MCU 4.42 Beta Firmware

[vk4tux]

Just a comment regarding the latest k3 beta firmware;

I have had intermittent problems with ssb TX power levels in the past with
some previous firmware versions.
I did all the usual advised fixes of tx calibration etc, however the tx
power level was hit and miss, and
sometimes I would need to re-ptt or hit the cw key to get full dialled power
level. Usually this happened
 most just after the rig was turned on and would sometimes settle down after
a while of use.

The power would sometimes slowly increase after starting off low level.
However in the latest firmware
I no longer have this issue. I am a happy K3 owner.

My only hope remains for some future firmware improvement to make the Noise
Reduction system work
reasonably well on ssb phone on the low bands, as at the moment it is
ineffective with audio loss and distortion.
I have experimented extensively with 6-1 being probably the best setting
from those available on a noisy 40m band.

 At the moment on low bands, I find K3 NR a waste of time. Use of
combination ATT and RF reduction is my best result here.

Adrian ... vk4tux

______________________________________________________________
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

[David Windisch]

CONTENTS DELETED

The author has deleted this message.

[Brian Hunt]

Here's another discussion by Lyle, KK7P, of Noise Reduction and what you
are getting with various settings.  Link to the thread here:
http://elecraft.365791.n2.nabble.com/Noise-Reduction-parameters-td457358.html#a457359

Brian
K0DTJ

On 9/21/2011 2:38 AM, David Windisch wrote:

> Hi, Adrian:
>
> I have found:   http://www.elecraft.com/K3/K3FAQ.htm  to be helpful.
>
> Here is a quote regarding NR:
>
> QUOTE
>
> How does the Noise Reduction work?
>
> NR depends on correlation of the present input with previous input. The
> system does not actually provide Noise Reduction; it provides Signal
> Selection. In other words, its default is to pass nothing at all. It has to
> believe there is a signal present, and then it builds a filter, or set of
> filters, around the spectral components of the signal it thinks is there.
>   Random noise has no correlation, voice has moderate but not perfect (unless
> you whistle a pure tone) and CW has excellent correlation.  As a result,
> noise is heavily suppressed (no filter is built to pass it), voice is
> partially suppressed (hence it needs some additional gain to compensate for
> this effect so the same AF level will produce about the same audio level
> with a moderate S/N speech signal) and CW is hardly suppressed at all (hence
> it does not need any gain boost).
>   NR is not recommended in Data Mode.  Data is already getting a matched
> filter in the demodulator. You might lose a few symbols as the NR settles
> around the signal, and it might suppress a very weak signal that you could
> otherwise copy.
>   NR in the end is intended for modes you listen to.
>
> UNQUOTE
>
> Hth.
>
> Brgds,
> Dave, N3HE
>
>
> --
> View this message in context: http://elecraft.365791.n2.nabble.com/MCU-4-42-Beta-Firmware-tp6813673p6815465.html
> Sent from the Elecraft mailing list archive at Nabble.com.
> ______________________________________________________________
> 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
>
>

______________________________________________________________
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

[vk4tux]

Thankyou for the links regarding NR. I hope possibly in future firmware
updates that a signal selection
 scale will provide more bleed-through at the -1 setting, allowing for a
less distorted starting point with
 better/more audio level.

 In addition hopefully the comparison signal selection process algorithms
can be improved to better discern noise
 against audio.

After a year with the K3  and experimenting with settings I am happy with
the K3 as I have learnt to deal with
 various signals on various bands. I quite often have local op's telling me
they could not hear the station I was talking too.

For me having both the 2.1 and 1.8 k 8 pole filters is a plus and I usually
listen between 1.8 to 2.1 k width. I use AGC threshold
at 8 and slope 0. RX Eq set to N1EU pink noise roll-off settings (red).

Some time ago I listened to a USA 40m net on a noisy band. I went through
every NR combination with 6-1 giving the relatively
 best result for NR settings.

Then I turned off the NR and turned on the ATT (RF gain already back to 2
o'clock) and what a difference. The noise had
 become suppressed and transparent, and every weak signal was copied with
good clarity compared with the best NR
 setting I could find, which still sounded like mush in comparison. So I
learned using ATT was the way to go.

The K3 obviously has plenty of sensitivity to spare. It just makes me wonder
how Yaesu's DNR works etc, as it seems very
 effective without the same issues. Maybe they use a different method?

Adrian ... vk4tux

______________________________________________________________
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