I'm a recently converted SW engineer to EE (Graduated less than 3 years ago, BS in Comp Eng, focus on Digital hardware) and I've been made the hardware lead on a project here at work. The project is in the final testing phase, and the module deals with 3G cellular, with a custom designed antenna our in-house RF guy made up. My question is that we're doing some pretty intense testing, and I'm seeing all these graphs with SWR readings, and I have no idea what I'm supposed to be seeing. Typically, the Y-axis is 1 - 15dB, and the X-axis is Freq (usually in GHz).

An issue I know is occurring (from meetings) is that we're having spurious emissions on one of the harmonics, which is "bad." My question is "Why is this bad, and how can I not look like a complete idiot when looking at future graphs?" Should the dB be higher or lower when you're looking to attenuate? When working in the GSM signal band (we're targeting both EU and USA standards), what should the graph look like?

As a followup, instead, should I just smile and nod and let the RF guys handle this?

  • \$\begingroup\$ You must remember that a dB is a logarithmic ratio of two numbers. \$\endgroup\$ – dext0rb Jan 6 '14 at 20:49
  • \$\begingroup\$ +1 for "just smile and nod and let the RF guys handle this". Sounds like a good plan :-) \$\endgroup\$ – Anindo Ghosh Jan 12 '14 at 6:49

VSWR of a quad band GSM antenna with variation of the feeding point

I will quote out of my M2M Iot Cookbook:

Changing the position of the feeding point d has an influence on both - the resonance frequencies and the antenna matching. An antenna is matched if the reflection coefficient is less than -6 dB. Before I go ahead to explain the further graphics I will analyze the upper graphic more detailed. If the parameter S1,1 is 0 dB then all power generated by the radio module will be reflected to the module. For some radio modules this is a dangerous situation. The module could break and even if they do no break the rejected radio wave could force electromagnetic radiation inside the cellular module and outside at other parts on your PCB as well. Remember that the target was to generate up to 2 Watt TX power on GSM 850/900 and to radiate it to the air. The gap between the - 6 dB on the lower end of the curve and the higher end of the curve is called bandwidth of the antenna. As we can see the curve is crossing the -6 dB border in four times. The antenna is resonant in two bands. The two resonances are already covering the GSM 850, the GSM 900 band, the GSM 1800 MHz band and the GSM 1900 MHz band. GSM 850 and GSM 900 are very close together and overlap in the frequency range and with GSM 1800 and GSM 1900 it is the same. Please note that the European UMTS 2100 is overlapping with the GSM 1900 band. If you have a closer look to some quad band GSM antennas then they often already cover the UMTS 2100 MHz too. Anyhow, the antenna under test in this chapter offers not enough bandwidth to cover UMTS 2100. The biggest bandwidth we get at the position d of the feeding point X =3. A one to one copy of the origin documentation is maybe not the best choose. It makes maybe sense to move the feeding point of the antenna a little bit to the right. More about the optimization will follow later in this chapter. Be aware that a reasonable reflection coefficient only shows you how many power will not rejected from the antenna. If the radio wave will not be reflected the wave has two options. Option one is that the antenna will convert the power in thermal energy. Just replace your embedded antenna to a 50 Ohm resistor and detect the result. No energy will be rejected. The reflection coefficient will look perfect. If you measure the emitted TX power, then it will be zero. A good antenna will show a reasonable reflection coefficient at its data sheet and will tell the antenna efficiency as well. The antenna efficiency is the value of the generated energy in comparison to the emitted energy. The antenna efficiency is an important parameter and is a part of this simulation as well. Even more, the simulation includes the plastic enclosure too.

Just have a look on of the four curces. All curves are passing the -6 dB border in four times. - 6 dB is the same like a VSWR of 3. A VSWR is ok for an embedded antenna. For an external antenna the target is - 10 dB or a VSWR of 2. As told before the VSWR is an indicator. The important parameter is the antenna effiency.

See also Nothing will interfere your embedded antenna?

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.