Techniques to Reduce Interference?

[tnavressdog]

I was wondering what steps you guys take to reduce interference on your UAVs. In my case, my 900Mhz radio modem / transceiver seems to affect my R/C receiver drastically. What should I do to reduce or eliminate this problem? Thanks for any help.

[workshop]

Send us a block diagram and some photos so we can see and understand your setup.

Here are the directions I'd look...

Increase physical distance between components
Use individual power supplies
Use optoisolators on signal lines
Use linear regulators rather than switching PS
Position Antennae for minimum pattern overlap
RF Shield R/C System
Clean up wiring

Jeff

[danstrider]

I just had some troubles with an RF modem as well that seemed to be bleeding over into servos. I figure I'll add on Jeff's list a bit...

- copper foil (or just aluminum foil even) wrapped around the servo lines and servos
- triple check have no ground loops, especially going to the modem (this can be a smoking gun if you find one)
- antenna location and orientation between the radio modem and RC receiver plays a big role too .... if you have coax running to move your radio modem antenna away from the receiver, you may try getting better coax or increasing seperation between the coax and RC antenna
- a counter-poise added to the RC receiver can be a cure-all sometimes too. It is a length of wire the same length as an antenna simply tied to receiver ground and run 90deg to the RC antenna. From what a prof at school told me, it helps reduce RF interference getting to the RC antenna and has worked for me before.
- putting the radio modem in a metal box (faraday cage really) can help if it's not the antenna itself
- reducing the transmit power can help sometimes
- if you have a coax or if your 900MHz antenna is running near a servo or near the receiver itself, try adding more seperation or changing the orientation of the parts to one another. Even if this isn't a final solution, at least you'll know why things are interfering.
- ferrite rings or chokes can work
- you can get an opto-iso box from Jomar if the servos have trouble but the receiver is getting signals okay
- if you have an antenna coax bundled with servo wires, that can cause cross-talking
- ground loops really are my first choice to go looking for. I can explain if you need more, but I'd try wikipedia or google for a diagram. Ground loops can put noise on an RC receiver in a heartbeat

Hope this helps!
If you post some photos of your setup, we may be able to help catch some things we missed in our lists. It always helps to have another few sets of eyes!

Dan

[danstrider]

Copper shielding for servos:
http://www.rcgroups.com/forums/attachment.php?attachmentid=970592

[tnavressdog]

Great information guys. I wasn't able to read them until I did my own bit of problem solving but I will definitely use these suggestions in the future. I mostly solved the problem yesterday by moving the receiver away the modem and putting it in a different orientation. I also reduced the transmit power of the modem after remembering the basic rule that you should only use as much power as is necessary. 25 miles range should be plenty... 60 miles was a bit overkill and the power was flooding the system. I flight tested the system yesterday and everything is working perfectly. Thanks for the help guys.

[tnavressdog]

I guess I'll add another piece of advice to this thread just in case someone else solving interference issues can use it. I could use full transmit power and have zero interference by pointing the modem antenna directly at the R/C receiver. As some of you guys already know, the transmit power is weakest in this direction.

And Dan, I heard that adding a counterpoise on the receiver can actually increase the range of the R/C receiver! It's an interesting mod if you want to do some high altitude or long distance flying. It increases sensitivity and will allow the receiver to detect a weaker signal however if someone is on your same channel, they are more likely to affect you. I've always wanted to try the mod for longer range but never needed to yet.

[danstrider]

I was curious tnavressdog, are you using a Piccolo by chance? The 60/25 mile may have given you away.

Good to finally have someone corraborate the counter-poise info; I had never heard another modeler even mention one. I was beginning to wonder if my prof was just blowing smoke, hehe.

Another check I'd be interested to have you comment on if you have a chance. The other day a buddy and I spent an hour trying to figure out the range of our Piccolo setup. We found that with either the transmitter or the airplane on the ground, we got some range X. When simply picking up off the ground to waist height, the range doubled to 2X. When holding the transmitter (antenna collapsed still) and the plane wayyy above head, the range went up to 3-4X. With the Piccolo on, we saw about half the total range of a stand-alone RC system. Perhaps you could comment.

Dan

PS: I think I may have asked you this before, but are you an AUVSI Student UAV Comeptition team member by chance?

[Unterhausen]

Dan,
There was something on the Xbow uNav sourceforge site about range. A person said that they got better range by optoisolating the uNav from their receiver. I can imagine that optoisolators and ferrites might go a long way in a system like these.

[tnavressdog]

Hi Dan,

Actually, I am not using the Piccolo so sorry I can't help you. I have noticed that the range does increase significantly when off the ground with other setups however. As for my current setup, my R/C range check is about the same or maybe slightly less than with the modem off.

I used to be an AUVSI team member with MIT but not anymore. It was definitely fun while it lasted but I've graduated since. If I remember correctly, I am your year. You have a Piccolo now? Are you doing full-time work with the system or just for fun? I'll be in the work force soon but I'm still in school for now.

[Unterhausen]

the XPS guys were talking about how their range increases considerably if they lift the airplane off the ground. That's a 2.4GHz system

[kd7ost]

I've used 900 Mhz at 600 mw AV on board a 72 Mhz controlled aircraft without any problems. Of course mine wasn't a spread spectrum modulation so maybe it doesn't directly relate. I also put a 2 meter APRS system onboard the same aircraft sending packets at 750 mw using an Alinco DJ-S11 radio and a rubber duck antenna and still had no trouble with them working together. (Although the 2 meter packets nailed my BTA AS06 and 07 units) The AV antenna was a rubber duck pointing straight down in a vertical orientation. I used a ground plane under the fuselage to help direct the radiated lobe downward. The 72 Mhz receiver antenna was oriented horizontal along the fuselage axis and was above the ground plane.

Distance for range increases as altitude increases. The radiated energy moves outward and the ground itself causes interference by bouncing the signal that radiates towards it. We fly balloons to 20 miles high and 60 miles downrange with APRS packet traffic at 100mw without missing a packet. Altitude as well as good antenna's and a clean RF cable path gives great range. In fact we hit 2 meter repeaters in Salt Lake City Utah from Boise Idaho once we're up a few miles.

Dan

[deh6]

I heard that adding a counterpoise on the receiver can actually increase the range of the R/C receiver Yes. I found that adding a 1/4 wave ground leg/counterpoise was the only way to get a decent range with a flying wing (servos in wings). The other methods helped, but not enough.

Since that experience, I've done some computer modelling using NEC2 of the antenna and servo/battery wires. Using the input impedance I measured on the Futaba R-168DF (and at 50MHz; 72MHz would be different) I computed the mismatch. The loss for various antenna configurations ranged from 0 - 15db. The 1/4 wave "ground wire" running opposite the antenna wire was one of the better cases. The direction the servo wires run from the receiver has a big effect so one can't just make a flat statement that the counterpoise will help in all cases. It is impossible for the servo wires to be far enough away from the antenna/counterpoise wires to not have some effect, and when they are close, e.g. two or three inches, the effect increases (and shielding is of no help since the shield becomes a factor).

Best would be an antenna tuner/coupler, but lacking that one could experiment with changing the a) the total length (which will have a secondary effect), and b) the ratio of antenna to counterpoise (which affect the reactance). The total length should be less than 1/2 wave (80 inches on 72MHz and 116 inches on 50MHz), and shorter may actually match better. The ratio of "antenna" to counterpoise will have a big effect.

If one looks at the whole installation as antenna (which it is), then it is really an off-center-fed-dipole. One leg is the 39" green wire and the other is the mess of servo and battery wires. The feed point is the receiver inbetween. The radiation resistance is largely dependent on the total length. What varies greatly is that the reactance depends on where it is fed. I measured the input impedance of the R-168DF (50MHz) as 71 -J 118 which means that it expects a lot of reactance in the antenna. Best would be to have a tuner/coupler that is used to trim up the match. That's on the ever increasing list of things to do. Without resorting turning or measurement, one is stuck with "trying things," and adding a counterpoise should be on the list.

[Unterhausen]

This discussiong is good, but it's going right over my head. I'd like more info about a counterpoise. The ones I saw are different than described above.

Mr RCCam has a good idea, use the RSSI signal that probably is present in your receiver to measure receiver gain.
http://www.rc-cam.com/ant_exp.htm

[deh6]

use the RSSI signal that probably is present in your receiver to measure receiver gain. Yes, that is a very good technique. One cannot connect anything to the airplane to make the measurement because any connection affects the antenna system. Mr RC CAM's technique is a simple and correct way to make the measurement. The quick way to get a feel for what is happening, as well as see the effect of different configurations, is the computer modelling, but that only approximates. That needs to be followed by measurement.

I'd like more info about a counterpoise As applied here the counterpoise is merely a wire connected to one of the ground pins of the receiver and run in the opposite direction of the antenna. Or, it could be two wires running down the wings with the antenna running down the fuse.

[jtprouty]

I found that putting sensitive items in electrostatic discharge (ESD) bags (the silver bags that electronic items are shipped in - hard disks, electronic components etc.) greatlly reduces the interference. Data modems were the worst offenders. Put the modem in an ESD bag and try to shield the airframe from the antenna with bags as well. It's light weight and works really well.

One word of caution about the ESD bag - they conduct electricity and can short out any components with exposed circuitry.

Happy flying,

Jimmy

[deh6]

One word of caution about the ESD bag - they conduct electricity and can short out any components with exposed circuitry. Are sure about that? I checked 1/2 dozen of different ones I had around with an ohmmeter on the 20Meg scale and measured no resistance. I even punched through them and tried the edges, thinking that maybe the conducting layer was in between insulating layers.

Some types of these bags are only for tribo-electric protection; protection from generating a static charge due to rubbing, so these would not be effective for shielding. One web site said that for shielding the material would have a resistivity of less than 104 ohms per square.

Seems like a great idea, but it may not work with just any silver colored bag.

[deh6]

Ahah! I cut out a small square (about 1 1/2") of the "silver" bag that my Freescale processors came in. It reads open circuit on the ohmmeter. I placed it in the microwave oven (on small paper plate just in case). I punched the "cook" button and within a second there were all sorts of blue sparks and spitting.

It looks like this stuff is ideal. 1) it won't short the parts, 2) it is rf opaque (well, at least at microwave freqs...)

[jtprouty]

Not sure about all of the bags but I was able to get a reading on several of the bags that I put an ohm meter to. Maybe it's because I punctured down through the outer surface of the bag. Have to try in on some of the more recent bags things have come in.

I had really good results in reducing RFI in UAVs I worked on by using the ESD bags. Also put the autopilot in one to make sure it didn't get hit.

Happy flying,

Jimmy

[deh6]

Another experiment--How good is the electrostatic bag material shielding?

Plan: measure the RSSI voltage of a Futaba R-168 receiver on the bench, with the transmitter close-by (ant collaped), then wrap the transmitter up with the electrostatic bag material and measure the RSSI. Repeat using aluminum foil.

With the transmitter unwrapped the RSSI was virtually maxed out (the dbm versus RSSI curve gets rather flat near max) and somewhat lower when wrapped, so I moved the transmitter into the hall (about 12 feet away) to put the RSSI in a better range.

With the electrostatic bag material wrapped around the transmitter, ends folded and taped I got:
RSSI indicates a signal strength of -60 dbm unwrapped, -95dbm wrapped, so the loss by wrapping was about 45db. Pretty good.

Next, I tried the aluminum foil--
Unwrapped and transmitter a few feet away, the RSSI was maxed out.
Wrapped (in the same manner, ends folded & taped), and still only few feet away, the RSSI was at minimum. Well over 80db.

It looks like the aluminum foil is much better than the eletrostatic bag material (and obviously if completely sealed there would be no signal). However, the aluminum foil is definitely conductive! I keep trying to get some measure of DC conductivity with the electrostatic bag material, but I get nothing (I even tried the little piece that got "fried" in the microwave). The insulating property is a definite plus.

I didn't have enough bag material to double wrap the transmitter, so maybe that would have brought it on par with the aluminum foil.

[fly_asmara]

Hi Guys!

Boy! Am I glad you guys are talking about this!! I reached a dead end on this problem early this year after a terrifying experience using MP2028 autopilot with the 24XStream Maxstream radio modem. I have been using the MP2028 standard issue radio modem for awhile (3 years+) and decided one day to upgrade the modem to the 24XStream after hearing a great deal of good things about it. After setting up the new modem, and do a ground test, my team decided to test fly it. How do I set it up? Well, I followed Maxstream motto - just plug and play. The communication during ground test seemed ok. (Feel free to shout to me 'NO!')

So here we go. Full throttle, the UAV is rolling on the runway. Just as the UAV lift off the ground using RC (ie Pilot in Control (PIC) mode), the autopilot suddenly switched to Computer in Command (CIC) on its own!! Yeap guys! On ITS OWN! My heart went to my mouth. I have never ever had a runaway UAV before. For a brief moment, I saw at the corner of my eye the Ground Control Station (GCS) indicated the RC was OFF. The indicator went RC ON again and we regained RC control over the aircraft again. Tension was high. The aircraft was still climbing at this point with engine full throttle. We banked the aircraft to come back and land immediately and then it happened again, PIC switched to CIC automatically, only this time the aircraft was flying straight at us, low altitude and still climbing but on CIC!

The PIC and CIC mode seemed to change intermitently almost a pattern was emerging - like some kind of 'interference'. I really cannot put my finger on the actual cause of this - whether the 24XStream modem was interfering with the autopilot, or whether it was interfering with the RC receiver. The radio modem (enclosed in its original casing) was placed next to the autopilot, and adjacent to the autopilot is the RC receiver. They are all side by side. This was the same set up when I used the standard issue radio modem from Micropilot but I had no problems like this before. Could I be wrong about using 24XStream?

Oh, we managed to land the UAV. No one or property was hurt. The airframe was slightly damaged. We got a small window of opportunity when the aircraft flew infront of us when the PIC held still long enough for us to down the plane. Guys, needless to say that was the first and the last time I tried this experiment. I am not an EMI expert, so I would be so grateful if you could help me explain what could have happened that day. Is it interference? Or do I need to set up the radio or autopilot somewhere for them to work together? I would like to use the 24XStream radios if possible as I've invested a lot of money to get them.

I read the threads with great interest and will take the tips onboard. But if there's someone else who have gone through the same experience but emerged a champion, I like to shake your hand my friend, and would appreciate to hear from you. :).

Thanks guys. Peace.

--
fly_asmara

[danstrider]

I don't have a good answer about the near fly-away, but you might look into an override switch to take control away from the Micropilot in critical situations like that. The override gives the manual RC link an absolute way to take Micropilot out of the critical path getting your stick inputs to the servos. University of Texas Arlington said they had a flyaway once and now they are happy with the override capability they have added. You can find more info about doing a homebrew system or skip to near the end of this thread and read about the one from Jay Francis
http://www.rcgroups.com/forums/showthread.php?t=360746

Let us know how it turns out!
Dan

[jtprouty]

I was also using an MP2028 when I came across the idea of using the ESD bags. We were having problems with the autopilot going into autonomous mode on its own. The bags took care of several RF related issues but I couldn't figure out why the autopilot was kicking in. We put the R/C unit (Airtronics Stylus) on an O-scope and found that in PCM mode it drops frames which, although it's not a problem for the PCM receiver, the autopilot sees it as losing the signal from the TX and switches to autopilot control. We switched to a JR 9303 and the problem was solved.

I really like the M2028 and think their latest upgrades are really great. I'm in the process of developing some smaller UAVs, both electric and gas powered, and hope to be able to use the MP2028 again, funds permitting.

Happy flying,

Jimmy

[Unterhausen]

probably not a good idea to have the computer deciding if it's in control or not. That's the way mine works though. Trying to fix that.

[fly_asmara]

Hi Guys,

Thanks for your replies. Hey Dan, yeah, the irony of the situation was that I got an overide switch I made myself those days during tuning the autopilot, tested and all but didn't consider using it for the radio modem test. Totally underestimated the problem that might have caused by simply switching to a new radio modem. Will keep your tip in mind. Jimmy, I'm using a Futaba T9CAP and setting it up to work with MP2028 was interesting those days. But once it works, it works :). I think that's probably what happened that day during the radio modem test - somehow the autopilot thinks that there is a lost in RC tx signal and somehow it is caused by the radio modem. Hence, the CIC kicks in. Unterhausen, nope its NOT a good idea to have the computer deciding to take control of the aircraft especially when you don't want it to. Good luck with your project there.

Will have a go with this problem again next month with a smaller plane. Thanks guys.

--
fly_asmara