868MHz propagation question 1

[dgleeson]

Hello All

We have a propagation issue that is confusing us. We have two 868MHz modules on test, each capable of 500mW transmission.

We are testing to find the range of communication. We are testing them as follows. One remains in our office while the other is taken to the street outside. we are communicating charecter strings such as "Im at the lamp post number 1" etc. There is somebody at each end sending text strings in a continuous conversation. There is no protocol, error checking or error recovery.

What we have repetadly found is that the radio in the office can receive long after the radio in the street has stopped receiving. (The person with the radio in the street is moving away from the office.)

The radios are the same and the power supplies to the radios are the same. Indeed switching the radios gives the same effect.

The propagation paths between the two radios are the same in terms of distance. However the radio signal transmitted from the office travels through walls first before then traveling through open space. Its the opposite for the radio in the street, firstly traveling through space and then through the walls in the office.

Can anyone shed light on this phenomonon?

Many thanks

Denis

 

[pstuckey]

One possibility is that the radio on the street is moving closer to an interference source which is desensitizing it. Depending on what part of the world you are in, there are GSM bands, ISM bands, and paging bands near 868 MHz.

Peter

 

[VEBill]

Peter has identified the often-overlooked asymmetry (local noise) in what would otherwise be expected to be symmetrical.

 

[dgleeson]

Hi Guys

Thanks for your input.

We did think we had identified an interfeering source in one direction (when moving away from the office) so we started moving in the opposite direction. We found exactly the same effect.

We can change channels, so we will try this and see if we can move outside the interfeering source.

As far as GSM bands go in Europe 890–915 MHz OR 1710–1785 MHz So these shouldnt be our problem.

ISM is where we intend to operate and the greatest band occupancy should be 10%. So we should get communication even though we would get some interfeerance.

Paging - well this has very much gone out of fashion in Europe. But even if it was the cause then wouldnt it be intermittant communication. We have a constant problem.

Any other ideas.

Thanks again

Denis

 

[Comcokid]

I've done a lot of ISM (915 and 2.4 G) low power radio module testing. We always achieved repeatable results on a flat straigh strech of highway. Radios were tested 4 ft off the ground. A fixed number of data sets were taken at each distance interval and averaged.

Results were always statistically repeatable, including drop-out at longer distances due to ground refelections.

In your test setup, is the "radio in the office" connected to a computer, and the "radio in the street" is not? If so, the cable from the computer to the radio (serial, RS-232, or USB) may add to the antenna gain.

 

[biff44]

sounds like you need to put a narrowband bandpass filter at the input to your systems. It will filter out the out-of-band noise that often messes up reception.

 

[VEBill]

As hinted by biff44, the AGC loop is often based on the IF and is therefore a wider bandwidth than the rest of the receiver. Thus, interference from nearby channels can desensitize your receiver with no explicit evidence of interference at the receiver output.

These are the sorts of perfectly logical explanations that can explain the sort of asymmetry you noted.

 

[BOMBOVA]

Denis :
If you were in a screen room, with both modules, you would be in a controlled environment. You would be powering with a common power source and antenna link. Recriprocity would be exact and the laws of physics would be secure. (and,you would have a smile on your face.) Now, as you say, your environment of your units are different. One is fixed in the office, and the other is a wandering , out and about.
Well, now , lets think this through! The office is using a standard antenna, a power source and its semi shielded rf quite. The wander unit has a power source standard antenna and is in a variable rf noise locations. Interference, spurious noise, possible power supply noise, and fringing of the moving module, are your problem. I am actually surprised you have observed the results correctly. Subjective observation, often fails to recognize less than ten db of desensing. You may actually be as close as 6 db, but lets not forget the capture principle, and stick with ten. So there you have it. Pete, was rite on. Just remember it is a brutal world out there on unregulated frequencies. Make sure you have good margins in your path loss, say a ten db buffer over your full fidelity signal reception. and if not, well get the filter, for the wander and have a smile for what you get. Remember, we are not reinvneting physics, we are just getting the best out of it. " you are this savy an observer, your going to be one hell of a good rf system planner.", cheers, rtg yvr ca

 

[Higgler]

There's a law of physics that says you have the same RF loss going either direction9assuming you don't live in ferriteville). As an antenna engineer, it doesn't matter which antenna transmits or receives, you have to measure the same result in a passive system.
The only time I've seen a difference is when energy leaks around the antenna and gets into the receive circuitry, i.e. measuring S21 on a network analyzer between two antennas can make a big difference if the transmit antenna is close to or far away from the analyzer.
I'd go with the noise environment being different in the two locations, makes the most sense.

kch