Now we apply the signals across a non linear device and see a much different output. The readout measure a signal of 1Khz. How does a 1000khz and 1001khz produce a 1khz outout. The non linear action of the diode procuces 4 output posibilities. They are the sum, difference, and 2 orignials signals.

In this circuit I simply apply 2 signals to a resistor and the produce an algebraic sum or difference of their amplitudes. The resistor being a linear device it has no effect on the signal phase or frequency.

Years ago my friend Andy built the 40/80. It is a good example of the dual conversion as presented in the block diagram. What I have in mind this time is a JFET design rather than the MOSFET. I bought a box of K2539's for $5 and I tend to use them first. I also have some J112 and J175 to try. My design tool of choice is LTspiceIV. Other tools may give good results but LTspice is an old friend. The next contender is QUCS "Quite Unerversal Circuit Simulator". I have used QUCS just enough to see the possibilty. It is available is several packages for Linux, Windows and Mac. Now things to consider: Should I use a DBM or active detector? Single conversion or dual conversion? Thrown together with a minimum of parts. etc.... look at the 40/80 here: ***** https://radio.radiotrician.org/2018/11/dbm-and-dual-gate-mixer.html***** It uses a DBM and dual gate mxer. DBM maybe but no dual gates this time.

10/10/25 ready to start a new receiver.

I am thinking a JFET direct conversion receiver. it will consist of : RF Amp Mixer AF Amp LO Osc *I will probable use a 2n3904 for the AF output. Building a single stage is not to hard, the troubles start when feeding other stages. We will be looking at loses due to Z mismatching and shared power supplies. I will use LTspiceIV for the sim's. (more on this later) I reloaded my old HP desktok with a new OS and the tools for the project. (more on this as we proceed - I may use a couple of other tools?)