QUESTION

Question

Hi. Any ideas appreciated? I've got access to some basic RF testing equipment.

Answer 1

It shouldn't be too difficult to pin down the source. If you have access to a spectrum analyzer, a low-noise amplifier and EMC probes, then you'll be in business.

Put the device in an isolated environment. Use a shielded enclosure if you have one.

First, check for spurious emissions with the transmitter OFF. If you pick anything up, then you'll know it's being caused by an oscillator or switching power supply. If that's the case, you'll want to change your layout to accomodate these. There are other ways to shield against it… more on that momentarily.

If you don't spot anything with the transmitter off, try with the transmitter on and push traffic through the transmitter. This should highlight the source if it's not being caused by another radiating part.

And when you've found the source, either change your layout, look at ways to improve your shielding or design-in some more signal filtering.

Answer 2

What kind of measuring equipment do you have? In which frequency range do you want to measure? What interference do you already have in your device?

Beside a spectrum analyser you need measuring antennas. Such loop antennas can easily be made by yourself. These antennas are then narrowband. You then move the loop antenna over the board and can detect the location of the interference.

I would not switch off the radio module. It is usually enough to terminate the 50 ohm line with 50 ohm or 2 x 100 ohm in parallel. Everything which is measurable then still at energy comes in some form from your circuit board and/or your equipment.

Remember that 33 dBm transmission power with GSM can quickly lead to mixed products with other components on the board. Each semiconductor can also act as a mixer. A good example from the past is the security electronics in a battery. The battery was not designed to operate GSM modules and the semiconductors in the battery acted as mixers. The customer had harmonics in frequency ranges that were not multiples of the transmission frequency.

The error in the battery was analyzed with a remote GSM antenna. With the antenna removed, the error had disappeared. The battery was replaced with a laboratory power supply on a trial basis. With the exchange of the battery the error was eliminated.