Servo Interference from 3DR Telemetry Module

[x0054]

I have a 2.4GHz FrSky transmitter and receiver combo and a 915MHz telemetry module from 3DR. After plugging in the telemetry module, I noticed a twitch in the tail servo on my Tricopter. It twitches every 1 second or so in a little pattern. like T-T-T--T--T-T-T--T... T stands for twitch, wasn't sure how else to represent it

Removing the antenna from the module fixes the problem, so it's clearly radio interference. Any ideas? Is this a common problem? Maybe problem with the antenna? Or did I just get a bad telemetry module?

Any ideas. Thank You!!!

 

[Craftydan]

How long are your servo leads? how close are they to the antenna?

you're probably battling ElectroMagnetic Interference (EMI) on the servo's signal wire. effectively, the servo's signal line is acting like an antenna and the data pulses are mixing in with the control board's pulses to confuse the commands. I haven't had a servo do it, but I've had a cheap ESC do it before.

If you've got a ferrite:



a few wraps around should cut out the EMI. if it doesn't, the EMI is attacking your receiver or control board, and the servo shouldn't be the only thing twitching

 

[Cyberdactyl]

If you don't want to disassemble your servo wires and/or remove them from your copter, you can also get some reduction in interference by taking a stiff, single strand low gauge, thick copper wire ~10ga. . . cutting a length of about an inch and shaping it into an open ring. Wrap your servo lead around the ring around 4 or 5 times and clamp the circle closed. Or find two small ~1/4" inner dia. thick washers. Cut a thin slot into both washers, just enough for the wire to slip in, align them, wrap 4 times, then rotate the washers to close the gap.

It won't work as well as a thick ferrite ring, but will choke down the EMI a bit and maybe help make a difference.

 

[x0054]

Thanks! I am going to go dive into my pile of electronic scrap, see if I can find something to pull a ferrite out of.

 

[x0054]

So I found a ferrite in one of my old routers and pulled it out. It was big enough to wrap the servo wires through it 3 times. I tried placing the ferrite next to the receiver and next to the server. No luck, it did not improve the twitching. I also have an ESC mounted to the same receiver, and at the moment I do not see any twitching from the ESC, so it must be just the servo.

I did notice that wrapping the servo wire around a screwdriver and then touching the ground pin of the antenna fixes the twitching problem, but I am not sure if it's because it cancels the EMI or just reduces the efficiency of the antenna.

Should I look for a bigger ferrite? Try something else? Thank you!

 

[jhitesma]

One other idea you may want to look into is a different antenna. End fed verticals aren't very efficient and can cause what you're experiencing. Switching to a simple dipole may help.

 

[x0054]

I built a dipole antenna and tried it out. It works much better almost no twitching. The only time it twitches not is when I touch the servo wires to the antenna. I think I am going to get a different servo, after all, though. I tried 2 different servos from my other helicopter, and they are not at all effected by the telemetry module. So, it must me that the servo is just way to sensitive, or poorly shielded, or something.

 

[jhitesma]

Yeah, verticals are a huge compromise. The bummer about the dipole is you don't get the same even coverage pattern but instead more of a bowtie shape with two big nulls that could be a potential issue. But all that RF going to waste on the vertical and messing with your servos is at least being pushed into the ether instead of coming out as noise.

You may want to look into a simple ground plane design if you have the space for it. It's basically a vertical but with a few ground radials coming off the bottom for the signal to "push off" against so less RF is wasted. You may even be able to take the stock vertical and just add some ground radials to get the same effect. With 4 radials you should have a nice even coverage pattern again:



If you're working with full length elements (which at these frequencies I assume you are) then you don't need a coil like this diagram shows.

I've built quite a few of these antennas for 2m/440m operation and am always amazed at how well they perform for how easy they are to build. For 2m/440 I usually just use a chassis mount PL-239 with the driven element soldered directly to the center pin and four ground radials attached to the four mounting holes with small bolts and a touch of solder. Scrap solid copper romex works great for them making them almost free. They're a little tricky to mount in some applications...but if you have the room and a way to mount them they're one of my favorite low cost antenna designs.

Ferrites and shielding can help mask the effects of stray RF. But eliminating that stray RF is always the best strategy since it solves the problem and increases your efficiency putting more power into your signal instead of wasting it as noise.

 

[x0054]

Unfortunately space is at a premium on this model. It's a 12" x 12" tricopter with APM, GPS, Telemetry module, and a GoPro. So, it's packed to the gills at the moment. I think I should heave went with a bigger frame, to be honest. The ground plane looks like an easy build. I think I am going to use that when I build a bigger model next.

The dipole that I built uses 2 20 gauge copper elements soldered to the coaxial cable I scavenged from a WiFi router. I ran the coaxial cable parallel to the ground element and packaged the entire thing inside a plastic tube. Does that sound right? I am not sure if running the coaxial cable parallel to the ground element in the dipole is ok. It works ok at about 250 feet, but I did not get a chance to test it any further.

 

[jhitesma]

It's not ideal, but may be ok. Ideally you'd want the feedline coming off at a 90 degree angle. But I know full well how hard it can be to fit things in and sometimes compromises are necessary. I'd say give it a try and see how it goes, if you don't get the range you need then you may want to look into getting the feedline away from the ground element. Putting a few loops in the feedline to act as a choke may help as well - but again - space.

 

[Steve Fox]

I know this is an old topic but I thought I would add my experience.
I run a 433mhz telem module and I was getting horrific interference from it, the servos went crazy, my APM kept rebooting, I kept loosing gps my osd was jumpy and I had lines in my video!

It seems the shielding in some of these units is very poor and it was transmitting rf interference straight down the data cable and into my flight controller!

I put a ferrite coil on the data cable and it improved lot but pre coil there was still a lot of leakage, carefull placement of the telem module and all cables for the APM was the only way to get the problem completely under control.

Obviously, with a quad its not so easy but being a plane I was fortunate that I had the space to shift things about.