[linrad] Re:Mixers

[Lawrence Stoskopf <stoskopf@xxxxxxx>]

No real expertise here in electronics, but 15 years or so ago I did a 
fairly complete literature survey dating back to the early telecom 
studies from the 30's on.  There are a couple of issues with diode mixers:  

One is pretty easily dismissed:  When the diodes are either on or off, 
there is significant current thru the diodes and the small signal 
current adding and subtracting from the LO current is typically not 
significant.

During switching you want as fast a rise and fall time as possible as 
the signal voltage does modulate the switching current in that interval. 
 You want as high switching power as the diodes will tolerate to change 
the (sinusoidal) voltage as fast as possible to get out of that 
interval.  The diode then saturates and all is well.  That is why the 
higher level mixers have diodes in series and also resistors to limit 
the current during saturation.

Now at low frequencies, our faster digital circuits have fast switching 
times and operate from low voltages to saturation. Obviously at VHF, the 
transition time becomes a factor, not the on and off times until the 
waveform begins to look like a sine wave.

But I think there is an added factor.  With the FET or switch type 
mixers, the LO current is not in the signal path.  Or is it reflected in 
the turn on/off resistance change?  Off hand would think this is a 
different breed of cat.

The real alternative is the old beam deflection mixer such as the 7360 
where the signal current really never changes and is beam deflected from 
one plate to the other with almost no change in beam current.  Of 
course, it has other problems, can you say microphonics and heat.

Someone straighten me out as needed.

N0UU

 


LINRADDARNIL