What is the maximum range required ? What sort of RF power output ?

If the cars are constantly turning left in a confined area, then you might be able to use a modest gain antenna on the receiving end. It all depends on where things are relatively located.

Instead of using a ground plane, you might be able to use a half wave vertical dipole (perhaps less weight), especially on 458MHz.

If you're experiencing drop-outs, it may not be a gain issue. It might be multipath issue (although gain properly applied can help in some cases). Often, two antennas can be used as with wireless mic receivers and wireless routers.

JB,
I'd say use a dual frequency linearly polarized patch antenna (21.5 inches square, 1/16 inch thick using Rogers 6002 material) on the roof of the car with 6 added ground plane wires brought down the roof structural supports connected to the roll cage or ?metal frame (or is it all composite?). The antenna can be inside the car, it doesn't have to be located outside on the top of the roof (assumes the roof is carbon fiber).The roll cage won't hurt it as long as the 21.5 inch square is between the roll cage bars.


The other antennas in your communication system should be circularly polarized because the car antenna will be polarized vertically in the forward and aft directions and horizontally in the side to side directions. Your problem is a two antenna problem. If you improve your car, you still can have a problem.

kch
Antenna Engineer,

other notes;
 "Ground planes" by definition are usually one or two wavelengths in size. That's 6 or 12 feet at 160 Mhz and 2 to 4 feet at 458 Mhz. Small ground plane antennas have odd asymmetric antenna patterns. What antennas are presently used?

I assume you are trying to get a good omni antenna pattern so that you receive equally from all directions?

Do you know the loss components of the carbon fibre?
What's a "tub", the internal structural frame.

Maybe I'm about to learn something: Don't patch antennas typically have a main beam that is normal to the antenna, thus your proposed antenna would be a 'cloud warmer'.

Antennas with small ground planes have much lower peak gain and can be very omni. Standard dielectrically loaded patches with one wavelength ground plane have less then one dB peak gain since alot of current goes under the groundplane and radiates downward. 3 wavelength ground plane patches get above +5 dBi gain (depends on dielectric value of course). The gain is a straight line from the 1 to 3 wavelength ground plane size based on an article I have. Having a small antenna and a small curved groundplane lowers zenith gain and gets alot more energy on the horizon. Patches are often made into horizontally polarized omni's by curving them in a circle. Bandwidth increases too for curves patches compared to flat ones.

Patches are very cheap and easy to make too. You don't have to etch. He can hand trim them. I think he only needs one or two.
kch

Thanks for the details.

We should get some additional application data from Mr. Beacon.

If his receiver is always in the In-Field, and the race car is always going in circles and turning left, then he could beam the signal out the drivers side of the car (so long he doesn't mind losing data during spins and crashes).

If he is transmitting 2 watts (for example) and he is losing the signal from a race car that is only 500 feet away, then it may be a multipath issue more than an antenna gain issue. CP might be a good counter-measure.

If it is an off-road rally car (probably not), then it is a whole other issue.

There are different optimal solutions for different situations.

VE1BLL, Higgler; Thanks For the Responce. Some futher background info.
Vehicle: Open Top LeMans LMP2 SportsCar, All of the Chassis and bodywork is Carbon fibre composite (no metal Roll Cage, all Carbon ). The car has no roof, think of a two seater F1 Car with body work covering all of the wheels.
Data Telemetry TX(Car); 5 Watts, 458.575MHz, 1/4 Wave Dipole, Mounted behind rear roll hoop. ( cant go any higher )
Data Telemetry RX(Base station); Two Dipole antennas, approximatly 2m long and seperated by 1.5m, feeds into a diversity reception modem.(To help with Multi-path issues)
Voice Radio RX/TX (Car); TX 7.5 Watts, 180 MHz, 1/2 Wave Dipole, mounted in centre front of car. ( Very limtied other location to mount )
Voice Radio RX/TX Base Station; TX 10 Watts, 180 MHz, Single dipole antenna, Approx 1.5m length.
Maxium Range; Varies but in general a line of sight max range would be 1.5-2km.
Ground planes for both car antennas (data/voice)are simply as much copper tape as I can fit underneath the relavent area, this is currently not much but may be improved during the winter redesign. The Antenna sits in the middle of the copper patch and clamps through to the outside of the bodywork.
Vehicle Orientation; The nature of the circuits means that the cars can be at any angle relative to the base station, many circuits fold back on themselves or even run in reverse.
We have considered using a Yagi antenna for one of the diversity antennas and pointing it at the weakest area.
Sorry, Dont know what the loss component of the carbon fibre.
I am not familiar with patch antennae, We have complex compound curves of bodywork and many limitations on mounting positions, this may make such an antenna hard to fit, possibly. I shall look more at this sugggestion so that I am more up to speed with what you are suggesting, its a new field to me Im afraid Cheers, John.

Carbon fibre is quite lossy however that shouldn't worry you.  A common way to provide a ground plane for 1/4 wave dipoles mounted on CFC is to embed copper mesh into the gelcoat, inside or outside as practical.  You are not going to be too worried about some loss with the ranges and power levels you gave but I expect the receivers are going to want a very good signal to noise ratio to cope with the huge range of fluctuations you will encounter in this work, so a ground plane is necessary.

If it proves very difficult to provide a 1/4 wave radius ground plane you will achieve good performance 1/4 wave long tapes radiating out from the antenna base.  Try for at least 4 tapes if you are forced along this route but a continuous plane is best.  Remember the purpose of the ground plane is to decouple the coax braid and is as much a matching issue as a radiation issue.

<rough quote> "Telemetry: 5 Watts, 458.575MHz, 1/4 Wave Dipole. RX: Two Dipoles, 2m long, diversity reception"

5 watts is tons. Make sure the Xmit 1/4 wave dipole is vertically oriented to provide omni-directional coverage. If it were horizontal then it would have nulls off the ends. I don't understand the 2m long dipoles. Dipoles at 458 MHz should be just about one foot in length. We need further info on that discrepancy. Also, need to confirm that the base antennas are arranged to provide omni-directional coverage.

<rough quote> "Voice RX/TX: 10 Watts, 180 MHz, dipole"

You could probably use a simple rubber duck antenna inside the car. Use an elevated (on a mast) vertical antenna for the base station to provide clear line of sight with omni-directional coverage. With that power and such short range, you should be able to use a wet noodle as the antenna at each end.

I'm still voting for the patches.
The 1/4 wave shorted one's at 458 Mhz in FR4 are 4" long, 2" wide and 0.060" thick.
similar antenna at 180 Mhz would be 10" long 5" wide and keep the same 0.060" thickness.
These are narrow bandwidth antennas, JB - do you know what bandwidth is needed for these antennas, i.e. how many MHz.

This may sound odd, but adhering them to the bottom of your metal chassis facing downward into the track may work the best. You'd have a metal ground plane via the chassis metal. Also, ground reflections won't enter into the equation and multipath off other cars may be minimized since energy can travel under adjacent cars when line of site is blocked.
Last year I did a similar design for the underneath side of a palate that had military ordnance on it, the antenna patterns were pretty good. If you lined up all the palates, the energy propagated down all the palates and communication range was long.

kch

I reckon there'll be some experimentation needed to get useful performance from an underbody antenna.  These vehicles only have a couple of inches ground clearance.  Look at the F1 cars which have whips on them.  

A 458MHz whip should be 156mm long (allowing for end shortening effects) and a 160MHz whip = 445mm long.  The ground planes would be as large as you can practically get up to a full 1/4 wave which is 164mm or 469mm radius.

I once organized a contest to see what range could be had transmitting through a bit of wet string.  High transmitter power proved to be a disadvantage if the transmission was to last more than a minute or so and the winning "lubricant" was vintage shiraz wine.

Brian R,
Do you have any recommended reading/websites/info about the RF loss of composite fibre.
I know there are alot of different composites out there and I'd like to read up on the subject.
Thanks,
kch

PS: JB, what's your gameplan - cost effectively experiment yourself or hire someone to build antennas for testing? a possible hint.

I haven't an online reference.  I was referring to specific data we have used in military aircraft projects where composite skins are used.  There is a large variation in loss tangent depending on the matrix used, the impregnation ratio and the structure of the carbon filler.  CFC is much too lossy to be used as a radome even at VHF frequencies.

Generally when a conformal antenna is to be installed that section of the skin is made from a structural plastic (now there's a good scientific term).  

There are other issues relating to the use of CFC, titanium and aluminium.  CORROSION is the main one with 2000/6000/7000 series Al high on the anodic end of the potential scale while CFC is right at the cathodic end with Ti not to far from it.  Joining Ti to CFC is not too difficult but when joining Al to either we employ serious measures to prevent the Al from being corroded away by electrolyetic action.

This is an interesting article I found on google:
 
http://www.me.gatech.edu/jonathan.colton/microwave.pdf

BrianR,
Thanks for the input. I thought that carbon fiber would be ok at the lower frequencies. Sounds like it could be a good microwave absorber.
That hints my idea of antennas to radiate thru the carbon fiber structure may be a bad idea and the under side of the car sounds more promising.  

Where does one get a sample of carbon fibre? I'd like to do some radome measurements and build some antennas on them to build my knowledge base.

kch

Hi guys,
Sorry about the poor responce. Car preperation has forced me to concentrate on getting the datastreams and sensors working correctly instead. The infomation you have supplied is very useful and I'll try to respond correctly on monday.
FYI, Regarding metal floors, the cars chassis is a total carbon composite construction and ground clearance is only 30mm. If your interested in an overview of the car I'm dealing with...
http://www.dailysportscar.com/free/teams/rml/mgex264.htm
Many thanks, John.

Well after seeing a picture of the car I can see the difficulties in antenna placement.  I suspect that central to the fwd cowl is a likely spot, one that won't get smashed off in a "touch" like a mud guard would.

Higgler, thinking more of what you said, what would be worth experimenting with would be the losses with a thin wall radome made of CFC.  If there is enough margin in the path losses it could well work.  There is also the EMC issue of the engine management computer to consider if there's a lot of radiation inside the body with nowhere to go.

"...enough margin in the path losses..."

Well, with 5 or 10 watts over 2km max range...

Bandwidth is probably 'xx kHz', not MHz. JB never mentioned video.

JB never mentioned what the issues were with his current antenna systems (why can't he reach 2km with 5 or 10 watts now?).