The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> FM receiver sensitivity
https://designers-guide.org/forum/YaBB.pl?num=1185306772

Message started by wag_master on Jul 24^th, 2007, 12:52pm

Title: FM receiver sensitivity
Post by wag_master on Jul 24^th, 2007, 12:52pm

I'm designing an FM receiver so I looked at available IC's to see what the state of the art is performance wise. The Silicon Labs Si4730 has sensitivity = 2.2uV EMF at (S+N)/N = 26dB. If my math is right, 2.2uV = - 102 dBm in 50 ohms. Since the bandwidth of FM is approx 200kHz, the noise in that bandwidth is -174 + 50 log10(200k) = -121 dBm. So even with a noiseless receiver, the best SNR I can get at the minimum signal level is 19 dB, but the Si Labs part is claiming 26dB!

Any suggestions as to where I'm going wrong?

Title: Re: FM receiver sensitivity
Post by mg777 on Jul 24^th, 2007, 9:21pm

The emf spec is peak-peak. So 2.2 uV emf into 50 Ohm is actually -109 dBm, that's the 7 dB difference.

M.G.Rajan

Title: Re: FM receiver sensitivity
Post by wag_master on Jul 24^th, 2007, 9:28pm

That makes the best possible SNR 122-109 = 13 dB, instead of 26 dB.

I found the answer though- I didn't know about FM capture, which only needs a couple dB to lock on and reject noise and weaker interferers.

Title: Re: FM receiver sensitivity
Post by aaron_do on Jul 24^th, 2007, 10:13pm

Do you have the link to this FM capture? In fact the spec was 1.1 uV at 50 ohm so its even better. They quoted a test condition of (S+N)/N=26 dB. My initial thoughts are that their test condition is using a sinusoid not a 200 kHz signal but since you mention FM capture...

Aaron

Title: Re: FM receiver sensitivity
Post by didac on Jul 26^th, 2007, 10:58am

Hi to all,
FM wideband modulation was the modulation that I hate most when I studied Signals and Systems. It's the less straightforward modulation for hand calculations, and only you can manage it using a cosinus as a information signal.
The FM SNR is very tricky in fact due to the spreading effect, so it's not very surprising that numbers don't match. The expression for SNR at the demodulator input it's:
SNR_D=0.75*(B_T/W)^2*P*γ,
where B_T is the transmitted bandwidth(150KHz,200KHz it's the channel spacing for security reasons for a modulation index of 5),
W=audio bandwidth(15KHz),
P=power at the input of the demodulator
γ the equivalent SNR for a Baseband transmission.
With that datasheet is difficult to provide any number of their internal blocks, but it looks like the signal path for both AM and FM are common so picking their sensitivity for AM and assuming a transmitted bandwidth of 30KHz and a SNR=26dB I calculated a NF=21.814dB.
Regarding FM capture this effect relate to a close interference(in band interference),not noise(it's equivalent to say a C/I intrachannel) due to the fact that what matters in FM is maximum frequency deviation,closest interferences are less harmful that further ones(if they are of less amplitude that the desired signal).

If you are interested in a detailed analysis of FM characteristics take a look at the excellent book:
Communication Systems, McGraw-Hill, 1968,A.Bruce Carlson which is for my taste the best book on analog signal theory.
Hope it helps,

Title: Re: FM receiver sensitivity
Post by didac on Jul 26^th, 2007, 1:45pm

Errata:
using Carson's rule B_T=180KHz for m=5 not 150Khz, sorry for the mistake.