[Elecraft] K3 AGC THR doesn't go low enough - SOLVED

[John Chappell G3XRJ]

Problem solved.

When I built my K3 and came to the Owner's Manual setup procedures on 
Page 51 - 'S-Meter and RF Gain control'. Not being in possession of an 
XG2 I took the statement 'calibration is normally adequate using the 
factory settings' at face value - and didn't calibrate the rf gain.

Anyway, some of the responses to my original post pointed to some more 
in depth reading from W3FPR, K6NA and others.
I realised that my K3 simply wouldn't respond as described and had too 
much gain in the RF/DSP somewhere.

So I re-visited the RF Gain Calibration procedure - still without XG2. 
Brought up K3 Utility
Selected Calibrate RF Gain which brings up screen 'Welcome to RF Gain 
Calibration' which refers to using a calibrated signal source -
Nevertheless, click Next
Choose 'Use factory default'
Click  Next

PROBLEM SOLVED

The K3 now sounds and behaves as I would expect.
What a revelation.


It may be just my K3 slipped through some factory stage and needed the 
factory defaults loading.
However, my two nearest K3 owners - also self builds - suffered with the 
same problem I described, as I had discussed it with them before posting 
my original.

One has had similar success with the Utility Calibration the other less so.

However, if you suffer the same issue it would certainly appear to be 
worth running through the procedure outlined above.

It is a night and day transformation for my K3.

Now to order an XG2 for a final tweak.

73 John G3XRJ


> I like to listen to a receiver with the rf gain set to just below
> antenna atmospheric noise and AGC on - and not raising the rf gain when
> it is switched on.
>
> My K3 is srl 46xx and the only way I get the quiet background I desire
> is to turn the AGC off. This necessitates having the AF limiter turned
> on to save my ears and hopefully the speaker output chip.
>
> I've played around with various AGC settings and in particular the
> AGC THR - which if it would go 2 or 3 points lower would appear to
> address the problem.
>