[time-nuts] GPS from a window seat

[SAIDJACK at aol.com]

That's only a small issue in my opinion.
 
I've designed broadband systems that run at ~50% to 150% of self  resonant 
frequency, and it works quite well if you put a 50 - 220 Ohm resistor  in 
series with the inductor as the other member noted.
 
It works well because it basically is a single-pole system, and above SRF  
you get the parasitic capacitive loading into the series resistor, and  
below you get the inductive loading into the series resistor. At SRF there  
should be very little AC load.

The impedance of the 220Ohm in series with the inductor will be so much  
higher than the 50 Ohm line impedance that the increase in SWR due to  the 
bias-t will be negligible.
 
A problem is the stub generated by the solder connection to the inductor:  
that could work as a load capacitor if not designed properly.
 
If you have a known frequency (such as 1574MHz) then one very good bias-t  
is a quarter-wave transmission line made with coax cable, or a PCB trace.
 
The 1/4 wave transformer will totally keep the AC load to a minimum if  
designed properly, and should be wide-band enough to work well even with  
tolerances in its length.
 
We designed DirectTV front-ends with heavy-duty (1500 Amps!!) lightning  
protection at the end of the 1/4 transformer, and it worked quite well for  
frequency ranges of 900MHz to 2GHz.. The protection diodes were so big that  
their intrinsic capacitance was basically an RF short, but it didn't matter  
because of the 1/4 transformer.
 
bye,
Said
 
 
In a message dated 10/1/2009 21:50:09 Pacific Daylight Time,  
bruce.griffiths at xtra.co.nz writes:


Sure, but what about production tolerances?
In any case the  actual SRF isn't specified for some parts.
Some of the ferrite filled  conical chokes work  well.

Bruce