Hi I want to implement a narrowband FM transmitter and I find all features about power and noise in the FM Modulator datasheet, but nothing about bandwidth.

The transmitter will work at center RF frecuency of 408.01250 Mhz being the channel bandwidth 25 kHz (408-408.025 Mhz).So the modulated signal would have a 25 kHz bandwidth.

I'm been searching for weeks the FM modulator in many manufacturer websites and finally I found this one with integrated PLL.

But the thing is that I can't find the bandwidth, I find some Frecuency Deviation tables and Modulation depth but nothing to work around with Carlson Rule or something similar.

Do you know where are located the related bandwidth specs on modulator datasheets like this one ? Maybe it can't be used for FM narrowband because it seems to have a Deviation Frecuency around -/+15 kHz.


The bandwidth of FM is in principle infinite. Filtering FM signal to some limited bandwidth causes distortion. You seemingly know Carlson's rule, which is a thumb rule for the needed bandwidth when one wants smallish distortion. I have seen also B=2*(fdev+2fm) for ultra low distortion. These really are empirical rules which give a starting point for simulations or other more exact ways to test which system fullfills your distortion and noise requirements.


  • Decide how high frequencies you really must modulate and detect. Speech is intelligble even after limiting it from 300Hz to 3,3kHz.

  • Find how large deviation you need to push the noise down (you can extract the original modulating signal with low noise from noisy FM signal if you have large enough deviation and the S/N gets over the FM treshold

  • You can get the wanted deviation if your modulating signal has the needed amplitude (0,5Vpp/15kHz; taken from the datasheet)

FM theory is complex. With Carlson's rule calculations are quite simple, but definitely unexact. An example: If you have 3,3kHz max baseband signal frequency and 25kHz channel space, you can have 9,2kHz peak deviation.

Can it carry your signal with low enough noise needs S/N data of your radio system and some spec of the wanted performance. You also must define the needed hf emphasis.

I recommend some measurements with a built system if, possible. Another way is to do simulations. If it doesn't work in math, it cannot work in reality. Succesful simulations still do not quarantee the real functionality, but give some hope.

  • \$\begingroup\$ Yeah I did it in Matlab but I coudn't find a microphone working at these frecuencies.Anyway thanks, just wanted to get the first aproach to real modulator devices and got it thanks. \$\endgroup\$ – EduardoG Dec 1 '17 at 21:58

From page 4 of the datasheet:

You want ±12.5 kHz, so nominal 420 mVpp on VAFIN.

  • \$\begingroup\$ But this would be the frecuency deviation really? not the bandwidth of the modulated signal \$\endgroup\$ – EduardoG Dec 1 '17 at 16:25
  • \$\begingroup\$ I mean if I apply the Carlson rule D=2*(fdev+fmod) the Max fmod would be 0 Hz when a microphone usually goes from 20 Hz to 16 Hz \$\endgroup\$ – EduardoG Dec 1 '17 at 16:28
  • \$\begingroup\$ Will you directly apply the audio to the modulation pin? Or will you boost the high-frequency part of the Audio? \$\endgroup\$ – analogsystemsrf Dec 1 '17 at 17:23

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.