İmage Rejection at Low IF

[cmkrkmz]

Dear Friends,
I need a solution for downconversion, I need to reject the image for low IF frequencies (about 1 MHz IF). I could not find image rejection mixer for such a low IF frequency, and so my noise figure for my entire system increases about 3 dB. Do you have any solution?
Regards

 

[G4BCH]

Design your own image rejection mxer with the output quaradture coupler centred on your IF.

 

[cmkrkmz]

Dear G4BCH
How can i design such a quadrature at such a wide octave and low frequency,
my IF band should between 300 kHz-20 MHz
regards

 

[FvM]

"between 300 kHz-20 MHz" is not what we expect for "about 1MHz". You'll better give a full specification.

 

[biff44]

you either need a image reject mixer configuration, or a 1st if at something like 10.7 MHz, bandpass filter, then downconvert to 2nd IF of ~ 1 MHz.

 

[mikerf]

There are various approaches

1. Use a Zero-IF approach. (quadrature mixer); You'll end up with I & Q at baseband which will need capturing, almost certainly with a pair of ADCs and a bit of software to pull out the correct sideband

2. adjust the architecture so your IF increases by 10 MHz or so, thus you now have a suffuciently wide frequency range between your wanted and your image that a filter prior to the mixer can allow you to swamp the mixer noise figure with whatever amplifier preceeds it.

3. build an image reject mixer yourself, but use two high frequency 90 degree splitters, one for your LO and one for your RF. Traditionally you get a 90 degree splitter on either the LO or the RF and do the other 90 degrees in the IF, which is bad for very broadband applications. The up side is that you can then choose easily which sideband you want and can even allow it to switch actively, great for direct modulation/demodulation. In those occasions where broadbanding is required a 'zero IF' is usually chosen as 'perfect' 90 degrees from two ADCs/DACs is available. But you can put that 90 degree shift on the RF instead, then add a zero degree splitter/combiner after the mixer at the IF... it won't be as high performance as possible, but you don't need that, 6dB or so of unwanted rejection (two lots of HF 90degree splitters mean double the phase/amplitude errors) is good enough if it's only the image noise you are worried by.

 

[cmkrkmz]

Dear FvM, biff44 and mikerf,
Thank you for your posts. Fvm is right, I should be more clear about my IF range, Around 1 MHz is a wrong expression, sorry for that.
My IF frequency should be between 300 kHz-20 MHz, so you can see that it is very large octave. So there is no quadrature- 90degree shifter that can involve 300khz to 20 MHz, I am not even sure that it can be done.
Do you have any ideas guys?
Best Regards,

 

[FvM]

A differential 90-degree phase shifter can be implemented for a wide range, but it's a complex filter. Several other options have been already suggested by mikerf.