[Elecraft] Next Question: DATA A with K3s and Fldigi

[Conrad PA5Y]

Don, it depends on the hardware. I would be very surprised if the Audio codec used in the K3S can reach full scale with the windows mixer at max. I expect that it is limited to -1dB FS. Also I believe that there are further limits imposed i WSJT-X. Therefore there should be no reason for clipping to occur in the CODEC.  I expect that there is sufficient range in the mic gain (or whatever it is called) on the K3S. This gives the input signal the maximum SNR which is the socially responsible thing to do.

Until I have checked the behaviour of the windows mixer with the CODEC I will not be using it. As I have said previously I don't trust the windows mixer not to degrade the audio due  This may well have been resolved but historically was of concern.

I think that I will decide what to do when I have made some proper measurements. It's not that I don't believe you, I just like to check things for myself.

73

Conrad PA5Y


________________________________
From: Don Wilhelm <donwilh at embarqmail.com>
Sent: 17 June 2019 17:31
To: Conrad PA5Y; elecraft at mailman.qth.net
Subject: Re: [Elecraft] Next Question: DATA A with K3s and Fldigi

Conrad,

Do check and adjust the Windows soundcard controls to mid-range. When
set to extremes, it can cause clipping and distortion.

With the K3S, the internal soundcard will identify as USB AUDIO CODEC.

73,
Don W3FPR

On 6/17/2019 11:21 AM, Conrad PA5Y wrote:
> Don of course it is not a power control, that is obvious to me.
>
> I am very fortunate to have a lot of RF and Audio test equipment
> available both at home and at work so I will measure what works best
> and try to understand why. I will be happy to share my findings. I
> will of course read your article. I am loathe to use the Windows audio
> mixer, I would rather use an external passive attenuator. Maybe the
> Windows mixer has improved, I need to check that with an audio
> analyser. It used to be awful so I built an external passive
> attenuator some years ago.
>
> I will be interested to see what happens to the K3S TX composite noise
> when the power control is used. There is a certain very popular radio
> where the AM noise comes up by 20 dB when the power is reduced from
> 100W to 30W. I would be really pleased if the K3S power control works
> properly!
>
> I bought the K3S purely for its TX PN performance.
>
> 73
>
> Conrad PA5Y
> ------------------------------------------------------------------------
> *From:* Don Wilhelm <donwilh at embarqmail.com>
> *Sent:* 17 June 2019 16:03:02
> *To:* Conrad PA5Y; elecraft at mailman.qth.net
> *Subject:* Re: [Elecraft] Next Question: DATA A with K3s and Fldigi
> Conrad and all,
>
> The POWER control in WSJT-X is a misnomer.  It is actually an audio
> level control.  Do NOT use it to control the power with an Elecraft
> transceiver - set the audio correctly and use the power knob to control
> the power.
>
> While other amateur transceivers can be made to control the power by
> varying the audio level, that can be disastrous for Elecraft transceiver
> Power Control.  Read the relevant article on my website www.w3fpr.com<http://www.w3fpr.com>
> <http://www.w3fpr.com>
> for a full explanation.
>
> Set the WSJT-X POWER control for 30% to 50% and do the same for the
> soundcard control in Windows Mixer, then set the LINE IN gain in the
> K3/S to somewhere near mid-range.  By manipulation of those 3 audio
> level controls, you should be able to achieve the 'magic' indication of
> 4 bars solid on the ALC meter with the 5th bar flashing - that is the
> "NO ALC" point for the K3/S, KX3, and KX2 transeivers. The lower 4 bars
> of the "ALC" meter are there as an aid for setting the audio levels.
> Set the desired power output with the POWER knob - do NOT vary the audio
> level.
>
> 73,
> Don W3FPR
>
> On 6/17/2019 9:38 AM, Conrad PA5Y wrote:
> >
> > Historically there were all kinds of low level artefacts produced by
> Windows audio algorithms, including the volume control. This was
> caused by insufficient precision leading to truncation artefacts. I
> have not tested it for a while and so it may no longer be the case.
> The result when I did test it was spectral smearing and some AM on
> what should be a constant envelope signal.  I trust the power control
> in WSJT-X far more than that from the Windows mixer. These effects
> only really matter during weak signal conditions when your TX signal
> is on or near the noise floor. at the receiving station However they
> do make your signal wider.
> >
> > Also be careful how you vary power, reducing audio carelessly will
> invariably decrease the signal to noise on your TX signal which will
> also affect a stations ability to receive you.
> >
> > My K3S arrives this week and I will make some tests to see exactly
> what happens with TX composite noise when power is reduced with the
> power control and by reducing audio drive from the PC. Unless someone
> else has measured this already?
> >