Steve Ellington, N4LQ said in part:
>
>1. For reception of cw in heavy QRN, would one want a fast or slow
attack
>time? Newer rigs seem to have faster attack but are hurt more by
qrn.
Hi Steve,
I understand your theory about audio AGC with relatively slow attack
times being better in high static conditions and I agree with you.
I own an old Argonaut 505, an Omni-D Series B and a Corsair II. The
Argonaut & Omni hear better in noisy QRN conditions than the Corsair
does. As a test, you can turn a light switch on and off and listen
to the audio on a dead band like 10 meters with each receiver. The
pop is louder on the older rigs but the S meter hardly moves and
there is little or no noticeable gap in the steady hiss audio out
from the receiver. With the Corsair, the pop isn't as loud, but
particularly with a slow AGC decay time constant there is a
noticeable pause/gap before the hiss is again audible from the
receiver.
I've found that fast AGC time constants help when the noise bursts
are high because each time there is a strong crash, it takes less
time for the AGC to decay and allow the receiver audio output to
recover to somewhat normal volume. With a slow AGC decay time
constant, a loud static crash will 'blank' the receiver audio output
for a second or more which causes the operator to lose more
information during that gap.
I've modified the Omni to use the FAST/SLOW QSK switch as an AGC time
constant selector. I forget the actual values but I think I changed
the capacitor to about half of the stock value for fast AGC and about
double the original value when slow AGC is selected. The Argonaut is
stock and the Corsair is designed with selectable FAST/SLOW AGC on
the front panel.
The basic design of the two older AGC circuits are similar while the
Corsair has the new 'hang' AGC design. With the Corsair, the AGC
attack is quite fast and the time constant VERY long. A second
circuit detects the audio output and quickly discharges the time
constant capacitor if there is no or little audio output for about
two seconds or more (about a half-second in FAST mode).
Collins had an excellent AGC circuit in their KWM-2 and S-Line
receivers which used dual time constants. The R-C circuit for time
constant had TWO pairs of parallel resistors and capacitors hooked up
in series. The component values of these were selected to produce an
unusual effect. While listening to a SSB signal that peaks S-9 on
the meter, the voice peaks cause the meter to flicker noticeably at a
syllabic rate between S-6 and S-9. When the operator pauses or stops
speaking, the meter would then slowly drop from S-6 down to the
ambient noise level. The receiver audio sounded almost like a
transmitter with 6 to 12 dB of audio speech compression. Strong
static crashes didn't 'mute' the receiver so badly because a
percentage of the AGC voltage had a fast decay time constant (the
S-meter swing between S-6 and S-9 in the above example). The
remaining percentage of the AGC voltage had a slow decay time
constant which reduced the 'pumping' sound in the audio output. The
ratio of the values of the two discharge resistors sets the amount of
'fast AGC' or varies the width of the S-6 to S-9 response in the
above example while the capacitor values in the circuit control the
decay time of the voltage across each resistor.
Another technique I've seen Steve is to use a very carefully chosen
level of clipping or limiting in the audio going to the AGC detector
or receiver audio output or both. The clipping level should be set
so that there is NEVER ANY clipping on steady level signals of any
strength. The clipping occurs during the time period when a fast
rise time strong signal first hits the receiver up until the AGC
voltage is developed and the receiver gain is reduced to a steady
level for that strong signal. The improvement is two-fold... First
a single narrow pulse of strong signal is clipped and doesn't last
long enough for the AGC detector to develop a high AGC voltage so
there is much less of a 'gap' in output receiver audio or muting
effect after a strong signal. Second, if clipping is also applied to
the signal going to the receiver output amplifier and
speaker/headset, the operator's ears are somewhat protected since the
peak audio output is limited while the receiver is 'overloaded' due
to slow AGC attack time on strong signals. Obviously there is some
intentional distortion introduced here but the effect is pleasant if
the proper clipping level(s) are correctly set.
Sorry to ramble so much but perhaps these ideas will help you better
understand how your equipment is working or how to modify it for your
intended purposes.
73,
Bob Helms, AF5Z
af5z@inetport.com
--
FAQ on WWW: http://www.contesting.com/tentecfaq.htm
Submissions: tentec@contesting.com
Administrative requests: tentec-REQUEST@contesting.com
Problems: owner-tentec@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|