RF filter => real need ?

Thread Starter

cam0

Joined Mar 23, 2010
25
Hi,

I am currently trying to build a frequency detector for UHF. Until now, my design was somthing like:

Antenna -> HF filter -> Amp -> rest of the system (detector)


But actually, I was wondering why not using the superheterodyn principle to downgrade the frequency, because I am having a hard time trying to find filters for these frequencies (I found SAW filters...). I am not afraid of aliasing, because the signal that will be transmitted is around 9600bauds (and I probably won't use the signal). So the new design would be:

Antenna -> (Amp) -> mixer -> IF Filter

Nothing new here, this is just the usual theory, but to me it has only advantages:

  1. filters are easy to build / find for the lower frequency.
  2. No aliasing in case I want to use the signal.
  3. I can use a basic oscilloscope to view the signal.

Am I right ? The rising question is what should I use to generate a signal at around 300MHz ? I found VCOs having very wild frequency range, but actually I am fearing not being very precise with this kind of device. Also I have a dumb question: the signals (from antenna and from oscillator) going in the mixer must have the same amplitude, right ? so is this something do I have to care about ? Finally, would it be a good choice to have a chip doing (mixer + oscillator) all in one ?

Thank you

c0
 
Last edited:

Thread Starter

cam0

Joined Mar 23, 2010
25
Hi Bertus,
actually, I am taking the ideas from Ham. They propose to put a HF filter, but sincerely, AFAIU, it seems to me we don't really need it, we can just mix the signal and filter it afterwards.
 

bertus

Joined Apr 5, 2008
22,270
Hello,

When you do not use a filter at the entrance of the reciever,
you will recieve two frequencies.
Fosc + MF and Fosc - MF.
For example, you have an oscillator of 310 Mhz and a MF of 10 Mhz,
you will get the signals of 300 (310 - 10) and 320 (310 + 10) at the MF output.
You see, the input filter does not have to be very sharp, but must reject the mirror frequency.
The higher the MF frequency, the easier the filtering is.

Bertus
 

Thread Starter

cam0

Joined Mar 23, 2010
25
Yep, it makes sense. But what raises my question is that usually, ham put RF filters in the RF stage. Actually If I am coming in the mixer with a frequency of 250MHz, it will be trivial to filter it in the IF stage. So finally, what's the goal of this RF filter since we can do easier processing in the IF stage ?

Thanks !

Edit:

I am not sure that I am very clear hear. Basically, if I don't add a RF filter in the RF stage, I will have - let say - a whole range of frequencies from 0 to 500MHz for example (white noise). Then, if I provide to my mixer an oscillator frequency of 280MHz, the frequency I'll be interested (initially 300MHz) will be dropped at 20MHz. The unwanted frequencies at 500MHz will go at 220Mhz and 780MHz and the frequencies near 0 will go around 280MHz. Thus, here I could use a narrow bandpass filter for the 20MHz in the IF stage, they are pretty straightforward to do. So what is the benefit of pre-filtering the RF ? Ok, I'll have a hole in the IF band, but since the amplitude of the interested frequency will be much greater than the "noise" (I call noise the frequencies that are different from 300MHz [the frequency we are interested in] in this example), there will be no big deal here. Again, my assumption is that the amplitude of the wanted frequency will be bigger than the noise in the IF stage, that's why I am thinking about removing the RF filter.
 
Last edited:
Top