[hpsdr] ALEX - Input protection

[Graham Haddock]

Hi Henry:
Comments inline.
--- Graham / KE9H


Henry Vredegoor wrote:
> I was somewhat aware of the energy distribution in frequency of static noise
> caused by lightning etc., but in my previous post I was referring to the
> direct DC path to ground, the same idea of the solution as pointed to by
> Ray, as posted by Achim DH2VA, not the transformer effect of filtering
> itself.
> But interesting to read and learn anyway!!
>
> If I understand you correctly, you are using 2 X 10 Kohm resistors for this
> (I am surprised that these can be so high in value!), but wouldn't such a
> direct DC path to ground work better? (the lower the impedance to ground for
> DC / "LF" static, the better?)
>   
Any static you can hear in the radio is going to be at some RF 
frequency, and should not
be affected by the DC path to ground.  The DC path to ground is to bleed 
off any constant
static charge that accumulates on an antenna. Any value of resistance 
from 1K to 100K will work.
100K is commonly used on transmitter connectors, so it doesn't absorb 
too much of your
transmitter power.  A time constant in milliseconds is fine.

All it has to do is bleed off the charge faster than the voltage can 
reach the TVS avalanche
voltage and trigger any spurious discharge events.  This DC path is NOT 
the one for lightning to
seek ground.  That must be closer to the antenna, and before the 
lightning ever gets inside
the building, much less your equipment.  Good reading on the Polyphaser 
web site.  See my
previous email to Chris.
> It was a (too-) quickly typed posting I guess, because I meant of course the
> "ground" or "return" OF THE ANTENNA SYSTEM being connected to the "ground"
> connection of the input winding of the broadband transformer.
> In an ideal situation this antenna system "ground" or "return" could be
> "isolated" from the shack/receiver/safety ground; the output of the
> broadband transformer could then be connected to the shack/receiver/safety
> ground on one side so that unbalanced filters could still be used?
>
> Also would a galvanic isolation be useful wrt. avoiding ground loops in a
> HPSDR setup or is this a non-issue in Mercury/Penelope setups and only of
> importance if ALEX would be used as a BPF in a Softrock/Janus/Ozy like
> setup?
>   
Mercury should not have the RF ground loop problems that SoftRock does.  
If you put ALEX
on a SoftRock, I would use the floating input on the SoftRock as you 
describe above, where you hook
it to the ALEX output.
> Another thing I am wondering about is if it would be beneficial if a step-up
> of impedance on the output of such a transformer (say 600 Ohms or so) would
> give better results for the passive filters?
> I played around somewhat with filter simulation software and if I remember
> (and simulated!) correctly, it seemed to be of influence. Or am I wrong?
>   
You are correct, that you can change the effective Q of a filter by 
changing the impedance. You might
want to go up or go down in impedance, depending on the kind of filter 
design. 

The filter program I am using recognizes the Q of the inductors and includes
this in the results.  If the shapes are good enough, then my plan is to 
go with the simpler circuit.

--- Graham / KE9H

==


 1186538858.0