[Elecraft] Re: KX1 AGC

wayne burdick n6kr at elecraft.com
Sat May 29 00:14:24 EDT 2004


On May 28, 2004, at 6:31 PM, Ed Parish K1EP wrote:

> I decided to see how the KX1 would do with a lot of strong signals in 
> the WPX CW contest.

Ed,

The KX1 uses audio-derived AGC. A certain number of cycles of audio is 
required to charge the AGC detector, amounting to some 10 ms at a 
typical CW pitch. Because of this, all audio-derived AGC systems have 
some lag in response to very large transient signals. The K2's AGC is 
derived at a much higher frequency (about 150 kHz), resulting a charge 
time in the tens or hundreds of microseconds, thus eliminating the 
leading-edge effect. But this takes many more components than the 
simple AGC system used in the KX1.

To migitate the effect somewhat, the KX1's audio amplifier is designed 
to limit the peak-to-peak voltage of the audio signal during the 
detector charge time. The limit is around 4 Vpp and is due to the low 
supply voltage used for the LM386 (6 V). If we had run the LM386 from 
the 9 to 12 VDC or higher supply voltage, the signal could have been as 
much as 8 Vpp on such transitions, resulting in about 4 times the power 
at the headphone jack. So the limiter saves your ears, although it 
still allows a brief period of distortion.

In summary, hearing a burst of nearly square-wave (clipped) audio at 
the leading edge of a very strong CW signal is normal, and can't be 
easily avoided. I'd recommend using a narrow bandwidth setting for the 
crystal filter during contests, which can also lessen the leading-edge 
distortion effect somewhat because it limits the slew rate of 
transients.

73,
Wayne
N6KR



More information about the Elecraft mailing list