HF General Purpose FET Mixer File: http://groups.yahoo.com/group/soft_radio/files/SDR%20Spice%20Analysis%20Files/HF_General_Purpose_FET_Mixer.asc
This is a low noise FET version of the mixer concepts I have been working on. And its harmonic content is suppressed a little more. And its bandwidth response is somewhat flat on up to 28 MHz. In fact it tends to dip up in gain slightly at 28 MHz for a slight peak there and by 30 MHz it is back down around the level of the rest of its bandwidth (via AC Analysis plot). This also indicates that the noise is up a little on the far end of the spectrum. Which is the way the noise plot looks. The FFT plot shows that the 30 MHz region is its -6 dB bandwidth point, and so the AC Analysis and FFT Analysis never agree the same on the bandwidth. It is not as wide as the previous model, but this is for HF only, and does the job in the software analysis. I felt that something that would be usable in a transmitter as well as a receiver was more like the vision I want the circuit to have. And so I want to keep as much of the upper order harmonics down below 60 dB as much as possible right from the start. A 30 MHz harmonic has to be watched however. It uses the 2N2608 JFET transistors. And has one 3 element (Pi) low pass filter section at its output. You might be able to use one filter on the mixer for 80 to 20 meters. And so it can be a 3 element or five element type whichever you prefer. And then on the upper bands use another filter on the mixer. Wide band mixers will have harmonic outputs, and can have lots of noise. But lowering the noise tends to lower the harmonics a little also. Low noise up around 15 to 30 MHz is important on shortwave. In turn, since this mixer is flat from 1 to 28 MHz, it would be nice for a receiver section, and since it has low noise, even nicer. This model of the mixer has a noise plot also, so you can see the noise in nano volts per Hz 1/2 terms. Right click through the plots other functions and call up the dB scale. Which should put you in the range of analysis of -160 dB and below. In this mixer you do not want to add any other ferrite beads since this will change the response and can even attenuate some of the HF band. The existing ferrite beads are from 1.5uH to 2uH. I guess this would be classified as an exotic RF mixer circuit. If you want, you can add in a 30 MHz notch filter at the end of the mixer before the Pi network filter. And run that in LTspice.
