Page 1 of 1

zeor or low if

Posted: Thu Dec 01, 2016 5:47 pm
by vy73
What are the pros and cons of zero if and low if ?
Thanks vy73

Re: zeor or low if

Posted: Thu Dec 01, 2016 9:22 pm
by DaveB
With zero IF you have the problem of mirror images of strong station on one side of the centre frequency showing up on the other side of the centre frequency. To minimise this the I/Q channels need to be very carefully balanced. Current version of the SDRPlay API have this built in under the Advanced options - you need to tick the box to enable post tuner DC/IQ compensation. There is also a DC centre spike that can cause a problem but is also largely dealt with. Also phase noise - but I don't know whether this would really be a big problem unless you are operating with very weak signals in a very electrically quiet environment.

With a low IF architecture you have no zero IF mirror image, no centre DC spike and no phase noise. But what you do have is a low IF image - in the case of the SDRPlay with a 2.048 mbps sample rate you will find an image 4.096 MHz away from the wanted signal.

I found that this image was around 45dB down from the main signal - and made low IF a poor second to a properly compensated zero IF mode.

I think the low IF mode is there because it was part of the architecture of the Mirics chip used in the SDRPlay not something specifically designed into the SDRPlay


Re: zeor or low if

Posted: Sat Dec 03, 2016 5:27 pm
by vy73
thanks for prompt reply .Yes zero If has a problem with the mirrors around the center frequency , it is much like the problem the ancient regenerative receivers and the direct down
receivers had , but just alike . In analogue times SSB was generated by the phasing method or more like what you have in SDR now it is Weavers method generating SSB . Putting the system around you have an imageless receiver , or you could use it as an imageless mixer .
Both receiver and mixer do depend on exact balance of the oscillator and its exact 90 degree phase shift signal .
Therefore both systems Low If and Zero If require same accuracy of the I and Q signal path .
There would be no difference whether the unwanted sideband occurs in the Low IF and afterwards brought in baseband or in case of Zero IF when it is directly transferred into baseband . So what advantage remains is the LO IF system has no DC spike in the intermediate frequency .

With the Lo IF system : the aliasing occurring due to sampling frequency should become removed by properly down sampling and CIC filtering .

Anyone out there helping me with profound literature especially covering the LO IF pros and cons ? Appreciate your help -