K3 fw 3.25 NR

[Richard Van Tieghem]

Several posts to the list have indicated that the NR function in K3 firmware 3.25 sounds the same, regardless of which NR setting has been selected.  
As mentioned by another list member, firmware 3.25 has an anomaly that prevents a new NR setting from immediately taking effect.
The work-around is to change the NR setting (such as from F4-1 to F4-4), then disable NR, and then re-enable NR (the new NR setting becomes active upon NR re-enable).  I've found that NR is much improved in fw 3.25, and is now usable for me with relatively weak CW signals.
Rich VanTieghem W2VT


     
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[KK7P]

> Several posts to the list have indicated that the NR function in K3 firmware 3.25 sounds the same, regardless of which NR setting has been selected.  
> As mentioned by another list member, firmware 3.25 has an anomaly that prevents a new NR setting from immediately taking effect.
> The work-around is to change the NR setting (such as from F4-1 to F4-4), then disable NR, and then re-enable NR (the new NR setting becomes active upon NR re-enable).  I've found that NR is much improved in fw 3.25, and is now usable for me with relatively weak CW signals.

The reason for this is that the LMS algorithm as implemented has a slow
"decay" of computed NR filter values.  Decay is a trade off between
response of the algorithm to signal that disappears.  A side effect is
that if you go from less aggressive F1-x to more aggressive (F4-x)
settings, the filter built around the signal remain in place and slowly
decays to the more aggressive setting.

If you start with F4-x and go towards F1-x, you'll find the algorithm
response to your changes is faster, due to the less aggressive nature of
the algorithm at those settings.

I'm looking into ways to get it to be more responsive as you increase
aggressiveness without compromising other parameters that must be kept
in balance.

Thank you for your feedback on this Beta code!

73,

Lyle KK7P

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[drewko]

A lot ot times I would like to compare, for example, filter setting
1-3 with filter setting 2-3. There is no way to do this currently
without stepping through four intermediate settings. It would be nice
if in addition to the current filter sequence (1-1, 1-2, 1-3, 1-4,
2-1, etc...), we could also step through the settings in this
sequence:

1-1, 2-1, 3-1, 4-1, 1-2, 2-2, 3-2, etc.

In other words, in the 4x4 matrix of filter settings we should be able
to move either row-by-row or column-by-column. This would allow much
more freedom to compare the different settings.

I suggest that when "NR F x-y" is visible on the display that the
VFO-A knob be used to move through the settings in this vertical
order. Then we can go through them in any order using the two VFO
knobs.

73,
Drew
AF2Z


On Thu, 20 Aug 2009 19:17:57 -0700, Lyle KK7P wrote:

>> Several posts to the list have indicated that the NR function in K3 firmware 3.25 sounds the same, regardless of which NR setting has been selected.  
>> As mentioned by another list member, firmware 3.25 has an anomaly that prevents a new NR setting from immediately taking effect.
>> The work-around is to change the NR setting (such as from F4-1 to F4-4), then disable NR, and then re-enable NR (the new NR setting becomes active upon NR re-enable).  I've found that NR is much improved in fw 3.25, and is now usable for me with relatively weak CW signals.
>
>The reason for this is that the LMS algorithm as implemented has a slow
>"decay" of computed NR filter values.  Decay is a trade off between
>response of the algorithm to signal that disappears.  A side effect is
>that if you go from less aggressive F1-x to more aggressive (F4-x)
>settings, the filter built around the signal remain in place and slowly
>decays to the more aggressive setting.
>
>If you start with F4-x and go towards F1-x, you'll find the algorithm
>response to your changes is faster, due to the less aggressive nature of
>the algorithm at those settings.
>
>I'm looking into ways to get it to be more responsive as you increase
>aggressiveness without compromising other parameters that must be kept
>in balance.
>
>Thank you for your feedback on this Beta code!
>
>73,
>
>Lyle KK7P
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html