Has anyone flown the Park Flyer 6 channel, SS system at Horizon?

If you are talking about the new DSS radio, I read a post from Horizon a few days ago that stated this is for Parkflyers. They are working on a similar system for bigger planes and specifically stated that this is not to be used for larger planes. I was reading about this radio a while back. Sounds like a major change.

On Mon, 7 Nov 2005 06:28:57 -0500, "Ed Cregger" wrote in :

The work is underway to make a spread spectrum (2.4 GHz range) system for RC aircraft, according to some pattern folks who have test-flown the rig.

I don't know how long it may take to bring it to market. I'm sure there are lots of regulatory hoops to jump.

Marty

It would really be nice to not have to worry about shooting down someone else, or being shot down by someone else. Especially today, with all of the "renegade" flying that is being done.

Ed Cregger

| I can't remember off the top of my head whether or not the system is | available yet, but it has me a bit excited. | | Is there any reason why this system won't be scaled up and offered for | general R/C usage?

I can certainly provide some reasons why it might not be, but they're not the `official' reasons --

1) power is severely limited on the unlicensed 2.4 GHz band, which will probably give you a good deal less range than we're used to.

(However, our current 72 MHz systems only have marginally more power, up to about 3x as much, which many transmitters don't even actually use, so this alone isn't a major limitation.)

2) the 2.4 GHz band is already chock-full of random stuff. Yes, spread spectrum is supposed to allow you to share the band with other applications, but it only goes so far, especially when dealing with other applications that aren't spread spectrum. Get enough interference, and your plane isn't listening to your controls anymore.

I suspect that the 2.4 GHz unlicensed band is the most heavily used chunk of bandwidth _anywhere_ in the radio spectrum.

Even microwave ovens run in that band, and they tend to leak some RF energy. If a microwave will kill your WiFi so easily, what might it do to your plane?

3) Their FAQ still mentions channels --

When the transmitter is turned on, the system scans the 2.4GHz band looking for an open channel. 79 channels are available. When an open channel is found, the system locks in and transmits on that channel as seen below.

and it's quite obvious that only one transmitter/receiver combination can use one channel. But spread spectrum doesn't work that way -- if you do have channels, channels can be shared by multiple users. Things just don't sound right.

This makes it sound like it's not spread spectrum at all, just a clever way of finding an unused channel. (And the GUID stuff is nice, but all it really guarantees is that nobody else controls your plane

-- it doesn't guarantee that you can control your plane.)

I've emailed them asking for more clarification on what this all means in the past, and all I got back was some more marketing speak about how perfect the system is, how interference is impossible, etc.

(And really, this FAQ is *full* of marketing speak. `For the highest security levels of radio communications, NASA, the U.S. military and government agencies like the CIA and FBI rely on Spread Spectrum technology' -- bah! I could write a whole post just on what's not quite right about that statement.)

Obviously the transmitter scans for an unused channel and uses that. But what if you've got two transmitters a half mile or so apart, seperated by a big hill so they can't pick each other up, and they each choose the same `channel' (it's not even really clear what a channel is in this context, but I suspect it's just like the channels we use now) and both take a plane off that flies over the hill. I suspect that somebody's going to crash.

In any event, I suspect that it's these limitations that are keeping them from `okaying' their system for use in anything large enough to cause some real damage -- they just don't want the liability. It probably works well enough for things you keep very close -- cars, park fliers -- but anything that can get 500+ feet away, problems could show up.

Ultimately, spread spectrum should be the solution to many of the woes affecting our radio systems. I'm just not convinced that Spektrum RC really has the correct implementation of it.

Doug- From what I have been able to gather, the system currently going to market for parkflyers has 100 mW output. This is strictly a guess, but I suspect the design borrows heavily from modules developed for appliance (e.g. cell phone) markets with huge sales volume. Output of

1 W is allowed in the band, but I suspect that is more than what off-the-shelf modules can deliver. If so, another 10 dB of link margin is to be had for the cost of a custom design of the Tx final. That may be the essential difference that would make it applicable to models that may be flown beyond about 100 meters from the Tx.

As I read it (including some other accounts), the band is divided into

80 slots of 1 Mhz each. The Spektrum radio searches for and selects two unused slots. The whole RF link is redundant, with 2 rx's, each with it's own antenna. There are several advantages to this approach: The antennas can be orthogonal, so if one should lose signal due to directionality ('omni' gain of a whip is in one plane only), the other fills in. That addresses the range-rate problem familiar in comms at UHF-Hi and above in vehicles that can quickly change orientation. The doubling of signal b/w is theoretically good for 3dB. Also, the permutions possible with pairs of channels from a group of 80 is 80^^2. The 32- bit GUID is used for digital sequential spreading, true spread spectrum stuff, though if you like me are more familiar with the frequency hopping implimentations it seems less intuitive. The Rx is bound to the Tx by the GUID, which might be thought of as a signal mask. Signal energy accumulates across multiple bins in the sequence representing a data frame as it passes through the mask, other signals not in proper sequence per the codeword only contribute randomly to the energy in the bins, like any other noise. There is gain in S/N proportionate to the number of bins that represent a data element (e.g., bit), so the 'GUID stuff' actuallly does contribute a lot to robustness of the link - often in the ten's of dBs. There's a lot dB's to make up for when going from 72 MHz to 2.4 GHz due to atmospheric absorbtion; I get ~30 dB disadvantage for 2.4 GHz without resorting to a calculator.

Abel

| From what I have been able to gather, the system currently going to | market for parkflyers has 100 mW output. This is strictly a guess, | but I suspect the design borrows heavily from modules developed for | appliance (e.g. cell phone) markets with huge sales volume.

I was thinking WiFi would be a closer match. Or Bluetooth, which is even closer.

| Output of 1 W is allowed in the band

Hmm, I was thinking 0.250 watts, but it seems you are correct. Our 72 MHz radios are capped at 0.75 watts, so that's an (miniscule) improvement.

| As I read it (including some other accounts), the band is divided into | 80 slots of 1 Mhz each. The Spektrum radio searches for and selects | two unused slots.

... two slots that it thinks are unused. What's unused at ground level may not be unused at some distance. But since it's really spread spectrum, it shouldn't matter that much.

| UHF-Hi and above in vehicles that can quickly change orientation.

`Picket fencing'. Yes, it seems a good plan.

The FAQ I read covered their car system, which doesn't do all of this.

| Also, the permutions possible with pairs of channels from a group of | 80 is 80^^2.

Technically it would be 80 * 79. Though assuming that the odds of a problem are 1 / ( 80 * 79 ) also assumes that there's no other users of the band. And hopefully the system is smart enough to not pick two adjacent channels, and it stays away from the frequency used (2450 MHz) by microwave ovens.

| The 32- bit GUID is used for digital sequential spreading

... and since they are using spread spectrum (the new faq is a lot better than the old one) then even if two transmitters are on the same channel, it should still work.

Still, I'm surprised they chose 2.4 GHz. 5.8 GHz seems a better bet, since it's not used so heavily, and the antennas could be even smaller. FSL is larger, but not too much so.

Also, the antenna design on the transmitter (a whip just like existing whips, but smaller) seems to be less than ideal. Since your TX is almost always pointed in the direction of the car/plane when in use, at least in the horizontal plane, it seems to me that a dipole antenna (which is practical considering how small it is) pointing to the left and right would be perfect, giving you even more range.

| There's a lot dB's to make up for when going from 72 MHz to 2.4 GHz | due to atmospheric absorbtion; I get ~30 dB disadvantage for 2.4 GHz | without resorting to a calculator.

The fact that you didn't give a dB/distance attenuation factor suggests a mistake was made, as atmospheric absorbsion would indeed be a function of distance.

Looking it up, at 2.4 GHz, it's approximately 0.01 dB/mile, which is utterly insignifigant. (Cite:

Now, there is free space loss, but that's totally different. And yes, it's larger for 2.4 GHz than 72 MHz.

Their new FAQ also has some errors. For example, this one --

So when FCC devised the 2.4GHz rules, they set about a completely new system: they mandated that all 2.4GHz devices will transmit at less than 1 Watt.

... which is patently incorrect. For example, the ham band overlaps the unlicensed 2.4 GHz band, and the power limit there is 1500 watts (assuming it's not spread spectrum) with no limit on antenna gain.

Or this one --

Since all devices on the 2.4GHz band are required to avoid other users (or else they wouldn't receive FCC type acceptance)

That would only affect spread spectrum uses. Not all 2.4 GHz band equipment is spread spectrum. For example, non-spread spectrum 2.4 GHz cordless phones and wireless video cameras are known to wreak havok with WiFi, and they could totally knock out one of these R/C systems, especially the video system, probably using 6 MHz of bandwidth. Hopefully the DX6 is smart enough to make sure the channels it chooses are at least 6 MHz apart.

In any event, I was looking at the specifications for the old Spektrum system for my original post. (Their web site could benefit from some work.) The new plane system, DX6 looks a _lot_ better. And with a $200 price tag for 6 channel TX, RX and 4 micro servos sounds like a hell of a deal.

Of course, apparantly it's not even out yet. `Early December'.

Doug-

Yup, I estimated free-space path loss rather than what I incorrectly stated. It's a function of 20 log (freq) + 20 log (dist). At any arbitrary distance, the difference is reduced to a function of the ratio of frequencies^^2, and that amounts to the ~30 dB given.

Abel

Wow! If I remember right -30 db is a 1000 fold reduction in whatever - signal strength in this case? Sounds like the bird flu is a much bigger threat than 2.4 GHz. :-)

Red S.

You remember right (I got a ratio of 1100 +/- small change to arrive at ~30 dB), and the 'whatever' is indeed signal strength. The increase in transmission loss just due to the move from VHF to the upper end of the UHF band leaves you on square one with a 30 dB deficit in S/N assuming the noise part is white across the RF spectrum (it isn't really, but will do for this discussion). That deficit has to be made up in processing gain, and that is what spread spectrum is all about. If we allow that 10 kHz signalling bandwidth is adequate for R/C as we know it, 2 slots of 1 Mhz each provides for a 200 fold increase due to processing gain, so it ain't so bad as it looks.

Abel

| Wow! If I remember right -30 db is a 1000 fold reduction in whatever - | signal strength in this case?

The frequency component of the free space loss formula is there _strictly_ to take into account the effective receiving `area' of the receiving antenna. It also assumes isotropic antennas, which do not exist in the real world.

A 2.4 GHz antenna of a given type (like a half wave dipole) is 33 times smaller (in one dimension) than a 72 MHz antenna of the same type, and so the received signal is 1100 times smaller, assuming that the transmitted power remains the same. However, this is not a problem for us at all -- as the noise received by this antenna will also be 1100 times smaller, and it's the S/N ratio that really matters here. The weaker signal is easily compensated for by simply setting up the receiver's internal amplifier for more gain.

Since the channels are 1 MHz in size vs. 10 KHz in size as they are now, Abel is right -- the S/N ratio needed should be very low, much lower than what we have now. Really, with one watt to work with, and a clear line of sight (which is hard to fly without) it shouldn't be difficult to get all the range we're used to out of a 2.4 GHz system, and more.

Also, since the antennas are so small, they could easily include a dipole with some gain in exactly the right directions (on the TX anyways), which would give you more range.

In any event, in the radio world, a 30 dB difference in signal strength isn't _that_ big a big deal. And since the noise is reduced by a similar value, it's pretty much a non-issue here.

| Sounds like the bird flu is a much bigger threat than 2.4 GHz. :-)

Is threat the right word here? I only see a few bad things here for the hobby in general here --

- Spektrum appears to have `patent pending' all over their web site, which suggests that they're patenting everything in sight, even though what they're doing is really only natural (obvious) extensions of existing technology, so most of it shouldn't pass the `obvious to the layperson' requirement of the patent office. Unfortunately, this requirement is very weakly enforced, and so it's entirely possible that Spektrum may end up locking up the entire market via patents, making it impossible for anybody else to make R/C spread spectrum devices in the US.

- They also like to use the word `proprietary', which suggests that other vendors won't be able to make equipment that interoperates with this equipment. So it would be like PCM -- your RXs only work with TXs by the same vendor.

- I'd have suggested the 5.8 GHz band instead of 2.4 GHz, due to less noise, but ultimately it shouldn't be a big problem. Since Spektrum is the forerunner here, other vendors are likely to make things that work similarly.

(I'm guessing they've adapted WiFi or Bluetooth equipment for this, and that's why they chose the 2.4 GHz band.)

- Giving users equipment that doesn't require frequency management may help cause the users forget about frequency management if/when they move up to large planes on the 72 MHz band.

- Having R/C gear on 2.4 GHz will mean that you can't have a 2.4 GHz transmitter in your plane (unless it's spread spectrum as well, though the small/cheap video systems we've used usually aren't.) Even having a 2.4 GHz video transmitter in your plane with 72 MHz R/C gear may crash _other_ nearby planes with 2.4 GHz gear. (It all depends on how close the video TX is to the RX, how close the R/C TX is, and which frequencies are being used by both. Since a video signal is usually 6 MHz in size, it could knock out both frequencies used by the DX6 if you're not lucky.)

- 2.4 GHz is absorbed more by the human body than 72 MHz. I personally don't think that cell phones cause cancer, but certainly,

2.4 GHz systems are going to be a bigger danger than 72 MHz ones there. (And this may be why Sketrum hasn't added the dipole antenna I was mentioning -- the gain of the antenna would increase the RF energy directed at your plane, and decrease it aimed at the people standing next to you, but it would also increase the RF energy aimed at you. Though a fancier antenna could undo some of that, but you'd look sort of dorky with a TX with a cantenna on it. Though perhaps we're used to looking dorky?)

- If everybody starts using R/C equipment on other bands, the FCC may decide to take the 72/75 MHz bands from us. Of course, this would be many years away, and if it does come to pass, we're not likely to care that much anymore (since if everybody's really on 2.4 GHz, that means nobody's on 72 MHz.)

- With a 3" RX antenna, that may not be long enough to get it out of a carbon fiber fuselage. (Though I guess you could extend it to 9" which should work reasonably well.)

| > Yup, I estimated free-space path loss rather than what I incorrectly | > stated. It's a function of 20 log (freq) + 20 log (dist). At any | > arbitrary distance, the difference is reduced to a function of the | > ratio of frequencies^^2, and that amounts to the ~30 dB given.

The club president for one of the local flying clubs bought one of these a few weeks ago. He wanted it mostly for indoor and nitro helicopters, but he started flying his Zagi Tazz (

with the Spektrum radio system.

Horizon Hobby won't put an exact range on the system, at least not on their web site at this time. Some reviews say it has a maximum range of 1000 feet or so, while I've read other claims that say up to half a mile.

While I don't know what the maximum outer range is, I can say the Spektrum radio handled an 85mph + flying wing without a hiccup. The owner of the radio system has been really happy with all aspects of it and the Spektrum transmitter has a lot of nice helicopter and airplane mixing functions available.

Radio glitches and accidental radio overlap seem to occur at indoor flying events more frequently for our club. I can't imagine anyone who flies indoors not loving this radio system. While outdoor flying may enjoy better frequency management for traditional radios, there is no reason this wouldn't be fantastic for outdoor helicopter and parkflyer planes.

Having seen this system at work, it is quite impressive.

I got mine just last week. Replaced a GWS Rx in a Hell Raiser (11 oz foamy Ultimate), and have flown it a couple of times. It does seem like I've forgotten something heading to the flight line, as no frequency pin is needed. Range check is facilitated by depressing the 'bonding' button on the back of the tx to reduce power output, a nice touch. The feel is different without a long whip antenna; it only needs one a bit less than 4 inches long. The Rx is only 7 gms IIRC, and that includes the sturdy case. Four micro servos included, and along with a digital Tx that has decent mixing options and digital trims and 10 model memory, it is a good value at 200 bucks. The only thing I've found not to like is the lack of expo on rudder - this surface gets used a lot on foamy 3D models, so it seems an odd omission.

Regarding range, I have seen a report of a static ground test at 0.7 mi, however, at 2.4 GHz range rate is an issue, and so usable range will be affected by vehicle dynamics (causes phase jitter in the rx'd signal). The more rapid the change in vehicle attitude, the more range will be affected. I think that is why Horizon is reluctant to put a number on it. For Ed Cregger, I see no reason it won't be scaled up for general R/C. The current offering has an output power of 100 mW, probably constrained by what chips are available off the shelf in the comms appliance market. One W is allowed in the band, so there is 10 dB more to be had for the price of a final amp stage to boost the power out. The Rx would appear to need nothing more than larger bus strips for supplying power to high torque servos. Abel

| For Ed Cregger, I see no reason it won't be scaled up for general | R/C. The current offering has an output power of 100 mW, probably | constrained by what chips are available off the shelf in the comms | appliance market. One W is allowed in the band, so there is 10 dB | more to be had for the price of a final amp stage to boost the power | out.

They could also get perhaps another 3 dB or so by replacing the small whip antenna with a small dipole that sticks out to either side of the TX from the top. Like this --

--------\ /------- -------------------- | | | [ .] [. ] | | | | \-------/ | --------------------

(ooh, ASCII art!)

Since you are almost always facing your model, with your TX in front of you, the plane would always be in the main lobe/beam (though it's more of a donut than a beam) of the antenna. Unfortunately, so would you, and meat is known to absorb 2.4 GHz signals very well, so there's going to be some concern there. Perhaps some sort of shield/dish should be added?

| The Rx would appear to need nothing more than | larger bus strips for supplying power to high torque servos.

13 dB of signal gain should certainly be enough.

Though I'd hope they'd give you more channels. Six just isn't enough. Seven is a lot better, though I'd suggest nine.