On Tue,5/5/2015 4:11 PM, Paul O'Kane wrote:
Therefore, to get the ball rolling, I suggest that
excessive bandwidth be defined, in part, as follows -
CW - width greater than 500Hz at 40db down
SSB - width greater than 6kHz at 40db down
Having studied this matter for a while, and having done some
measurements of my own, I'll throw these thoughts onto the pile. First,
some numbers.
My 2008 vintage K3 driving a 1980 vintage Ten Tec Titan to 1500W is 360
Hz wide at -40, 470 Hz wide at -50.
Before the September 2014 firmware update and set for 6 msec rise time
(it's slowest), a neighbor's FTDX5000 was 480 Hz wide at -40, 800 Hz
wide at -50. After the update, the numbers are 410 Hz wide at -40, 600
Hz wide at -50. The greatest improvement with the new firmware is below
-50 -- it's 1.7 kHz wide at -60 before the update,
I'm measuring with a P3/SVGA hooked up to a K3, and the rig under test
is driving a 500W dummy load. I'm using dits in the range of 30-35 WPM.
I'm measuring in peak mode, and accumulating peaks for a while, so it's
a worst case measurement.
SO -- since the FTDX5000 was the widest of the modern rigs ARRL tested,
I'd say your 500 Hz at -40 is a pretty good limit. Setting the rise time
faster will cause it to be wider. Being careless with amplifier tuning,
or using AGC between amp and rig, or using a dirty amp, will all make
the signal wider.
On SSB, I'd call 6 kHz wide at -40 pretty tight. That's what a K3 does
with strong LF rolloff driving a legal limit power amp; an FTDX5000 is
slightly better running 50W in Class A, slightly worse running Class AB.
I'd call 7 kHz at -40 more realistic. SSB gets real wide real fast with
distorted audio and those same problems with power amp setup. Again,
these numbers are for accumulated peaks. One way to tell that it's an RF
problem, NOT an audio problem, is when the waterfall shows horizontal
lines going both sides of the carrier on voice peaks.
The really dumb part of badly distorted, overdriven audio is that it
makes the guys that do it really hard to copy.
73, Jim K9YC
_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest
|