RC Scale Aerobatics
Mike Hurley [[email protected]]
The mysteries of RFI become personal for the author
For the 2005 season I completed one of the nicest airplanes of my life: a 40% Carden Extra 330 that was custom-built with a bunch of personal modifications and a completely hand-fabricated carbon-fiber internal structure. This airplane was the pinnacle for me so far. But the year turned out to be wrought with frustration. My beauty was plagued with a mysterious radio-frequency interference (RFI).
RFI problems can be some of the most frustrating in the RC hobby. Failure to fix the problem can be, and often is, disastrous. As you will learn as you read this, I am by no means an expert in dealing with RFI. But hopefully my problem-solving approach may help others solve their interference troubles.
Initially I thought I was getting good range checks. I went by the book and used the method recommended by Horizon Hobby from the company's Web site. You can find it at www.horizonhobby.com/Explore/Article.aspx?ArticleID=1079.
In the article Horizon explains how to set up your Pulse Code Modulation (PCM) radio so that the fail-safe feature is used to determine your system's distance to failure. It works fine, but I was also getting tiny delays in servo operation. These delays could easily be seen when comparing the movement of the elevator halves, each half being operated by a separate servo on a separate channel.
Since the delays could be seen at any transmitter-to-receiver range but the fail-safe did not kick in, I wasn't sure that I was experiencing real RFI. So I called JR's research-and-development manager John Adams and asked him about what I was seeing. He said that indeed the problem was caused by brief radio dropouts and I should not see them once the problem was resolved.
The difficulty in determining the cause of the RFI was exaggerated by the complexity of my airplane's internal features. First of all, I am running full-length tuned pipes, putting a great deal of metal in close proximity to the radio equipment and spanning the full length of the engine and control systems.
Mike's RFI-Problem-Solving Process
- One receiver pack was close to the ignition switch, so it was moved.
- Replaced the ignition module.
- Replaced the ignition switch, battery, and wiring.
- Replaced the spark plugs.
- Checked throttle linkage for loose connection and vibration points.
- Changed throttle linkage.
- Replaced throttle servo.
- Changed the propeller. This could be a twofold problem; propellers on full-scale airplanes have been known to build up an electrical field. Also, specific vibration frequencies could be inducing RFI.
- Changed antenna location (took it out of the airplane).
- Changed receiver location.
- Checked each servo and battery connection going into the receiver, one connection at a time, by unplugging the lead and doing a range check.
- Changed receivers.
- Changed module on the transmitter, and in doing so also changed channels.
- Changed transmitter.
Second, I built custom-made carbon-fiber servo trays and exhaust enclosures. The carbon ran from the very front of the motor box to the aft section of the radio-equipment hatch. This carbon turned out to be highly conductive.
The elevator synchronization delay problem only occurred when the engine was running, so it clearly had to do with either the ignition system or the vibration the engine produced. Fortunately I was able to replicate this problem in my driveway at home, and throughout the many hours of testing I did have a lot of fun getting to know my neighbors. Of course, now they think all RC airplanes span 120 inches and have 150cc gas twins with full-length tuned pipes!
In trying to find the source of the RFI I looked for the usual suspects, such as electronics that were too close to the ignition box and parts of ignition and receiver systems that were too close to each other. I also considered isolating the ignition battery and switch from the rest of the electronics. My throttle servo linkage is metal, so I wondered about possible vibration in the linkage.
Luckily I have many spare parts on hand, so I didn’t incur much cost in replacing components.
The accompanying sidebar contains a summary of how I tried to solve the RFI problem. This might be a good, logical process when looking for problems of your own. I performed each task one at a time, and testing was done before I moved to the next change.
On each occasion no positive result was found, but one of them may be the magic change for your particular problem. As you go, remember to make adjustments in a slow and methodical way and test after each modification for a change in response or range.
I use a single receiver in this airplane, so I wondered if idle vibration might be overloading the receiver from all the servos fighting to stay centered during idle. Therefore, I went to a dual-receiver setup, splitting the model down the middle. Both antennas remained outside the airplane. After finding no change I went back to a single receiver.
I spent roughly a week messing with the tuned-pipe system, making sure the pipes didn’t touch the headers and swapping headers and pipe positions around. I even replaced the pipes to make sure a baffle wasn’t loose inside one of the pipes, causing vibration-induced RF. I made a considerable effort to make sure there was no metal-to-metal vibration or contact anywhere in the exhaust system.
I found that part of the carbon tunnel I had built had a crack that was loose and vibrating, so I epoxied that together securely. I also reinforced the thin-walled tunnel (2-ounce total weight!) with strips of fiberglass to help reduce vibration.
I split and twisted the leads going to the back of the airplane. I changed all the wiring in the airplane. I rebuilt each servo (I sent them in for service and to be checked by the factory techs) and placed them in new locations.
I added toroidal rings (chokes or filters) to each new lead. I figured that since I couldn’t get a response from all the changes in the airborne systems by moving antennas and changing receivers, channel, and modules, maybe the RFI was coming in through one of the leads.
At this point I started suspecting the carbon parts. Since there are three separate carbon structures in the model, opposing electrical charges built up in the carbon could emit RF. So I grounded all the carbon parts together.
Although I had moved the receiver around a bit previously, I started to wonder about its proximity to the tuned pipes, so I moved it a total of four times. It is now high inside the turtledeck.
At this point a range check using the fail-safe method gives a range exceeding 100 feet, which is acceptable. But I still cannot eliminate the dropouts on the ground. This frustrating process took approximately three months and grounded the airplane for most of the summer. I stopped short of ripping out all the carbon fiber and rebuilding the model with completely new components.
I’ve flown the airplane roughly a dozen times since I gave up trying to make it perfect, and it hasn’t gone into fail-safe—not even for a split second. Although I was unable to completely cure my aircraft’s ailment, the methodology for isolating an RF problem is sound. Now I understand why some people refer to RFI as black magic or voodoo. If anybody has a line on a model-airplane exorcist, drop me a note! MA
The Carden 40% Extra 330. Initial range checks seemed fine, and Mike had almost 30 flights before he started to realize there was a problem with the airplane.
Transcribed from original scans by AI. Minor OCR errors may remain.
