Day/evening coverage

[Iden]

I have an application requiring two-way RF data communication at a location in Southern United States.

I am operating 1mW at abotu 900MHz. Communication over a short range (few hundred feet) can be difficult during the day, however, at night the range is far more than required.

Any clues as to why this may happen. Is it synomonous with the day/night transmission of the lower bands (CD and AM radio), where AM radio transmitters in Canada must actually be directed north during the night? Can anybody explain this?

 

[iroquoix]

Relatively low frequencies like mediumwaves (0.5-2MHz) or shortwaves (2-30MHz) are reflected by the ionosphere. This does not happen to waves in the UHF range (>500MHz) under normal circumstances. Your application is a simple line of sight application where you possibly may have reflections by more or less conducting material (walls, roofs, cars ...).
I suppose that the mentioned effect has its reason in manmade noise which is quite a lot higher during daytime than at night. High speed LANs and WLANs, computer equipment, wireless communication equipment and such modern gear raise the noise level. Most people stop working at night. Watching the noise level with a spectrum analyzer will show this.
Many receivers of data communication equipment have very broad input stages. So if some other wireless communication equipment is operating in the same area, even its transmitting frequency is quite different, it will lower the sensivity of your equipment. This could be prevented by using good filters between antenna and the data-transceiver with a high selectivity.
Just a small story from my job department: A guy who is working with me bought a new car which has a small key-chain remote control to open the cars doors. This remote control is working on 433MHz. Another guy from my department is a ham who started transmitting on 439MHz to make a little joke when the car-owner tried to open the car to drive home. The car did not open its doors! He was very astonished and we had something to smile about.

 

[Comcokid]

Sounds like you are using some device in the 900-925 MHz ISM band. DD3LY is right - there are no long range propagation issues at this frequency. This is especially true given the simple nature of ISM band devices - the receivers are generally not extremely sensitive, the transmitters not powerful (1mw for yours), nor will the antennas have much gain.

Common interference at this frequency is from cell phone towers in the 800-890 MHz range, 930-950MHz range, and cordless phones which share the same band.

Also look for something which might be changing in the near/direct path between your devices - i.e. a large metal garage or bay door which closes as business shuts down in the evening, Fence gate, semi truck/trailers being parked, etc. You don't have much path margin to fool around with.

 

[Iden]

Thanks for the comments. As I thought, there is no propagation issues in this band. Thank you for confirming that.

We have other versions of this device that operate at about 870MHz. Due to the Cellular band, we cannot use them in North America. As you would expect, the cellular interference swaps us.

The receive device actualy has a low noise floor at -100 dBm.
Iden

 

[iroquoix]

The sensivity (or noise floor) of a receiver does not express how good it wilbe able to receive a small (wanted)signal in the presence of a much bigger (unwanted) signal which is in the same frequency range.

If you make a walk with a friend on the beach during storm and the waves are rolling towards the coast you have a very limited range if you just talk with normal volume to your friend. During silence (the sea is calm and "looks like a mirror&quot your friend will be able to go away several meters and will still understand what you are talking about...

 

[photistor]

Typicall issues which effect day time vs. night time tranmissions are multipath, however I believe fad margins due to thermal effects can still be an issue at 900MHz. Since the distance seams to be short one would hope that fade margins due to thermal effects would not be your problem. The angle between the Tx & Rx should be looked at with respect to interferrence. Do you have unobstructed line-of-sight between the TX & Rx or are there RF blocking features in the way? I agree with DD3LY, using a spectrum analyzer will help troubleshoot the problem.

I know they may be overkill for your application but here are some RF link papers to refer to:

http://www-classes.usc.edu/engr/ee-

ep/475/Radio_link_design.pdf

http://mendel.ee.washington.edu/rainfall/lec10.pdf

http://staff.aub.edu.lb/~webfea/courses/EECE503/Handout 7 - Cellular coverage-1.doc

 

[FAA]

Most likely the problem with the device is that it does not have enough receiver selectivity. Selectivity is a receiver jargon term that basically tells you how much your receiver can reject a signal in an ajacent band. This quantity is measured in dB and typically needs to be quite high. For example cell phones require about 70-80 dB of rejection. This means an adjacent signal will be attenuated by 70-80 db compared to the desired RF input. Selectivity is achieved by bandpass filters in the RF front end of the receiver. Typically the cheaper the filter the less rejection you have. So I beleive the problem is mainly with the comm. device that you are testing rather than the RF radio environment. The RF environment is a problem but a well engineered receiver will be able to compensate for radio interference. The bottom line is if you want better performance you'll have to pay more for better hardware

 

[buding]

Hi!

I would like to add one more point of view. Antenna used i such a device is always pointed, in some way, towards the ground. And, as everybody knows, it gathers the thermal noise from the ground. During the night situation is much better. To avoid such a problems You have to make good design on RX side ( low noise, filtering, etc..

Other conclusions from above posts are also correct and it is very hard to estimate what has the biggest influence. The best way is to take Your equipement and test it in totalt different enviroment.

Good luck,
Buding.