[TenTec] INRAD Filter Question

[Dr. Gerald N. Johnson, electrical engineer]

Ringing may depend on the CW frequency as it passes through the filter.
In the center it likely will ring the least, near the passband corners
where the phase delay changes most rapidly, it will probable ring more.

I have an Inrad 400 Hz filter in a Kenwood. I'm glad that Kenwood has
filter selection independent of mode because that 400 Hz filter (should
not be closely related to an Inrad filter for a Tentec) begins to ring
on CW at 25 or 30 wpm.

Its hard to make sharp enough corners in the ladder filters that TT uses
to get them to ring.

Listen to lightning static. If its a crash the filter is ringing and
frequency didn't make any difference. Crystal and mechanical filters
made for radios other than TT fold up under lightning static and convert
what should be clicks into long crashes. They often convert line noise
that should be a buzz into continuos noise. TT stock filters (like in my
Corsair II) click on static crashes, but do elongate the click some and
buzz on line noise as they should. The only selectivity I've found to
handle lightning static better is the multiply tuned 85 KHz IF's in my
BC-453 (plus converter) with the rods pulled for maximum selectivity.
The passband is nearly gaussian which is nearly optimum for passing
static as short clicks.

On the air, in the summertime (northern hemisphere), notice how the
Yacomwood users fold their tents because the lightning crashes are
killing their reception and notice you can still copy with the stock TT
filters. That's also a strong sign of ringing. As far as I can tell
(from long use) Collins mechanical filters are the worst of the lot and
most crystal lattice filters (typical of Inrad) are nearly as bad.

A filter response that comes back up after a notch at band edge is
characteristic of a crystal lattice filter Tchebyshev frequency response
whose ringing profusely is only surpassed by Cauer frequency responses
for strong ringing. Both sacrifice time response for steep skirts.

Once upon a time a QST author noticing the arrival of crystal filters,
opined that steep skirts and square shoulders were the optimum for
filter responses. Filters have rung badly ever since. He didn't consider
the effects of pulses and ringing. I may go to extreme, I figure time
response is everything under adverse conditions and I can sacrifice
those square shoulders and steep skirts for filters that don't hide the
signal in ringing hash from static.

In other brands of radios it is essential to have the optional CW filter
in crowded CW bands (like during CW contests). The bandpass tuning of
the TT radios makes that far less an essential. Its easy to use one
slope of the 9 MHz filter and the other slope of the 6.3 MHz filter to
make a pass band narrower than the CW filter, such as 100 Hz and it
responds well. I do have a CW filter but passband tuning with the stock
SSB filters does quite well and I had that CW filter sitting on the
shelf for several months wondering why I'd bought it.

And the stock SSB filters easily narrow down to 1.8 KHz for SSB, though
cascaded 1.8 KHz filters in both IF's will give more rapid adjacent
channel rejection. Still the lack of front end crunch makes that far
less of a problem.

73, Jerry, K0CQ

-- 
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.