[K3] Perfect for Field Day S&P ... Limited Production 700 Hz 8-Pole Filters

[Chuck Guenther]

I designed analog and digital filters for processing flight test vibration measurements in the aerospace industry.  I understand how cascaded filters work, and the function of a roofing filter versus the DSP filtering in the K3.  I have owned and operated my K3 since mid 2008, and have turned the bandwidth knob a time or two during the past several years!

I prefer to keep my roofing filter bandwidth close to the DSP bandwidth, for maximum attenuation of close in signals.  My experience running stations during Field Day is that callers are sometimes rusty on their CW skills and unfamiliar with the equipment they are using.  A fairly wide bandwidth (500-600 Hz) is sometimes needed to capture these callers.  Yes, turning the bandwidth knob will accomplish this, and should work in most cases.  But some roofing filter will be in the chain.  My preference is that it be 700 Hz rather than 2.1 or 2.7 KHz for the additional skirt selectivity and protection of the A/D from close in signals. 


I am not affiliated with Gary, other than being a satisfied customer.  (I used his "Rework Eliminators" when I built my K2, and I have used the 700 Hz filter for several years now.).


73,
Chuck  NI0C 
______________________________________________________________
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

[Edward R Cole]

I wonder if any of you have seen this:
Pn = KTB
Pn noise power
K = Botzmann's Contant = 138 E-20
T = temperature in Kelvin (0C = 273K)
B = bandwidth in Hz

convert to dB:
Pn = Log (KTB) = -198.6 + LogT + LogB

If B = 2.7 KHZ, LogB = 44.3 dB
If T = 290K (room temp), LogT = 24.6
then Pn = -198.6+24.6+44.3 = -129.7 dBm

which a pretty good MDS for a HF receiver.
MDS defined as minimum detectable signal or when SNR = 1
Ps = Pn

So what if you use a 500-Hz filter?

that merely becomes the Log (0.5/2.7)*Pn = -7.3 -129.7 = -137 dBm
Does anyone recognize this as close to the published MDS for the K3?

So Pn reduced 7.3 dB when the bandwidth was reduced from 2.7 to 0.5 KHz!

Hello!  The noise reduced.  Why? Because the noise bandwidth was reduced.
It is to be expected!

Now your homework is to calculate Pn for 2.1 KHz to 250 Hz reduction. ;-)

This is accurate if there is no AGC operating which will change
receiver gain, including both Pn and Ps.  For that math I suggest you
contact Elecraft.  No, I am not serious about the last comment,
unless you are a communications engineer and plan to design a receiver.

73, Ed - KL7UW
---------------------
From: Bill Frantz <[hidden email]>
To: [hidden email]
Subject: Re: [Elecraft] [K3] Perfect for Field Day S&P ... Limited
         Production 700 Hz 8-Pole Filters
I am using:

13 KHz: 0 dB
2.7 KHz: 0 dB
2.1 KHz: 0 dB
250 Hz: 6 dB

When the 2.1 and 250 filters switch in, the signal level in the
earphones doesn't seem to change, but the noise level does
indicating an improvement in S/N. Perhaps they are sharpening
the cutoff of the DSP filter removing some of the off-frequency noise?

Cheers - Bill, AE6JV



73, Ed - KL7UW
http://www.kl7uw.com
     "Kits made by KL7UW"
Dubus Mag business:
     [hidden email]

______________________________________________________________
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

[Joe Subich, W4TV-4]

 > If B = 2.7 KHZ, LogB = 44.3 dB
 > If T = 290K (room temp), LogT = 24.6
 > then Pn = -198.6+24.6+44.3 = -129.7 dBm

All of which means absolutely nothing since the proper number to use
for "B" (bandwidth) is the final detector bandwidth which can be as
little as 50 Hz *for a trained human ear*.

A conventional product detector following an IF where the selective
element is a crystal filter after the mixer has a detector bandwidth
of several KHz (the noise bandwidth of the IF) yet the real bandwidth
that determines MDS is the bandwidth of the user's ear/brain filter not
necessarily that of the IF filter or the product detector.  It is quite
easy for the trained ear to detect coherent (e.g. CW) signals that have
a negative SNR of one were simply measuring signal power vs. integrated
noise power in the product detector output.

73,

    ... Joe, W4TV


On 3/3/2014 1:22 PM, Edward R Cole wrote:

> I wonder if any of you have seen this:
> Pn = KTB
> Pn noise power
> K = Botzmann's Contant = 138 E-20
> T = temperature in Kelvin (0C = 273K)
> B = bandwidth in Hz
>
> convert to dB:
> Pn = Log (KTB) = -198.6 + LogT + LogB
>
> If B = 2.7 KHZ, LogB = 44.3 dB
> If T = 290K (room temp), LogT = 24.6
> then Pn = -198.6+24.6+44.3 = -129.7 dBm
>
> which a pretty good MDS for a HF receiver.
> MDS defined as minimum detectable signal or when SNR = 1
> Ps = Pn
>
> So what if you use a 500-Hz filter?
>
> that merely becomes the Log (0.5/2.7)*Pn = -7.3 -129.7 = -137 dBm
> Does anyone recognize this as close to the published MDS for the K3?
>
> So Pn reduced 7.3 dB when the bandwidth was reduced from 2.7 to 0.5 KHz!
>
> Hello!  The noise reduced.  Why? Because the noise bandwidth was reduced.
> It is to be expected!
>
> Now your homework is to calculate Pn for 2.1 KHz to 250 Hz reduction. ;-)
>
> This is accurate if there is no AGC operating which will change receiver
> gain, including both Pn and Ps.  For that math I suggest you contact
> Elecraft.  No, I am not serious about the last comment, unless you are a
> communications engineer and plan to design a receiver.
>
> 73, Ed - KL7UW
> ---------------------
> From: Bill Frantz <[hidden email]>
> To: [hidden email]
> Subject: Re: [Elecraft] [K3] Perfect for Field Day S&P ... Limited
>          Production 700 Hz 8-Pole Filters
> I am using:
>
> 13 KHz: 0 dB
> 2.7 KHz: 0 dB
> 2.1 KHz: 0 dB
> 250 Hz: 6 dB
>
> When the 2.1 and 250 filters switch in, the signal level in the
> earphones doesn't seem to change, but the noise level does
> indicating an improvement in S/N. Perhaps they are sharpening
> the cutoff of the DSP filter removing some of the off-frequency noise?
>
> Cheers - Bill, AE6JV
>
>
>
> 73, Ed - KL7UW
> http://www.kl7uw.com
>      "Kits made by KL7UW"
> Dubus Mag business:
>      [hidden email]
>
> ______________________________________________________________
> 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

[Wes (N7WS)]

Joe is right on.

Back when EME was "real", i.e copied by ear, it wasn't unusual to copy signals
with a negative SNR.

A little story:  I went to work for Hughes Aircraft here in Tucson when I was 25
years old.  At the time I was given a hearing test as part of my physical exam.
(So they told me later).

About 15 years later, at the time I was working two-meter EME,  I had to attend
a meeting in another building.  After the meeting, as I exited that building and
headed to the security gate a Plant Protection Office ran to his truck, jumped
in and proceeded to turn on the siren just as I walked by.  I was nearly
deafened by the sound.

The next day my ears were still ringing (to this day I have tinnitus) so I went
to the Medical Department and reported an on-the-job injury. As part of their
evaluation they gave me another hearing test.  The test was almost like
listening for EME signals; detecting weak tones buried in noise and pressing a
button when I heard them.  Their conclusion was that not only where my ears
fine, my hearing had actually improved since I took the first test!  Clearly
impossible, but training made it seem so.

My receiver (in fact with the exception of the coax, the whole station) was HB
and designed for high linearity and low distortion.  The final hardware BW was
200 Hz, provided by a passive, LC audio BPF.  The final BW was set by the
tracking filter between the ears.

Wes  N7WS


On 3/3/2014 12:59 PM, Joe Subich, W4TV wrote:

>
> > If B = 2.7 KHZ, LogB = 44.3 dB
> > If T = 290K (room temp), LogT = 24.6
> > then Pn = -198.6+24.6+44.3 = -129.7 dBm
>
> All of which means absolutely nothing since the proper number to use
> for "B" (bandwidth) is the final detector bandwidth which can be as
> little as 50 Hz *for a trained human ear*.
>
> A conventional product detector following an IF where the selective
> element is a crystal filter after the mixer has a detector bandwidth
> of several KHz (the noise bandwidth of the IF) yet the real bandwidth
> that determines MDS is the bandwidth of the user's ear/brain filter not
> necessarily that of the IF filter or the product detector.  It is quite
> easy for the trained ear to detect coherent (e.g. CW) signals that have
> a negative SNR of one were simply measuring signal power vs. integrated
> noise power in the product detector output.
>
> 73,
>
>    ... Joe, W4TV
>

______________________________________________________________
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