[TenTec] Roofing Filters and DSP BW

Bill Tippett btippett at alum.mit.edu
Sat May 24 10:27:55 EDT 2003

	This is a related to my previous question about the curves,
shape factor, ripple, bandwidths, etc. of Ten Tec's Orion filters.  In a
private E-mail, Earl K6SE said:

"In other words, if the BW is set to 6000Hz, the 20kHz is the active
roofing filter.  For a BW from 2400 to 5990Hz, the 6.0k roofing filter is
active.  From 1000 to 2390Hz, the 2.4k roofing filter is active.  From
100 to 990Hz, the 1.0k roofing filter is active (with the 500Hz filter
disabled).  The actual BWs where the next filter switched in or out might
actually be 10 or 20Hz less, i.e., 5980, 2380, 980Hz, etc.

My 500Hz filter switches in when I switch the BW from 380 to 370Hz (I can
hear a slight change), so it appears that Ten-Tec realized that some
users might have the INRAD 400Hz filter installed

My 250Hz filter switches in when I switch the BW from 200 to 190Hz."

	I also saw a note from AB7R which confirmed that his Orion also
switched the 500 Hz roofing filter in at a DSP BW of  370 Hz.  From this,
it appears that the DSP algorithms should operate in the flat areas of the
roofing filters, and that perhaps <? dB ripple within the DSP BW is the 
for automatically switching to the next lower roofing filter (i.e. for a 
500 Hz
6 dB BW filter, you get <? dB of ripple only in the middle 370 Hz area [or
less] of the roofing filter response).  This indicates to me that passband
ripple may be an important part of any roofing filter characteristic in
addition to shape factor.  The curious thing to me is that the automatic
mode seems to switch right at stated (6 dB ?) bandwidth points of the
6000, 2400 and 1000 Hz filters, but at lower points for the 500 and 250
Hz filters.  Perhaps, as Earl said, they did this knowing that narrower
filters would be used (Inrad's "400 Hz" is actually 455 Hz @ 6 dB and
their "250" is actually 315 Hz @ 6 dB).

	It's very easy to see Inrad's filter characteristics on their website,
but I wish Ten Tec or someone would publish the curves for all Orion roofing
filters (not just the optional ones), along with an explanation of Orion's DSP
assumptions for any roofing filter that precedes it.  Since they have given us
the flexibility of overriding automatic roofing filter selection, I believe 
would be important information for those of us who may choose to do so.

					73,  Bill  W4ZV

More information about the TenTec mailing list