900 mhz Long range antenna selection / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

900 mhz Long range antenna selection

1 2 next

Move Thread

LAN_403

SaberShip

Tue Apr 01 2014, 07:25AM

Registered Member #4546 Joined: Mon Apr 23 2012, 02:46AM
Location:
Posts: 8

Hi!
As a free time project a couple of friends and myself have gotten into the hobby of Near-space Weather ballooning. Already with 2 successful launch, track, and recover missions we are confident in our method of tracking and recovery and would like to start trying more "exotic/challenging things" with our payloads.

We are currently using 900 Mhz, 1 Watt Digi 9xTend modules for our real-time tracking, uplink, and downlink. (not the large and heavy xtend-pkg ones). The antennas we are using are: 2 identical 9dbi gain panel antenna, one on a car pointing up and the other on the payload pointing down and serving as the floor for the payload box. This is where our dilemma comes in, each antenna weighs in at about 1 pound (almost 1/3 of the payload mass), and weight is everything on these balloon flights.

Any help on a lighter replacement antenna for the payload would be appreciated. We have had absolutely no trouble with signal loss and we are almost always directly below the payload. Also, our distance from the payload is not likely to go over 15 miles LOS

I'm pretty new to RF and stuff so, thanks for any help!

PS. (I realize 1 Watt for our transceivers may be a little overkill, i plan to have the on-board computer manage the output power of its transceiver based on RSSI to save battery)

Andy

Thu Apr 03 2014, 09:59PM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

Hi SaberShip

If it was me, I would download Eznec

, and bruteforce designs.
You would probable be after hi dbi, I think 4 dbi adds 300meters at 2.4ghz or it might have been 300% for 4 dbi, so guessing about 15dbi would be a good range to aim for.
To make it smaller, you might need phase shifted pieces, with maybe 2-3 yagis side by side
You should beable to get 10dbi from about 5 elements, adding more slowly increases it, better to work on ground element reflections on the antenna etc

Hope it helps

Dr. Slack

Fri Apr 04 2014, 07:59AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

I'd use the assymmetry inherent in the situation, and increase the weight and gain of the ground element in order to tolerate a hit on gain in the flying one. You probably don't want too much gain in the air anyway, as with increased gain comes a tighter beam. Are you always going to be directly below the payload? I'm assuming you are not intending to point the balloon antenna. On the ground OTOH, you can track the balloon.

What polarisation are you using? The balloon can rotate with respect to the ground. If the balloon is transmitting EW, and the car is NS, you will receive nothing. At 45 degrees difference, that is much of the time, you're 3dB down. With circular polarisation, the relative compass orientation of the balloon and car is irrelevant, you are always at max gain.

I would suggest a lowish gain end-fire circular polarised airborne antenna, pointing down. Helicals are relatively wide bandwidth, so non-critical dimensions to tune to your transmitter frequency, unlike the Yagi, especially if made with thin wires instead of thick tubes for weight. You can mount a pair of Yagis with a phase shifter to generate circular polarisation, but I think a helical is much simpler in the air.

There's more info on them on the net than you can shake a stick at. Try wikipedia, or this calculator

At 900MHz, the diameter of a helical antenna is about 4" (lambda/pi). That feels like the diameter of a 2 litre PET drink bottle, which could be used as the wire support for a short airborne helical. I reckon that construction would bounce back for most hard landings, and when it didn't, it's dirt cheap to replace. It's also the diameter of sub-ground PVC drainpipe, which crossed my mind for the ground based antenna, but I'm not sure about the effect of the RF loss of that on antenna noise figure, might be better with polypropylene or the like.

Of course you could use crossed Yagis on the ground if you wanted, or even a single Yagi, taking a 3dB hit in gain. Because the airborne antenna is transmitting circular, a single receiving Yagi would never drop into a null. It might be easier to get the 3dB gain back with doubling the length, rather than crossing 2 with a phase shifter, though that would give you a tighter beam and more sensitive tuning. Compromises, compromises.

Make sure you build the Rx and the Tx antenna with the same handedness so they can see each other.

One other feature of helicals. They look cooler than Yagis, or at least spacier. Wear a tin-foil hat when operating the truck-mounted one, just to look the part.

SaberShip

Fri Apr 04 2014, 04:54PM

Registered Member #4546 Joined: Mon Apr 23 2012, 02:46AM
Location:
Posts: 8

Thank you both for taking time to reply, all of this causes me to realize how much I still don't understand about radio communication.
Here is a link to the antenna we currently have for the ground and the air http://www.ebay.com/itm/900-Mhz-Panel-Antenna-9dBi-ISM-902-928-915-/120380605303?pt=US_Directional_Network_Antennas&hash=item1c073e4377 In the description, to my surprise, it says that the polarization is linear. This confuses me because during our ground tests, as well as the first and second flights, AFAIK the antenna rotation never seemed to matter. Good communication was achieved with the payload throughout the flight; though you are correct Dr.Slack the payload does rotate with respect to the ground continuously, So I'm a little confused on this, did we just get really lucky?

I really like the idea of a cheap homemade helical antenna, though i am concerned with the landings as you mentioned. I may be wrong but isn't it possible for damage to the antenna to cause the driven element to short with ground and cause damage to the transceiver? I will try to think of ways to design it so that it doesn't break on landing, because it is important that we maintain short range communication with the payload after it hits the ground. This is just in case we are not able to be right were it lands, we have found that with our current antenna setup we can get range of almost a mile with the payload antenna just facing down into the ground like it landed. This is helpful for when it lands far from a road in a field or something.

My last question is if we did decide to use a lower gain (down facing) helical on the payload would we necessarily need to replace our panel antenna on the ground? (Can a circular polarized helical talk to a linear polarized patch antenna?) We really found the panel antenna to work well with the car, its easy to mount, durable, has a 9dbi gain, and a nice 64 degree (or so it says) beam angle. If not that is ok too, we could take a slight hit for beam width on the ground, however it is nice to have it as wide as possible in-case we fall behind or it travels to far from a highway.

Thanks again for all the help, I will look into helical antennas some more!

Dr. Slack

Fri Apr 04 2014, 07:08PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

You can recieve a circular polarised wave with a linear polarised antenna, you just fail to collect half of the transmitted signal power, 3dB. This is in contrast to a linearly polarised wave where depending on the orientation you get all, or nothing, or most of the time a few dBs low. So, use your ground based patch before you go building anything different. You could cross your two patch antennae and combine them with a quadrature coupler to pick that 3dB up again

It can be difficult to get 'no signal' with crossed linear antenna, any scattered signals will fill in your null. It's not uncommon for a patch antenna, any antenna really, to have a bit of cross-mode leakage, especially if things, like people, are near it, so you wouldn't have to get that lucky not to see a null casually, especially if you have plenty of power. The time you will see it of course is the time you *really* don't want to see it.

SaberShip

Fri Apr 04 2014, 09:37PM

Registered Member #4546 Joined: Mon Apr 23 2012, 02:46AM
Location:
Posts: 8

Sounds great! Ok so just did some tinkering with that helical antenna calculator you posted. Looks like I could whip up a small helical for 915mhz (center of the FHSS) and it would be considerably lighter than the panel we have on the payload now.

I'd like to keep it as short as possible without losing to much gain tho, if we keep the same panel antenna on the ground for now any idea what would be a safe but minimized gain for the payload antenna? (probably not an easy question to answer... I understand). It is looking like I can get roughly 7.79 dbi gain with only 2 turns at quarter wavelength, would it not recommended to have a non whole number of turns like say 1.5 turns? 1.5 turns would get about 6.5dbi which to me doesn't sound to bad, but we'd hate to lose contact toward the top of its flight.

Andy

Fri Apr 04 2014, 10:04PM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

You would probable want to do a link budget, workout all the losses from the transmitter to the receiver and see if its above 0, and allso, weather the receiver gain/sensitivity is higher than the losses.
transimmter = 90%
cabling = 95%
Antenna = 70%(70% of max distance)
Antenna = 70%
Diff = 75%(25%>0) passed
Sensitivity of receiver 65%(spec sheet)
25%+65% =90% >=100% failed

The elements don't have to be 1/4 length, its just the most eff, with the two elements, make a third one in the direction you don't want the signal to go, and make it slightly large, if the thing is 2.4cm,make the ground element 2.6cm(2.4ghz).
Things that can increase the dbi, is the distance the elements are away from each other length ways, the sizes of conduct(the source say 1mm, the second one 10mm), and the length of the directors

Proud Mary

Sat Apr 05 2014, 01:26PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

As you have no control over the attitude of the antenna on the balloon payload, you should not use a directional antenna on your balloon.

We can guess that the payload swings beneath the balloon like a pendulum, whose swing may be modified by windage, so that it will perhaps describe a figure-of-eight at some times, and an ellipse etc at others. At any rate, a directional antenna on the payload might only point at the ground station occasionally - or not at all - so is not something you should waste your time on.

I will suggest a single quarter-wave element - a piece of wire 8.3cm long (for a quarter wave radiator at 900 MHz.) If you attach a small weight to the end of the wire, it will reduce (if not completely correct) its angle to the vertical, so that - ideally - the wire will radiate in all directions at right angles to the vertical round 360Â°.

You will note that when this quarter wave radiation is directly above you its signal will be least, but as it drifts away, you will benefit from the monopole's tendency to produce good low angle radiation - just what you want!

In practice you are likely to find that the antenna must be made slightly shorter than 8.3cm, as stray LC coupling and loading effects will have reduced the radiator's natural resonant frequency. Correction for this is easy to do empiracally with a home made signal strength meter and some wire cutters: start with the radiator at, say, 9cm long, and clip off a millimetre at a time until you get the maximum signal.

As for the ground station, a simple Yagi antenna would give you good information about your balloon's direction of flight, and is easy to match to coaxial cable.

Dr. Slack

Sat Apr 05 2014, 07:28PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

As ever, ask on a public forum, and you get as many different answers as respondants.

Andy seems to be describing a Yagi. If pointing downwards this will be tranverse polarised. So dodgy to receive with a transverse polarised receiver. But as you already know, or should have worked out from reciprocity, if you use a circular polarised receiver, you can read it with no nulls, albeit 3dB low.

Mary is suggesting a very simple omni, vertically polarised so no problem with balloon direction. It does have one teeny tiny ever-so-slight defect of a null pointing straight down. From your earlier posts talking about chasing the balloon, you say you spend much time under it, or nearly so. For the vertical lambda/4 to be received, you need a bit of slant at the receiver.

Your call as to which you go with.

Andy

Sat Apr 05 2014, 08:07PM

Registered Member #4266 Joined: Fri Dec 16 2011, 03:15AM
Location:
Posts: 874

SaberShip , Do what ProudMary recommends, but a couple of questions, How far does 1dbi at 900mhz at 1 watt travel, and
if you have a 5watt 2.4ghz 25dbi dish, and its directed at a 100mWatt omni at approx 3-5dbi, what would be the range you could setup a network?Do a link budget.
With two dishes can travel 15km above...minus the 3-5 degree line up :)

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob