Radio Technique
George M. Myers
MORE on Third-Order Intermodulation (3IM) avoidance. I have just received a very nice letter from Eldon Wilson, AMA 5247, of San Angelo, TX. I will share parts of it with you now. Mr. Wilson writes:
"...you may wonder if your messages are ever considered. Let me relate how your 'group frequency control' (July '85) worked.
"...Our recent West Texas Fun Fly saw 93 transmitters impounded, of which 31 were on the old frequencies. On the new frequencies were 29 AM and 33 FM transmitters. In two days of flying, with four to six airplanes up at a time (75-640 is still being used), we logged 194 flights without a single case of interference or unexplained crash. Being one of the originals in RC (since 1950), I consider this a record.
"...Our flight stations were kept a minimum of 25 feet apart, and all transmitters were kept a minimum of 15 feet from the runway boundary.
"...To be sure, we have the usual industrial transmissions to contend with: paging systems and telephone relays within two miles of the field and FAA transmitters 5 miles away. We don't have TV4.
"...Some of our local fliers claimed that our site was unfit because of unexplained interferences. I gave the club your group suggestion two months back and suggested that they try it. Immediately, the unexplained crashes by experienced fliers ceased. Only the novices continued to have 'interference,' not having learned that most interference is caused by the loose nut on the stick.
"...You might (look at a map) wonder how a place like San Angelo, TX can ever get 93 fliers together for just a Fun Fly. I've tried for years to get the AMA to have a meet designed for every type of flier, and every age from little kid to grandfather. And, it's real easy to do! We have no competition other than Spectators' Choice. Our prizes are won on a raffle basis. One chance with your name on it is put in the box for every flight you make. Then we hold a drawing.
"...This is our seventh year, and it just grows and grows. It must be our high entry fee (nothing). We double it each year, and still they come back. We have people from 700 miles away attend each year, and our average guest drives at least 140 miles.
"...It's quite a sight to see everything from 1/2-scale down to Electric flying at the same time. Our Spectators' Choice ballot box contained over 400 ballots for the two days. This quick note is Texas style, as we like to talk a lot. EW"
Well thanks, pardner. I like your style. GMM.
Back in July I gave you a "universal" five-station scheme for grouping frequencies, which is the scheme that Eldon used. Since then, I've had several calls for other groupings (four, instead of five, flight lines, for instance). This month, I'll show you how to make up your own groupings on the spot, in order to fit the particular group of transmitters or number of flight stations you use at your contest, Sunday flying, or whatever.
We saw, in last month's test results, that the most common worst interference comes from Third-Order Intermodulation (3IM). Considering only the 17 frequencies now in use in the 72 to 73 MHz band, there are 24 combinations which will produce 3IM. I call each combination a "triplet," because it consists of three frequencies mathematically related so that two times the frequency of one, minus the frequency of the second, equals the frequency of the third member of the triplet. For example: 2 × RC40 − RC42 = RC38 (because 2 × 72.590 MHz − 72.630 MHz = 72.550 MHz, which is the frequency of RC38). It works the same for all the other triplets listed in the Triplet Table.
Now, if all we want to do is avoid 3IM, we must start by knowing how many flight lines we will have and how many transmitters will be used on each RC channel or frequency.
Since it takes three transmitters to make a 3IM problem, if we fly a field with only two flight lines, we cannot simply grab any transmitter and go flying. Consider the guy in the pits who says, "Pete's up on RC38 and Joe's up on RC40; it's OK for me to turn on RC42 and do some testing." He'll be doing some testing, all right — he'll be testing Pete's receiver's ability to reject 3IM. That could lead to some annoying glitches. From this, we learn that every flight line should keep a copy of the 3IM Triplet List handy, to take care of guys who want to make throttle adjustments, control checks, etc. on the ground.
Now that we have that thought firmly in mind and we have decided how many flight lines we want to accommodate, the next step should be to look at the population: to find out how many transmitters are on each channel or frequency. Then, if we find large differences in the numbers (like 25 guys on RC48 and none on Pur/Whi), we should look for the reason. At the 1985 Nats, the reason would have been the operations of a cement-truck dispatcher on Pur/Whi, which was known and publicized prior to the contest. You can get 3IM with those kinds of transmitters, too.
On the assumption that every flier should have the same number of opportunities, each flight should be allowed as much time as any other flight, and flying should finish on all flight lines simultaneously. To do that, our aim is to make all of the flight lines contain the same number of transmitters—or as close to that as we can get. To get such a distribution, perform the procedure given in the "Flight Line/Channel Assignment Procedure" section below.
Let's say that the following frequency population is what you face at a typical contest.
Channel Assignment Worksheet
Chnl Assign Flt. Ln. 1 Flt. Ln. 2 Flt. Ln. 3 Flt. Ln. 4
Prohibited Chnl
Population
Totals
Here's the suggested format for a worksheet, which can be used to help plan what RC channels to use on multiple flight lines. There's an example of how to use it in the column. It's easy!
Note: Treat any known interference as being present on every flight line, with a population of zero.
Note that the population on each flight line is about as identical in size as we can hope for. We can now pat ourselves on the back for a job well (and quickly) done. See how easy it is? The distribution isn't perfect, but it's probably better than any single, universal arrangement will provide, particularly for small groups like this. When you get to larger groups, like Mr. Wilson's 93 transmitters, you are probably better off with the universal arrangement (fewer arguments and less work). The point is, it is easy to make up a scheme that is most efficient for your situation. Racing pilots do it every time they make up heat matrix cards.
You must take care of the "image frequency" problem by treating Brn/Whi + RC12 + Yel/Whi as a linked set that has to be put on the same flight line if Yel/Whi is present with one or both of the others. It's even possible that there will be so many of them in the set that you'll have to assign that set first. I've shown a more usual situation in the example.
Likewise, you can take care of Second-Order Product interference by treating Brn/Whi + RC38, Blu/Whi + RC42, Red/Whi + RC46, Pur/Whi + RC50, and Org/Whi + RC54 as linked pairs. I haven't found Second-Order Product to be much of a problem, so I left Org/Whi and RC54 on different flight lines in the example.
There are some things I haven't shown. For example, Hams fly on six meters (54 MHz), which can be mixed in with 72 MHz channels whenever needed. Likewise, there are still some who fly on 27 MHz (but I wouldn't, considering all the kilowatt amplifiers rolling around the country in trucks). They really simplify balancing the flight lines.
On the other hand, this process can be made much more complicated in Canada, where they are using virtually all 50 of the interstitial frequencies. There, they have about 600 3IM triplets to check. We can have the same problem here in the U.S. in 1991, if no changes are made in the AMA Phase-In Plan (and if the existing receivers aren't improved).
The method I have just explained will work for 24 or 600 triplets, or any other number. The only difference will be the amount of head-scratching involved. Perhaps someone will reduce it all to a computer program.
In setting up the conditions for last month's tests, I made the observation that separating the 3IM interferers by distances of five, 10, and 20 feet didn't help reduce 3IM. This is true enough, but it's subject to misinterpretation. Some folks who previously viewed the column for me came back with the question: "Does that mean that no amount of separation between the interferers, regardless of distance, will help?"
Flight Line/Channel Assignment Procedure
- Take the channel with the largest population and put it on Flight Line 1. Put that channel number in the "prohibited" box for all the other Flight Lines. Cross it off your list of channels to be assigned.
- Take the next largest population and put it on Flight Line 2. Put that channel number in the "prohibited" box for all other Flight Lines. Cross it off your list of channels to be assigned.
- Examine the 3IM table. Put the third member(s) of any 3IM groups on FL1 and FL2 so as to keep their populations balanced. Put the same channel numbers in the "prohibited" boxes of any remaining Flight Lines that have not yet received channel assignments. Cross them off your list of channels to be assigned.
- Assign a channel that isn't prohibited to Flight Line 3. If possible, it should be the third largest population. Cross it off your list of channels to be assigned. Put the channel number in the "prohibited" box of all other flight lines.
- Examine what you have now for new 3IM partners. Write them in FL1, FL2, FL3, etc., until you run out of Flight Lines.
- Pick another population group. Find a FL where it isn't prohibited. Assign it there. Cross it off your list of channels to be assigned. Add its number to the "prohibited" box of all other flight lines.
- Repeat steps 5 and 6 until all assignments have been made.
The answer to "Does separation help?" is: No — not unless the separations are very large. If we separated them by miles, there would be no 3IM. Actually, when you spread out transmitters on 40 kHz spacing by 40 ft., the 3IM effect disappears. Not so, however, for transmitters on 20 kHz spacing, which have to be spread out at least 80 ft.
Let's say that you have decided to ignore 3IM and deal with it by spreading four flight lines on 80-ft. intervals. Set them back 50 ft. from the runway centerline, then look at the landing situation: you need a Control Transmitter Advantage (M) of 245/50 = 4.9 just to break even! (You got the "245" by adding the three 80-ft. intervals between the first and fourth flight lines, then multiplying by the "miss-the-pilot" distance.) Add a little welt of turf and land a little closer to the guy at the far end—say 25 ft.—and the Control Transmitter Advantage you need jumps to 245/25 = 9.8 multiplied by the wet-turf effect (a factor of 1.7 is a good guess), which comes out to be 16.7. If you'll look at last month's Table 6, you won't find an MAS equal to 16.7 for frequency spacings under 20 kHz. Receivers will have to be more selective in 1991 than they are now. Till they are, you had better pay attention to 3IM!
There are still people who favor splitting the band into AM and FM segments. If they would but study the list of 3IM triplets for a few seconds, they would see how futile that idea is. I repeat last month's test results so they may realize that 3IM triplets cover the band, and that their influence is almost independent of the frequency spacing. The most significant indication is that one or more FM transmitters in the triplet makes the interference worse. So, the data tells us that the only certain road to a better 3IM situation is to ban FM entirely. But I don't think that idea will sell, either.
George M. Myers 70 Froehlich Farm Rd. Hicksville, NY 11801
Transcribed from original scans by AI. Minor OCR errors may remain.
