tester”.

https://www.amazon.com/dp/B012LZKTDO