Hi everyone,
I’m on a team that plans to launch a two stage rocket, and we have separate telemetry systems for the two stages. While the second stage will fare quite well because the nose is fiber glass, I’m somewhat perplexed about the first stage. I’m using an Xbee-900 Pro for that and I can’t have the antenna inside as the body is aluminium and will knock out RF. Does anyone have any experience with handling this issue and perhaps can offer some suggestions?
Thanks!
Siddharth
Maybe a ‘patch’ antenna on the outside of the rocket body.
Or, a dipole antenna built into one of the fins (if non-metalic).
Either would require an XBee module with an antenna connector then a short piece of coax to the antenna.
How large is the rocket body?
4 inches outer diameter, the total rocket length is about 12 feet right now, the first stage electronics sit about 7 feet off the base…I’ve been looking for a patch, but sparkfun has a patch only for 2.4 GHz. The other concern with a patch is insulating it from the outer body of the rocket…We currently have a rubber duck which turns 90 degrees about an elbow… One idea that I was thinking of was to maybe just have the elbow slightly outside the rocket body, and turn it 90 degrees to have the antenna pointing up or down and then protect it from the air blowing past it.
How you googled “900MHz patch antenna”? There is a good bit of info on both commercial antennas and DIY antennas. Here is one:
http://fpv-community.com/forum/showthread.php?tid=2
It may also be possible to cut out part of the aluminum body, a window, to mount a patch antenna flush with the outside surface of the body tube. This could minimize the air drag.
A 1/4 wave antenna sticking out the side of the rocket body would work if the air speed is not too high (too much air drag and possibly bending the antenna) or a rubber duck as you described could work.
What is the expected maximum air speed?
4in dia X 12 ft long. Oh my that’s some rocket. Do you have a link to more info and pictures?
Thanks for your help!
It’s still in the assembling process, scheduled for a launch in mid-March, so unfortunately I don’t have any pictures yet…I’m just thinking ahead 
So the max airspeed is a problem, it hits Mach 2, but after stage separation, so the first stage goes up to about 0.8 Mach roughly. So I really don’t want the antenna to stick out sideways, it’s going to fly off very quickly…with the 90 degree turn, it’s pointing along the rocket, but still needs something to protect the most upstream part from exposure…
Also, my problem with a patch is that it’s not omnidirectional… so there may be intermittency as the rocket rotates to stabalize…
What about using a fin? Are they metal?
If not metal maybe an antenna could be built into the leading or outer edge of a fin.
Another possibility would be to have the antenna(1/4 wave) hang down from the outer trailing edge of a fin.
You could do some mock-ups to see what the antenna pattern would be and some range testing.
I thought of that.Also, my problem with a patch is that it’s not omnidirectional… so there may be intermittency as the rocket rotates to stabalize…
Maybe if two receivers are used, spaced to each side of the rockets path, so that at least one receives the data.
Question: How are you going to keep the two transmitting XBees, 1st & 2nd stage, from interfering with each other?
Maybe another solutions is use only one XBee in the upper stage. Then have a 1st stage data set to the telemetry module in the upper stage. It shouldn’t be hard to find a connector that will separate when the 1st stage drops away.
The second stage telemetry module is an off-the-shelf thing, communicates at 70cm wavelength, so no interference issues. The philosophy here is to do the first stage in a built-from-scratch way and the second stage in a predominantly off-the-shelf way. The other reason why I have an independent module is that we’ll continue to have full telemetry after stage separation.
The fins are predominantly composite, i think. The trouble with an antenna there is having a cable that long going down the body of the rocket, and on the fin…yeah some mockup testing would really help.
Worst case, of course, one just logs in the data into an SD card and recovers it on landing.
try a wire inside the aluminum. I was surprised when i put an XBee (2.4GHz) inside an aluminum “Bud” box. Range was not hugely diminished. It may be due to that metal not being ferrous.
Failing that, for 900MHz… an antenna in the nose-cone would have an RF shadow straight down to the ground. But if your receiving antenna is say 20 degrees off nadir, which it probably is, a nose cone antenna would probably do well. That could be a 1/4 wavelength of hookup wire not coiled but stuffed.
Those little stubby helical antennas for 900Mhz are awful.
2.4 GHz doesn’t have the atmospheric range though…I expect this to work up to about 7000 feet, which 2.4 GHz will not do…2.4 GHz may just be a frequency high enough to penetrate through metal, but scatters quite easily in air.
The solution to that is a directional antenna with Gain on the ground. A small Yagi will increase the line-of-sight range considerably.sidkmoorthy:
2.4 GHz doesn’t have the atmospheric range though…I expect this to work up to about 7000 feet, which 2.4 GHz will not do…2.4 GHz may just be a frequency high enough to penetrate through metal, but scatters quite easily in air.
Hmm…That’s something I could look at…
Just wanted your opinion…for this purpose…do you think an xbee module with a U.fL connector might be better than one with an rpsma connector?
I’d go U.fl - use short “pigtail” from that to SMA or RP-SMA to mate with the antenna you select.
As to the line of sight range at 2.4GHz… it’s not “scattering”, it’s simply the path loss from attenuation at 2.4GHz, since it is LOS and no terrain. Path loss per distance unit (mile, Km, etc) increased 6dB per octave as the frequency increases. Not as much as you’d think. Also, antenna gain is easier to get at higher frequencies because antenna size reduces - so this often offsets much of the greater path loss at higher frequencies.
The math is simple and accurate, to calculate the received signal strength and margins, unless you use a directional antenna on the rocket - which I would not do because of size and packaging problems.
The main issue is to get a lot of antenna gain on the ground-side - like 12dBi or more which leads to a reasonably easy to point antenna. If you use a dish or some such, it’ll be too hard to keep aimed.
There are many on-line path loss calculators. The math is simple (I have a spreadsheet).
Transmitter power + transmitter antenna gain (if any) - path loss (distance dependent) at 2.4GHz + receiver antenna gain. Vs. required signal strength. This is often called a link budget. (in air to ground, you don’t have to allow for loss due to “Fresel zone” clearance (terrain).
For 802.15.4 such as XBee Series 1 PRO (not non-PRO, low TX power)… and the PRO have better receiver sensitivity, here’s an example estimate:
Here’s one arrangement of the formula for path loss at 2.4GHz. The path loss goes into the link budget terms I listed above.
Distance (km) = 10(Free Space Path Loss – 32.44 – 20log10(f))/20
The XBee PRO would be
TX power 60mW which in dBm units is about 17dBm. On the rocket, let’s say a wire antenna right on the XBee is used. Key is the antenna’s view of the ground - not passing through the engine area.
we have, for a 1Km path (100dB of path loss at 2.4GHz, less at, say, 900MHz) and a ground side 14dBi antenna and 2dB of coax losses (keep that short and low loss!!!)
+17 + 0 - 100 + 14 - 2 = -71dBm at the receiver input on the ground. And vice-versa for the uplink.
The receiver sensitivity is about -95 for 802.15.4 at 2.4GHz with the XBee Pro’s LNA. So the margin is
-95 - (-71) = 24dB. That’s excellent. But real-world, let’s change the rocket side to have loss due to the housing and antenna not seeing the ground well. Change the +17 - 0 to -17 - 6 and do the margin.
Maybe the best rocket-side antenna is just a 2.4GHz “dipole” which has about 2dBi of gain. That could be a hole in the housing, RP-SMA connector, and a thin-wire antenna of the right length - running perpendicular to the housing. Thin to minimize aerodynamic impact. Not parallel to the housing as that would make it rather directional, disadvantaging the “down” direction.
If you used 900MHz instead of 2.4GHz: path loss decreases from 100 to about 91dB. 9dB better. But that 9dB is still small in comparison with the total path loss at either frequency. Also, the ground-side antenna won’t be 12 or 14dB because that much gain at 900MHz is an antenna size that’s unwieldy.
If you double the path length, then the -100 become -106 in the formula. And so on. It’s the laws of physics in RF, “inverse square law”, double the distance, etc. It’s why we can talk to satellites outside the solar system!
Hi Steve,
Thanks for your advice. I think the frequency-dependent part of LOS loss is in fact, due to scattering. There’s is of course the simple 1/r^2 loss because of the increase in surface area over which the energy is distributed.
Thanks,
Siddharth
Wow, Steve really nailed that. Thanks for this explanation, it helps.
The best solution is to emulate the big expensive liquid fuel rockets that fly to orbit with super expensive payloads…
Run your antennas feedlines from the radios to the appropriate locations on the rocket body where your multiband dipoles are strung parallel to the rocket body. These signal feeds would be proper shielded mini coax from the SMAs or whatever interface jack the radios have, minimizing line loss and making the link to the antennas as neutral to affecting vSWR as possible, so that the Coms link performance is mostly determined by how accurately you phase and match the antenna array.
Now, the antennas are 1/2 wave dipoles with the counterpoise (virtual ground) section of the dipole is above the antenna feed point. The driven element extends down the rocket body from the antenna feed point in the opposite direction.
Since the body as metallic, Use 10mm rubber standoffs to separate the dipole elements from the rocket body.
This will naturally create a directional condition with signal gain existing on the side of the rocket with the antenna.
So then you use the fins to spin the rocket real fast so you now can get rapid blips of telemtry synchronized by the Roll axis rotation rate…;0)
no. kiddings…
You phase together two more Dipoles of the same common feed line forming a tuned array of 3 diploles 120 degrees apart.
Do the same methodology on any other RX/TX subsystems onboard, feeding RF to location of dipole feed points with shielded coax right into open radiating elements positioned the same way…if only two feeds (I/O) to two onboard radio subsystems needed, it would be one band operating via tri-phased rocket body dipoles(3) 120 degrees separation, and the other antenna array for the 2nd feed just laid out the same way but out-of-phase 60 degrees (in relation to the 1st array’s locations) so both antenna arrays share congruent placement around the rocket body exterior, with no blind spots and minimal fluctuation. Each 60 degrees around the body you have a dipole from one of the two arrsys mounted parallel to the body, and with driven elements facing “retrograde” (back at earth) during flight…the next 60 degrees around is a dipole from the other array, with this alternating arrangement.
This provides basically a vertical polarized antenna system you can use with verticals on the ground station, or some directional Yagi’s (recommended anyway) arranged with vertical polarization mounting for GS…or, co-phase the Ground Station antennas (vertical, plus the directional Yagis also) for tracking diversity , or rapid-switch between omni directional and directional segments during flight depending on need.,.
One other method would be to employ LHCP or RHCP polarization and rout the antennas “around” a desired location of rocket body (cork-screw) with a mounting arrangement winding around the rocket body 'no tighter than 30 degrees pitch, leaving enough separation from itself to allow maximal tuning potential (lowest vSWR we can get).
Paying attention to whether the onboard antennas are wound right hand or left hand polarized orientation, matching this with RHCP or LHCP (? hand circular polarized) antennas. These can be aimed directionally at the rocket also, like a beam Yagi for enhanced signal quality, or can be static mounted and used as omni directional antennas.
I would lean toward the dipole orientation arranged as described with counterpoise elements facing toward the nose, and drivens facing earth.
AS long as you give the elements some non-conductive stand-offs (10mm should do) SWR results with both antenna arrays for both radio subsystems operating at same time should be more than acceptable…in fact, they should provide tuning latitude to allow both your antenna arrays quite exceptional tuning potentials, with very low vSWR at the radios onboard.
The key here is feedline quality, matched to a carefully measured dipole array sporting decent SWRs (or as good as can had). The math can be kinda hairy constructing a tri-phased dipole array, theres a deviation in there that makes each singular dipole by itself high SWR (out of tune for single operation by itself) , but when you get that common length just right its magic.
Signal propagation from tri-phased vertical polarized dipole arrays it excellent in most cases.
One of my alltime favorite homebrew antennas for Ham radio has been tri-phasing 3 longwire dipoles with 3 feedlines, each dipole singularly tuned for standalone operation, and using a 3-into-1 antenna switch (mine were homemade with toggles) so you can choose Omnidirectional operation with all 3 being used at once, or selecting single diploes to use facing the direction of the target station, boosting performance by emulating a type of beam antenna with the two unused segments of the array acting as passive reflectors. Fun and easy little arrangements thay can be, like having a directional setup without all the cost and trouble of having a tower installation and Beams antennas to aim.
Anyway, that was slightly off topic, butr wanted to express the confidence of this arrangement while phased.
Wow. Thanks for that reply! I was actually thinking about the same thing for a while, the only problem is that the complexity of that system renders it very difficult to put together and test in the short time frame I have. For now, I’ve decided to settle on this:
https://www.linxtechnologies.com/en/pro … tennas/wrt
with the counterpoise shorted to the rocket and the dome sticking out. I’ve simulated it and it seems like it may work. I will test it out this weekend and see how that goes. If this fails, I’ll just log the data onto an SD card and hope that my rocket makes it.