[R-390] Multicouplers - active vs. passive

[2002tii]

Lester wrote:

>The obvious reason not to use passive is the degradation in Noise Figure due
>to the splitting loss. In situations where ambient noise is significantly
>above the receivers front end noise(NF), then passives would be practical.
>It clearly is difficult to build a wideband amplifier that is superior to
>the TRF front of the 390s, in terms of IM/Distortion performance.

If you build an RF amplifier like the one in a TenTec RX340 (not the 
switchable BFG16A RF preamp, the RF amp comprising six J310 FETs in 
push-pull parallel) [schematics are available on the web], you will 
get 6 dB of gain with input-referred noise approaching -130 dBm in a 
3 kHz bandwidth, harmonics around -80 dBc at 100 mV input (-7 dBm), 
and enough headroom to shrug off nearly anything that doesn't melt 
your antenna.  Run with a 12 V supply, the 1 dB compression point is 
greater than 4 Vrms input (+25 dBm).  Follow this amplifier with a 
passive splitter, and you will have a state-of-the-art 
multicoupler.  System gain will be ~ +3 dB with a 2-way splitter, 0 
dB with a 4-way, and -3 dB with an 8-way.

I know of no commercially-available multicoupler, past or present, 
that can match this performance (of course, I have not seen every 
commercial multicoupler ever produced).

For anyone who wants to build one:  I typically select FETs for Idss 
of 45-50 mA at 12 V (this needs to be pulse-tested with a duty cycle 
~30%, to mimic in-circuit operating conditions and prevent 
overheating the FETs) -- but 6 unselected FETs should work fine 
unless you are very unlucky in the draw.  L59 and L60 are unnecessary 
and can/should be omitted.  You will need to wind or find three 
suitable ferrite transformers, one of which (T3) must operate as a 
1:1 transmission-line transformer ("current-mode balun") at the 
lowest frequency of interest.  I replace C115 with a 100 uF tantalum, 
and add another 100 uF tantalum in parallel with C113.  Make sure all 
of the multicoupler outputs, including any unused outputs, are 
terminated in 50 ohms to maintain the rated isolation.

Best regards,

Don


Copyright (c) 2013.  Not for redistribution