I'm working on a mechanical device that uses three DSM44 servos, two FEETECH Mini Servo FT1117M, one JX PDI-6208MG Metal Gear Digital Servo, and one DS04-NFC 360 Degree Continuous Rotation Micro Servo Motor with Arduino controller and 6V 8A power supply to the shield.

The DS04-NFC 360 Degree Continuous Rotation Micro Servo Motor in a mechanical device must be kept distant about 8 meters from the controller.

So, I have to find a wire 8-10 meters long and attach connector male/female plugs with a crimping tool, or maybe I can make this with a short male to female lead servo extension cable wire cords, connected to each other to 8-10 meters length long.

I'm not a specialist in motors, powering, and wiring, but as I know, I have to use some kind of servo cable signal booster for such a task.

Can you guide me please, to understand how correctly calculate and provide such a long wire connection and power from the controller shield to a servo motor and separate devices from each other with the required distance

  • \$\begingroup\$ Pololu's a good source. But, just as a suggestion, while you can get part/product ideas from Amazon/eBay/Alibaba/etc. you should try to find distributors or retailers that provide products that come with datasheets from reputable manufacturers (that have a web site, put their name on the product, don't disappear in three weeks when you try to order more of the same item, etc.) \$\endgroup\$
    – JYelton
    Apr 20, 2022 at 17:37

3 Answers 3


Servo motors are a DC motor with a gearbox and internal closed loop feedback. They require voltage, ground, and a control signal. Based on the duty cycle of the control signal, the servo will move to that position (or speed, in the case of the continuous rotation servo).

Because they will try to move to and maintain a given position, if they encounter too much physical resistance, they can stall and demand high current (called the stall current). You need to be able to provide the high current, but also some protection in case such current is prolonged. In other words: when the motor is working against a heavy load, it will need higher current. If it gets stuck, you will want to have a means to disconnect or stop it (which may be as simple as a fuse or resettable fuse, or a controller that monitors current and cuts out as needed).

The wire you linked is 28 AWG, which is quite small. Larger currents need larger diameter wire which have less resistance and therefore lower voltage drop. If you look at a wire resistance chart such as this, you can see that 28 AWG copper is about 213 mΩ per meter. In an 8 meter length of cable you're effectively adding about 3.4 Ω of resistance. (Remember, both the voltage supply and ground form the loop, so that's 16 meters of wire.)

If all you did was leave the motor out and shorted that end, there would be 1.76 A of current flowing. (I = E/R) The motor may need more than that to perform well.

Increasing the wire gauge to 12 AWG which has a resistance of 5.2 mΩ per meter, reduces the total resistance to 83 mΩ (about a 40-times improvement). Now with a short, the available current is 72 A — more than enough.

Knowing this, you will want to do the following:

  • Use thicker wire (lower AWG #) for less resistance.
  • Keep power wiring as short as possible for low voltage/high current applications.

If possible, have your 6 V power supply much closer to the motor(s) and run separate wire to each motor depending on its needs. For example, if a particular motor needs 1-2 A and you limit the wire length to 1 meter, 28 AWG might be suitable.

Finally, the control signal for servos is very low current, so you don't have to increase its diameter in the same way. Obviously for long distances, you may need to use a thicker wire to impose less resistance as well.

All of this means that the typical 3-wire connectors used for servo motors are not going to accept the wire gauge you will need for long wire lengths. From the power supply to the motor will need to use bigger wire and connectors; basically whatever you want to use if you need to be able to connect/disconnect it. You might consider using commonly-available automotive or RC hobby connectors that are rated for higher current. At the motor connection, you can use the existing wires, but you'll have to decide if you need a connector there as well (connecting two dissimilar wire gauges is a little more troublesome). I would be tempted to remove the connector, strip the wire ends, and solder them together, or use some sort of splicing connector like a terminal block or "lever nut."

  • \$\begingroup\$ The wire resistances you quote are per 1000 metres, not per metre, according to the copper wire tables I checked. \$\endgroup\$ Oct 26, 2022 at 15:22
  • \$\begingroup\$ @PeterBennett Ah you are absolutely correct. I will edit accordingly, thanks for pointing that out. \$\endgroup\$
    – JYelton
    Oct 26, 2022 at 16:41

Your DS04-NFC 360 requires PWM signal, GND and 5V/6V supply voltage (<1A current consumption according to Amazon description).

Generally speaking, for power cables, the only thing that matters is the cable section (see AWG wire current chart online). The bigger the section the more current it will handle. In your case 1A is very small so you don't really have to care.

The other signal is the PWM signal. Problems with digital signal on long distances are both noise (if you have lots of noisy sources like high power motors ... you can twist GND and PWM if this is a problem for you ... see if it helps, or you can use a shielded twisted pair wire with shield connected to GND on both sides ... use that for really noisy environment) and DISTORSION (due to reflection and/or line capacitance : see "characteristic impedance theory" for more information). Basically you could end up with a signal that really doesn't look like a beautiful square PWM signal anymore. Using simulation tools like LTSpice and their non-ideal transmission line model you can see the effect of an 8-meter cable (if you know the cable specs of course) and experiment some strategies, or you can just play safe and use one of those "servo cable signal booster".

  • \$\begingroup\$ also note that those "servo cable signal booster" are nothing but simple high current non inverting buffers (particularity of those buffers : very low output resistor so it takes less time to charge all the line capacitance). Of course it might be easier to use a module than designing a PCB. For line capacitance estimation (btw PWM and GND wires) : you can use this online calculator : emisoftware.com/calculator/wire-pair-capacitance \$\endgroup\$ Apr 20, 2022 at 18:46

Consider that you are powering five 8 ohm speakers that only draw power when accelerated.

DSM44 servos draw 0.7A on a surge from 6V so the load appears to be about 8 Ohms DC max with a locked rotor but can surge with step changes to this value then decay so the spectrum.

If you understand Woofers audio systems need heavy gauge wire from the power source to push a cone fast with very low resistance wiring that is 0.1% of the source impedance or worst case 1% (Damping Factor =100) . That can be measured by the added % drop in voltage at max current.

let dV/dI=Zs the source impedance of the supply then the loss of the wire degrades the supply's load regulation error and raises it's source impedance.

Given 6V 8A with 1% load regulation error thus Zs= 1% of 6V/8A= 7.5 mohm .

Ideally, your power cable has a local supercap or Lithium or SLA battery near the device with a much smaller impedance than this to compensate for the wire resistance. This resistance will degrade the servo performance under repetitive motions. So lower is better with good damping then you will understand performance will suffer on stiffness and overshoot, resonance, if you use skinny wires to drive the motors.

The control signals should be shielded in twisted pairs isolated from these power wires that can carry fast impulse currents that may result in crosstalk to the control signals if you do not take care of this. (CAT 5 is convenient)

AWG 16 is enough if you have a big supercap near the load. 13.17 mΩ/m *8 m ~ 105 mohm which drops the steady state voltage by 0.8V but for short duration motions that may be acceptable or not.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.