mixer/adapter for swept analyzer?

[nenea dani]

Hello colleagues,
I wonder why most signal swept analyzers do not go below 9kHz or 5Khz the most performing ones. Is there any adapter to translate the 20Hz-20kHz spectrum for these swept analyzers?

[jjoonathan]

It's because the biggest capacitor in the high frequency capacitor kits from ATC and Johanson has a 3dB frequency of 9kHz when you drop it into the signal path as a DC block.

[nenea dani]

Why is a HF capacitor mandatory? Let me understand that, for example an high pas filter with 100 microfarads and 50 Ohms has a fcorner of 32 Hz? For compensation, there is a normalization function  in TG mode if I'm not mistaken.

[jjoonathan]

Competing priorities. RF instruments are made for RF applications which typically only care about DC-kHz for the purpose of providing power or flipping switches. Sometimes they need to tolerate DC, hence the blocking cap, but they generally don't need to analyze DC so they aren't good at analyzing DC.

Lower frequency applications exist, of course, so the idea is that you should use different equipment for this. Many techniques are practical at kHz that aren't practical at GHz so a good audio frequency analyzer will look very different from a good RF analyzer. Sound cards, high bit-depth oscilloscopes, dynamic signal analyzers, etc should cost less and perform better than trying to use an RF analyzer for something it isn't good at. If you really need both in one instrument, sure, you could upconvert, and you could also use a bias tee -- but make sure that you gain something in the combination, because otherwise these should really be two different instruments.

[David Hess]

The input coupling capacitor is one reason, but there are two more:

1. The frequency response of the first mixer only extends so low.  However for operation down to DC, the mixer ports can be swapped, although combined with removing the input coupling capacitor this makes the first mixer more susceptible to damage.

2. The increasing phase noise from the local oscillator makes lower frequency measurements noisy.

[nenea dani]

Dear colleagues, thank you for your attention. Please read my two posts in the section
 https://www.eevblog.com/forum/projects/front-end-for-a-swept-spectrum-analyzer/
I think you can help me with some advice. I built two versions, everything would be fine except the output level is too low.