900 MHz Long Wire Antenna - Oxymoron?

[Yashu]

I want to run a wire in PVC pipe under a racetrack to detect vehicles passing over it with 900MHz transmitters.

Does this sound insane? If not, what kind of antenna wire should it be?

Anybody know of another transmission medium which would be better (magnetic, SW, HF, ???)

 

[VEBill]

If I were you, I'd back-up a bit and try to find a better overall approach than the one you've described. DGPS beacons, IR emitters, bar codes read by laser scanners*, RFID tags*, etc. (*think: electronic toll booth systems)

If you must stick with the 900 MHz scheme for whatever reason, then a simple yagi (or corner reflector, or helical, or ...) antenna on a pole and pointed at the proper location would probably work much better than trying to use a buried longwire at these sorts of frequencies.

 

[Comcokid]

I think what VE1BLL indicated with a directional antenna aimed where you want to pick up the signal is probably the normal solution to such a problem.

However, if you must have a long-antenna-in-a-pipe solution for your application, then I suggest you look into what is called "leaky coax". This approach is used to provide wireless or cellular coverage in tunnels and underground parking spaces. I have no experience with this so I don't know what the limitations are.

 

[IRstuff]

You need to describe the requirements, not just what you think you want to do. Otherwise, you're going to get incomplete answers.

TTFN

 

[Higgler]

Excellent point IRstuff,

Give details Yashu,
What's your receiver sensitivity. What are you doing with the info. Describe whether the car has a transmitter or you're just detecting the car passing by, etc.

kch

 

[VEBill]

I've assumed that he wants to automate the racetrack, detecting when each car passes the start/finish line, probably to update the leaderboard. Or something similar.

Not sure where the 900 MHz came from...

 

[Yashu]

It's a race track thing.

Mission is to time stamp 2 wheel vehicles passing 'beacon' points on the track. Existing PVC about 18" down in ground already exist at 'beacon' points, but, are not necessarily deemed mandatory to utilize. Track is dirt and mud is everywhere.

Vehicles can conceivably have transceivers since this functionality will accompany the real mission of logging vehicle sensor data.

Max vehicle speed of about 60 MPH I guess. Shooting for ballpark of 10ms time stamp accuracy.

I know the popular technique is magnetic induction pickup, but, unsure of interference which may occur with a co-located magneto. Also, can't seem to find any info on commercially avail cheap mag ind sensors. except for auracomm.com (overkill for me).

So, was contemplating narrow band TDM bursts by the vehicles in response to a beacon sync pulse. Multiple vehicles over the beacon points is possible and would rather not miss any of this topology data. Envision RSSI tracking by the "beacon" receiver station @ this 100hz ping rate to more accurately determine actual vehicle location w.r.t. instantaneous crossing of the beacon point.

Grabbing log data across the 900 MHZ link is a plus, so, that's another reason I'm shying away from the ever popular magnetic technique.

Just fishing for novel ways of implementing system right now and not necessarily interested in re-inventing the wheel... of course unless there are substantial $$ savings to the customer.

I suppose 900Mhz signal loss thru 18" of wet mud will be substantial.

Any ideas are appreciated.

PS can you guess the vehicles?

 

[zeitghost]

Supercharged lawnmowers?

(I watched a lawnmower race on the tv yesterday...).

 

[IRstuff]

I think that you're better off with GPS, particularly some of the newest systems. Hemisphere GPS makes an OEM board that provides heading at 20 Hz and position at 5 Hz for under $1000 in single quantities. For $500 additional, the boards will do 20 Hz positioning as well. The datasheet claims 2.5 m @ 95% confidence for real-time non-differential operation. Real-time or post-processed differential performance is supposed to be better than 0.5 m @ 95% confidence.

They claim < $600 for quantities > 100 for the basic board.

http://www.hemispheregps.com/main/

We haven't actually bought one yet, but they appear to be substantially cheaper than a comparable system from Novatel.

The cool thing is that you could probably tie these boards into a PDA for datalogging and get the entire event datalogged. GPS also gives you speed and altitude, which might be of interest to your clients. Altitude performance specifications are published, so they're probably not terribly accurate, but having the time-history of a particular vehicle over the duration of an event might be invaluable for detecting areas for vehicle/driver improvements.

TTFN

 

[Yashu]

IR,
Thanks, but, WAY overkill/pricy. I can do a TI TRF6903 design for <$15 in parts and figure $10 for prepreg PCB.

Shooting to sell this thing for <$200, so, I figure I got to do it for $50.

Use of a leaky feeder coax over beacons @ $1/foot sounds OK.

Not interested in paying a company for what I can design myself.

 

[Higgler]

What about two antennas located end to end in one tube. If you place two dipole antennas end to end, space them in the tube ?2 to 4 feet away from each other and let the RF bounce off the car from one antenna to another. AC couple the detected RF signal to look for changes when a car passes over. It would make a baseband pulse change when the car enters and leaves the proximity of the antennas.
You may be able to do it with a simple CW transmitter and detector, may need to choose a lower frequency to penetrate the soil though. A more sensitive heterodyne receiver would be the next step up.


kch

 

[BrianR]

I don't think you'll get useful performance at 900MHz. Ground penetrating radar typically uses 20-150MHz and several Kw peak power and in wet clay that will not penetrate more than a metre or so.

I think your best chance is to use 40-100KHz with inductive transmitters on the bikes and a series of pickup coils buried in the pipe under the road. The transmitters could be modulated for individual ID.

 

[BrianR]

I should add that the pickup coils should be resonant at the transmitter frequency.

 

[VEBill]

I'm guessing that you need to put more thought into the basic operational concept (or at least tell us). Assuming you go with beacons (at whatever frequency works best) and you use a leaky coax under the track (or something similar); then how do you propose to determine the crossing time to within 10ms?

60 mph x 10 ms = 10.56 inches. My point is that you can't rely on detecting the signal to determine the position to that precision. Perhaps you could use Doppler shift, but you can see where that is going with complexity. Perhaps you could use two detectors and mark the equal amplitude.

What if two vehicles cross at the same time? If there beacons are free running then they might collide. Are they going to have intelligent TDMA?

So, what are your conceptual plans for detecting the crossing point in time to within 10ms?

The reason I ask is that I honestly can't see how it could be done on the cheap...


"...2 wheel vehicles....PS can you guess the vehicles?"

Hmmmm... Is that a trick question?

 

[Yashu]

Each vehicle is preconfigured for a 100us wide TDM time slot to respond in.. in response to the beacon sync pulse.

I figure about 100 time slots.. thus the 10ms dwell period.

Also, considering generating an RSSI curve for each vehicle (at this 10ms sample rate) and 'post processing' it [probably with some kind of simple digital LP filter] to search for peak RSSI which will dictate more accurately beacon crossing point.

Not sure if multi-path will be an issue here, but, use of over-the-track leaky feeder and LOS path to vehicle's transceiver should minimize fading.

PS: not a trick question... clue... sounds like a busy bee-hive at the start of the race.

 

[VEBill]

Guess: R/C motorbikes (riderless).

Your concept sounds reasonable.

I wouldn't bother trying to use 900MHz from 18" below ground. Probably a waste of time. See if you can get the system working with a simple (above-ground) antenna, and then you can optimise the antenna system a bit later.