Elecraft

K4 Question

Classic

List

Threaded

18 messages Options

Eric Norris-2

K4 Question

My foaming at the mouth over the K4 has been tempered by it having less
adjacent channel rejection than the K3 due to the different
architecture--at least until the K4HD comes out. I understand this, and
the reasons why. Thanks for your answers

I know I have asked this question before, but I want to be more specific.
My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away on
ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
other degradation, or would I be able to carry on a CW qso unmolested like
I can with my K3? What about an adjacent band like 3534 kHz or 10114 kHz?
Or is the answer I have to wait for the K4HD?

No speculation, please, I'm looking for a real-world or lab-world answer.

Thanks and,

73 Eric WD6DBM
______________________________________________________________
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
Message delivered to [hidden email]

donovanf

Re: K4 Question

Hi Eric,

We use more than a dozen K3 transceivers at the W3AO Field Day with
as many as four operating in the same band (CW, SSB, FT8 and GOTA SSB),
but not this year... (:

There's absolutely no trace of interference between four K3 transceivers
operating in the same band, despite the antennas being less than 250 feet
apart and no external filtering to minimize in-band interference.

73
Frank
W3LPL

----- Original Message -----

From: "Eric Norris" <[hidden email]>
To: "elecraft@mailman qth. net" <[hidden email]>
Sent: Tuesday, June 23, 2020 9:32:04 PM
Subject: [Elecraft] K4 Question

My foaming at the mouth over the K4 has been tempered by it having less
adjacent channel rejection than the K3 due to the different
architecture--at least until the K4HD comes out. I understand this, and
the reasons why. Thanks for your answers

I know I have asked this question before, but I want to be more specific.
My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away on
ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
other degradation, or would I be able to carry on a CW qso unmolested like
I can with my K3? What about an adjacent band like 3534 kHz or 10114 kHz?
Or is the answer I have to wait for the K4HD?

No speculation, please, I'm looking for a real-world or lab-world answer.

Thanks and,

73 Eric WD6DBM
______________________________________________________________
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
Message delivered to [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
Message delivered to [hidden email]

wayne burdick

Re: K4 Question

Administrator

Hi Eric,

S9+65 dB is about -8 dBm. Off the top of my head, this is far, far below what a basic K4 or K4D can handle, artifact-free, in-band, without the need for attenuation or additional filtering. When I get back to the lab I'm going to set up exactly this condition and get back to you.

Of course the out-of-band rejection is even higher.

A number of K4s will be used extensively during FD this year, including mine. I'll be taking advantage of the K4's low current drain (for its class) by running mine from a KX2 11 volt battery pack (3x 18650 cells). For at least an hour or so :)

73,
Wayne
N6KR

> From: "Eric Norris" <[hidden email]>
> To: "elecraft@mailman qth. net" <[hidden email]>
> Sent: Tuesday, June 23, 2020 9:32:04 PM
> Subject: [Elecraft] K4 Question
>
> My foaming at the mouth over the K4 has been tempered by it having less
> adjacent channel rejection than the K3 due to the different
> architecture--at least until the K4HD comes out. I understand this, and
> the reasons why. Thanks for your answers
>
> I know I have asked this question before, but I want to be more specific.
> My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away on
> ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
> perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
> other degradation, or would I be able to carry on a CW qso unmolested like
> I can with my K3? What about an adjacent band like 3534 kHz or 10114 kHz?
> Or is the answer I have to wait for the K4HD?
>
> No speculation, please, I'm looking for a real-world or lab-world answer.
>
> Thanks and,
>
> 73 Eric WD6DBM

______________________________________________________________
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
Message delivered to [hidden email]

Adrian-3

Re: K4 Question

In reply to this post by Eric Norris-2

An excessive strength FT8 signal is easy to remove with notch.

On 24/6/20 7:32 am, Eric Norris wrote:

> My foaming at the mouth over the K4 has been tempered by it having less
> adjacent channel rejection than the K3 due to the different
> architecture--at least until the K4HD comes out. I understand this, and
> the reasons why. Thanks for your answers
>
> I know I have asked this question before, but I want to be more specific.
> My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away on
> ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
> perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
> other degradation, or would I be able to carry on a CW qso unmolested like
> I can with my K3? What about an adjacent band like 3534 kHz or 10114 kHz?
> Or is the answer I have to wait for the K4HD?
>
> No speculation, please, I'm looking for a real-world or lab-world answer.
>
> Thanks and,
>
> 73 Eric WD6DBM
> ______________________________________________________________

Jim Brown-10

Re: K4 Question

In reply to this post by donovanf

On 6/23/2020 3:27 PM, [hidden email] wrote:
> We use more than a dozen K3 transceivers at the W3AO Field Day with
> as many as four operating in the same band (CW, SSB, FT8 and GOTA SSB),
> but not this year...

What power level, Frank?

73, Jim K9YC
______________________________________________________________
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
Message delivered to [hidden email]

Jim Brown-10

Re: K4 Question

In reply to this post by Adrian-3

On 6/23/2020 4:53 PM, Adrian wrote:
> An excessive strength FT8 signal is easy to remove with notch.

Yes, BUT -- the notch introduces phase shift, which can degrade
decoding. K1JT recommends running the IF wide and letting WSJT-X
separate signals.

I DO strongly differ with the advice in the WSJT-X manual for setting
levels. Instead, I recommend setting audio gains so that the green bar
indicating audio level as close as possible to the top without it
turning red on when strong signals are present. I find Slow AGC works
well for me, because both the strongest signals AND my local noise vary
widely.

Why do I recommend this? Simple. The practical dynamic range of a 16-bit
A/D is about 90 dB (96 dB theoretical, but linearity tends to degrade at
the low end of the range); if there's a 50 dB over S9 signal in the
passband and it's set to 78 dB on the green bar, we can decode signals
28 dB below S9. At 5 dB/S-unit, that's between S3 and S4. If the
strongest signal is 40 dB over S9, we can almost get down to S1.

This doesn't matter, of course, if your noise level is S8, but it
matters a LOT if it's a lot lower.

73, Jim K9YC

______________________________________________________________
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
Message delivered to [hidden email]

Adrian-3

Re: K4 Question

On 24/6/20 11:02 am, Jim Brown wrote:
> On 6/23/2020 4:53 PM, Adrian wrote:
>> An excessive strength FT8 signal is easy to remove with notch.
>
> Yes, BUT -- the notch introduces phase shift, which can degrade
> decoding. K1JT recommends running the IF wide and letting WSJT-X
> separate signals.
Ok on that, My experience is that it allowed decode of my target RX., I
did use narrow notch.
>
> I DO strongly differ with the advice in the WSJT-X manual for setting
> levels. Instead, I recommend setting audio gains so that the green bar
> indicating audio level as close as possible to the top without it
> turning red on when strong signals are present. I find Slow AGC works
> well for me, because both the strongest signals AND my local noise
> vary widely.

Yes I do that all possible gain and AGC turned off on low band use, with
green as high as 80dB with no red seen , for good weak signal decode.

Using low width roofing/dsp filtering around the signal boosts it to
great extent, using width/shift to focus.

______________________________________________________________
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
Message delivered to [hidden email]

Elecraft mailing list

Re: K4 Question

Once any FT8 passband is digitized, filtered, and decimated in a K4, there should be no need to ever apply AGC - just pass the bits over to WSJT. The decimated bit stream may exceed 16 bits, so the WSJT input processor would have to be modified to accept the additional dynamic range.

On Tuesday, June 23, 2020, 9:12:47 PM EDT, Adrian <[hidden email]> wrote:

On 24/6/20 11:02 am, Jim Brown wrote:
> On 6/23/2020 4:53 PM, Adrian wrote:
>> An excessive strength FT8 signal is easy to remove with notch.
>
> Yes, BUT -- the notch introduces phase shift, which can degrade
> decoding. K1JT recommends running the IF wide and letting WSJT-X
> separate signals.
Ok on that, My experience is that it allowed decode of my target RX., I
did use narrow notch.
>
> I DO strongly differ with the advice in the WSJT-X manual for setting
> levels. Instead, I recommend setting audio gains so that the green bar
> indicating audio level as close as possible to the top without it
> turning red on when strong signals are present. I find Slow AGC works
> well for me, because both the strongest signals AND my local noise
> vary widely.

Yes I do that all possible gain and AGC turned off on low band use, with
green as high as 80dB with no red seen , for good weak signal decode.

Using low width roofing/dsp filtering around the signal boosts it to
great extent, using width/shift to focus.

______________________________________________________________
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
Message delivered to [hidden email]

wayne burdick

Re: K4 Question

Administrator

In reply to this post by wayne burdick

Hi Eric,

I did a quick test using my home lab to simulate a possible FD scenario.

For this test I set up a K4D and a K3S with their antenna jacks connected directly together through a high-power attenuator. Receiver preamps were off. With this arrangement, the RX noise floor is minimized since there's no actual antenna involved.

I set mutual attenuation to 30 dB, a rough estimate of the path loss using dipoles 500' apart at 7 MHz. This is a pretty wild guess, though. Loss could be much higher if the antennas were oriented to avoid coupling, and it'll vary with frequency, terrain, actual distance, etc. Of course path loss could be lower with gain antennas at either or both ends, aimed at the other. (A situation generally avoided at FD.)

While transmitting with the K4D at 100 W and receiving with the K3S, I found I was engaging the K3S's carrier-operated relay. This is evidence that the path loss probably is higher than 30 dB in real-world scenarios. I dropped to 10 W on both rigs (10 dB down from 100 W) to avoid the confound.

I then coupled in a weak signal at the equivalent of about S2 (-113 dBm) as indicated on both receivers. When keying one rig, there was no evidence of desensing of this signal at the other, and only a very slight observed increase in the noise floor (as indicated by the respective panadapters).

Yes, the two radios have entirely different architectures. Each has pros and cons.

With an SDR like the K4, the fundamental limit on narrowband TX noise performance is the DAC. The K3S, on the other hand, has to shoehorn its 8 MHz IF transmit signal through a narrow crystal filter, adding ripple and group delay to complex signals (like voice and data). It also exhibits a characteristic "pedestal" of 15 kHz DAC noise that sits maybe 15 to 20 dB above the wideband noise floor.

When it comes to CW keying bandwidth, both the K3S and K4 have essentially identical (and excellent) performance due to an optimally shaped keying envelope.

73,
Wayne
N6KR

> On Jun 23, 2020, at 3:46 PM, Wayne Burdick <[hidden email]> wrote:
>
> Hi Eric,
>
> S9+65 dB is about -8 dBm. Off the top of my head, this is far, far below what a basic K4 or K4D can handle, artifact-free, in-band, without the need for attenuation or additional filtering. When I get back to the lab I'm going to set up exactly this condition and get back to you.
>
> Of course the out-of-band rejection is even higher.
>
> A number of K4s will be used extensively during FD this year, including mine. I'll be taking advantage of the K4's low current drain (for its class) by running mine from a KX2 11 volt battery pack (3x 18650 cells). For at least an hour or so :)
>
> 73,
> Wayne
> N6KR
>
>
>> From: "Eric Norris" <[hidden email]>
>> To: "elecraft@mailman qth. net" <[hidden email]>
>> Sent: Tuesday, June 23, 2020 9:32:04 PM
>> Subject: [Elecraft] K4 Question
>>
>> My foaming at the mouth over the K4 has been tempered by it having less
>> adjacent channel rejection than the K3 due to the different
>> architecture--at least until the K4HD comes out. I understand this, and
>> the reasons why. Thanks for your answers
>>
>> I know I have asked this question before, but I want to be more specific.
>> My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away on
>> ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
>> perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
>> other degradation, or would I be able to carry on a CW qso unmolested like
>> I can with my K3? What about an adjacent band like 3534 kHz or 10114 kHz?
>> Or is the answer I have to wait for the K4HD?
>>
>> No speculation, please, I'm looking for a real-world or lab-world answer.
>>
>> Thanks and,
>>
>> 73 Eric WD6DBM
>
>
>
>

wayne burdick

Re: K4 Question

Administrator

I should have mentioned that with 30 dB simulated path loss and 10 W TX, the received signal at each end is the equivalent of about +10 dBm, or "S9+83 dB". FWIW :)

73,
Wayne
N6KR

> On Jun 23, 2020, at 8:16 PM, Wayne Burdick <[hidden email]> wrote:
>
> Hi Eric,
>
> I did a quick test using my home lab to simulate a possible FD scenario.
>
> For this test I set up a K4D and a K3S with their antenna jacks connected directly together through a high-power attenuator. Receiver preamps were off. With this arrangement, the RX noise floor is minimized since there's no actual antenna involved.
>
> I set mutual attenuation to 30 dB, a rough estimate of the path loss using dipoles 500' apart at 7 MHz. This is a pretty wild guess, though. Loss could be much higher if the antennas were oriented to avoid coupling, and it'll vary with frequency, terrain, actual distance, etc. Of course path loss could be lower with gain antennas at either or both ends, aimed at the other. (A situation generally avoided at FD.)
>
> While transmitting with the K4D at 100 W and receiving with the K3S, I found I was engaging the K3S's carrier-operated relay. This is evidence that the path loss probably is higher than 30 dB in real-world scenarios. I dropped to 10 W on both rigs (10 dB down from 100 W) to avoid the confound.
>
> I then coupled in a weak signal at the equivalent of about S2 (-113 dBm) as indicated on both receivers. When keying one rig, there was no evidence of desensing of this signal at the other, and only a very slight observed increase in the noise floor (as indicated by the respective panadapters).
>
> Yes, the two radios have entirely different architectures. Each has pros and cons.
>
> With an SDR like the K4, the fundamental limit on narrowband TX noise performance is the DAC. The K3S, on the other hand, has to shoehorn its 8 MHz IF transmit signal through a narrow crystal filter, adding ripple and group delay to complex signals (like voice and data). It also exhibits a characteristic "pedestal" of 15 kHz DAC noise that sits maybe 15 to 20 dB above the wideband noise floor.
>
> When it comes to CW keying bandwidth, both the K3S and K4 have essentially identical (and excellent) performance due to an optimally shaped keying envelope.
>
> 73,
> Wayne
> N6KR
>
>
>
>> On Jun 23, 2020, at 3:46 PM, Wayne Burdick <[hidden email]> wrote:
>>
>> Hi Eric,
>>
>> S9+65 dB is about -8 dBm. Off the top of my head, this is far, far below what a basic K4 or K4D can handle, artifact-free, in-band, without the need for attenuation or additional filtering. When I get back to the lab I'm going to set up exactly this condition and get back to you.
>>
>> Of course the out-of-band rejection is even higher.
>>
>> A number of K4s will be used extensively during FD this year, including mine. I'll be taking advantage of the K4's low current drain (for its class) by running mine from a KX2 11 volt battery pack (3x 18650 cells). For at least an hour or so :)
>>
>> 73,
>> Wayne
>> N6KR
>>
>>
>>> From: "Eric Norris" <[hidden email]>
>>> To: "elecraft@mailman qth. net" <[hidden email]>
>>> Sent: Tuesday, June 23, 2020 9:32:04 PM
>>> Subject: [Elecraft] K4 Question
>>>
>>> My foaming at the mouth over the K4 has been tempered by it having less
>>> adjacent channel rejection than the K3 due to the different
>>> architecture--at least until the K4HD comes out. I understand this, and
>>> the reasons why. Thanks for your answers
>>>
>>> I know I have asked this question before, but I want to be more specific.
>>> My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away on
>>> ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
>>> perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
>>> other degradation, or would I be able to carry on a CW qso unmolested like
>>> I can with my K3? What about an adjacent band like 3534 kHz or 10114 kHz?
>>> Or is the answer I have to wait for the K4HD?
>>>
>>> No speculation, please, I'm looking for a real-world or lab-world answer.
>>>
>>> Thanks and,
>>>
>>> 73 Eric WD6DBM
>>
>>
>>
>>
>

donovanf

Re: K4 Question

In reply to this post by Jim Brown-10

Hi Jim

100 watts. Key to our success is the end-to-end alignment of our
3 element Yagis

73
Frank
W3LPL

----- Original Message -----

From: "Jim Brown" <[hidden email]>
To: [hidden email]
Sent: Wednesday, June 24, 2020 12:49:00 AM
Subject: Re: [Elecraft] K4 Question

On 6/23/2020 3:27 PM, [hidden email] wrote:
> We use more than a dozen K3 transceivers at the W3AO Field Day with
> as many as four operating in the same band (CW, SSB, FT8 and GOTA SSB),
> but not this year...

What power level, Frank?

73, Jim K9YC
______________________________________________________________
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
Message delivered to [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
Message delivered to [hidden email]

Vic Rosenthal

Re: K4 Question

In reply to this post by wayne burdick

Well, that would certainly be enough for me. I have two loud
line-of-sight neighbors, but they don't hit +10 dBm.
Unfortunately they have "dirty" transceivers, so it doesn't matter. I
suppose I should buy them K4s.

73,
Victor, 4X6GP
Rehovot, Israel
Formerly K2VCO
CWops no. 5
http://www.qsl.net/k2vco/
.
On 24/06/2020 6:38, Wayne Burdick wrote:

> I should have mentioned that with 30 dB simulated path loss and 10 W TX, the received signal at each end is the equivalent of about +10 dBm, or "S9+83 dB". FWIW :)
>
> 73,
> Wayne
> N6KR
>
>
>> On Jun 23, 2020, at 8:16 PM, Wayne Burdick <[hidden email]> wrote:
>>
>> Hi Eric,
>>
>> I did a quick test using my home lab to simulate a possible FD scenario.
>>
>> For this test I set up a K4D and a K3S with their antenna jacks connected directly together through a high-power attenuator. Receiver preamps were off. With this arrangement, the RX noise floor is minimized since there's no actual antenna involved.
>>
>> I set mutual attenuation to 30 dB, a rough estimate of the path loss using dipoles 500' apart at 7 MHz. This is a pretty wild guess, though. Loss could be much higher if the antennas were oriented to avoid coupling, and it'll vary with frequency, terrain, actual distance, etc. Of course path loss could be lower with gain antennas at either or both ends, aimed at the other. (A situation generally avoided at FD.)
>>
>> While transmitting with the K4D at 100 W and receiving with the K3S, I found I was engaging the K3S's carrier-operated relay. This is evidence that the path loss probably is higher than 30 dB in real-world scenarios. I dropped to 10 W on both rigs (10 dB down from 100 W) to avoid the confound.
>>
>> I then coupled in a weak signal at the equivalent of about S2 (-113 dBm) as indicated on both receivers. When keying one rig, there was no evidence of desensing of this signal at the other, and only a very slight observed increase in the noise floor (as indicated by the respective panadapters).
>>
>> Yes, the two radios have entirely different architectures. Each has pros and cons.
>>
>> With an SDR like the K4, the fundamental limit on narrowband TX noise performance is the DAC. The K3S, on the other hand, has to shoehorn its 8 MHz IF transmit signal through a narrow crystal filter, adding ripple and group delay to complex signals (like voice and data). It also exhibits a characteristic "pedestal" of 15 kHz DAC noise that sits maybe 15 to 20 dB above the wideband noise floor.
>>
>> When it comes to CW keying bandwidth, both the K3S and K4 have essentially identical (and excellent) performance due to an optimally shaped keying envelope.
>>
>> 73,
>> Wayne
>> N6KR

Eric Norris-2

Re: K4 Question

In reply to this post by wayne burdick

Hi Wayne,

How far apart in frequency were the TX and RX signals into the respective
radios? My guess is my neighbor is about 500ft away, but I will measure it
with a gps. He has never tripped my K3 or K3S COR when we are on the same
qrg. Thanks!

73 Eric WD6DBM

On Tue, Jun 23, 2020, 8:16 PM Wayne Burdick <[hidden email]> wrote:

> Hi Eric,
>
> I did a quick test using my home lab to simulate a possible FD scenario.
>
> For this test I set up a K4D and a K3S with their antenna jacks connected
> directly together through a high-power attenuator. Receiver preamps were
> off. With this arrangement, the RX noise floor is minimized since there's
> no actual antenna involved.
>
> I set mutual attenuation to 30 dB, a rough estimate of the path loss using
> dipoles 500' apart at 7 MHz. This is a pretty wild guess, though. Loss
> could be much higher if the antennas were oriented to avoid coupling, and
> it'll vary with frequency, terrain, actual distance, etc. Of course path
> loss could be lower with gain antennas at either or both ends, aimed at the
> other. (A situation generally avoided at FD.)
>
> While transmitting with the K4D at 100 W and receiving with the K3S, I
> found I was engaging the K3S's carrier-operated relay. This is evidence
> that the path loss probably is higher than 30 dB in real-world scenarios. I
> dropped to 10 W on both rigs (10 dB down from 100 W) to avoid the confound.
>
> I then coupled in a weak signal at the equivalent of about S2 (-113 dBm)
> as indicated on both receivers. When keying one rig, there was no evidence
> of desensing of this signal at the other, and only a very slight observed
> increase in the noise floor (as indicated by the respective panadapters).
>
> Yes, the two radios have entirely different architectures. Each has pros
> and cons.
>
> With an SDR like the K4, the fundamental limit on narrowband TX noise
> performance is the DAC. The K3S, on the other hand, has to shoehorn its 8
> MHz IF transmit signal through a narrow crystal filter, adding ripple and
> group delay to complex signals (like voice and data). It also exhibits a
> characteristic "pedestal" of 15 kHz DAC noise that sits maybe 15 to 20 dB
> above the wideband noise floor.
>
> When it comes to CW keying bandwidth, both the K3S and K4 have essentially
> identical (and excellent) performance due to an optimally shaped keying
> envelope.
>
> 73,
> Wayne
> N6KR
>
>
>
> > On Jun 23, 2020, at 3:46 PM, Wayne Burdick <[hidden email]> wrote:
> >
> > Hi Eric,
> >
> > S9+65 dB is about -8 dBm. Off the top of my head, this is far, far below
> what a basic K4 or K4D can handle, artifact-free, in-band, without the need
> for attenuation or additional filtering. When I get back to the lab I'm
> going to set up exactly this condition and get back to you.
> >
> > Of course the out-of-band rejection is even higher.
> >
> > A number of K4s will be used extensively during FD this year, including
> mine. I'll be taking advantage of the K4's low current drain (for its
> class) by running mine from a KX2 11 volt battery pack (3x 18650 cells).
> For at least an hour or so :)
> >
> > 73,
> > Wayne
> > N6KR
> >
> >
> >> From: "Eric Norris" <[hidden email]>
> >> To: "elecraft@mailman qth. net" <[hidden email]>
> >> Sent: Tuesday, June 23, 2020 9:32:04 PM
> >> Subject: [Elecraft] K4 Question
> >>
> >> My foaming at the mouth over the K4 has been tempered by it having less
> >> adjacent channel rejection than the K3 due to the different
> >> architecture--at least until the K4HD comes out. I understand this, and
> >> the reasons why. Thanks for your answers
> >>
> >> I know I have asked this question before, but I want to be more
> specific.
> >> My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away
> on
> >> ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
> >> perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
> >> other degradation, or would I be able to carry on a CW qso unmolested
> like
> >> I can with my K3? What about an adjacent band like 3534 kHz or 10114
> kHz?
> >> Or is the answer I have to wait for the K4HD?
> >>
> >> No speculation, please, I'm looking for a real-world or lab-world
> answer.
> >>
> >> Thanks and,
> >>
> >> 73 Eric WD6DBM
> >
> >
> >
> >
>
>

Eric Norris-2

Re: K4 Question

In reply to this post by wayne burdick

Hi Wayne,

That gives me almost exactly the overhead I need if my neighbor buys an
amp--or puts up a beam. Thanks!

73 Eric WD6DBM

On Tue, Jun 23, 2020, 8:39 PM Wayne Burdick <[hidden email]> wrote:

> I should have mentioned that with 30 dB simulated path loss and 10 W TX,
> the received signal at each end is the equivalent of about +10 dBm, or
> "S9+83 dB". FWIW :)
>
> 73,
> Wayne
> N6KR
>
>
> > On Jun 23, 2020, at 8:16 PM, Wayne Burdick <[hidden email]> wrote:
> >
> > Hi Eric,
> >
> > I did a quick test using my home lab to simulate a possible FD scenario.
> >
> > For this test I set up a K4D and a K3S with their antenna jacks
> connected directly together through a high-power attenuator. Receiver
> preamps were off. With this arrangement, the RX noise floor is minimized
> since there's no actual antenna involved.
> >
> > I set mutual attenuation to 30 dB, a rough estimate of the path loss
> using dipoles 500' apart at 7 MHz. This is a pretty wild guess, though.
> Loss could be much higher if the antennas were oriented to avoid coupling,
> and it'll vary with frequency, terrain, actual distance, etc. Of course
> path loss could be lower with gain antennas at either or both ends, aimed
> at the other. (A situation generally avoided at FD.)
> >
> > While transmitting with the K4D at 100 W and receiving with the K3S, I
> found I was engaging the K3S's carrier-operated relay. This is evidence
> that the path loss probably is higher than 30 dB in real-world scenarios. I
> dropped to 10 W on both rigs (10 dB down from 100 W) to avoid the confound.
> >
> > I then coupled in a weak signal at the equivalent of about S2 (-113 dBm)
> as indicated on both receivers. When keying one rig, there was no evidence
> of desensing of this signal at the other, and only a very slight observed
> increase in the noise floor (as indicated by the respective panadapters).
> >
> > Yes, the two radios have entirely different architectures. Each has pros
> and cons.
> >
> > With an SDR like the K4, the fundamental limit on narrowband TX noise
> performance is the DAC. The K3S, on the other hand, has to shoehorn its 8
> MHz IF transmit signal through a narrow crystal filter, adding ripple and
> group delay to complex signals (like voice and data). It also exhibits a
> characteristic "pedestal" of 15 kHz DAC noise that sits maybe 15 to 20 dB
> above the wideband noise floor.
> >
> > When it comes to CW keying bandwidth, both the K3S and K4 have
> essentially identical (and excellent) performance due to an optimally
> shaped keying envelope.
> >
> > 73,
> > Wayne
> > N6KR
> >
> >
> >
> >> On Jun 23, 2020, at 3:46 PM, Wayne Burdick <[hidden email]> wrote:
> >>
> >> Hi Eric,
> >>
> >> S9+65 dB is about -8 dBm. Off the top of my head, this is far, far
> below what a basic K4 or K4D can handle, artifact-free, in-band, without
> the need for attenuation or additional filtering. When I get back to the
> lab I'm going to set up exactly this condition and get back to you.
> >>
> >> Of course the out-of-band rejection is even higher.
> >>
> >> A number of K4s will be used extensively during FD this year, including
> mine. I'll be taking advantage of the K4's low current drain (for its
> class) by running mine from a KX2 11 volt battery pack (3x 18650 cells).
> For at least an hour or so :)
> >>
> >> 73,
> >> Wayne
> >> N6KR
> >>
> >>
> >>> From: "Eric Norris" <[hidden email]>
> >>> To: "elecraft@mailman qth. net" <[hidden email]>
> >>> Sent: Tuesday, June 23, 2020 9:32:04 PM
> >>> Subject: [Elecraft] K4 Question
> >>>
> >>> My foaming at the mouth over the K4 has been tempered by it having
> less
> >>> adjacent channel rejection than the K3 due to the different
> >>> architecture--at least until the K4HD comes out. I understand this,
> and
> >>> the reasons why. Thanks for your answers
> >>>
> >>> I know I have asked this question before, but I want to be more
> specific.
> >>> My QRM neighbor is S9+65 db on my K3S S-meter. If he is blasting away
> on
> >>> ft8 at 7074 kHz at that signal level, how would the plain K4 receiver
> >>> perform at 7034 kHz on CW? Would there be AGC pumping, RX desense, or
> >>> other degradation, or would I be able to carry on a CW qso unmolested
> like
> >>> I can with my K3? What about an adjacent band like 3534 kHz or 10114
> kHz?
> >>> Or is the answer I have to wait for the K4HD?
> >>>
> >>> No speculation, please, I'm looking for a real-world or lab-world
> answer.
> >>>
> >>> Thanks and,
> >>>
> >>> 73 Eric WD6DBM
> >>
> >>
> >>
> >>
> >
>
>

Greg Troxel-2

Re: K4 Question

In reply to this post by wayne burdick

Wayne Burdick <[hidden email]> writes:

> I set mutual attenuation to 30 dB, a rough estimate of the path loss
> using dipoles 500' apart at 7 MHz. This is a pretty wild guess,
> though. Loss could be much higher if the antennas were oriented to
> avoid coupling, and it'll vary with frequency, terrain, actual
> distance, etc. Of course path loss could be lower with gain antennas
> at either or both ends, aimed at the other. (A situation generally
> avoided at FD.)

I think 30 dB loss is an extremely conservative test and that very few
people will see coupling that strong.

At my club's 2019 Field Day, with help from several, I made measurements
of received signal strength at several of our stations with various
antennas, with a nominal 100W test carrier. Measurements are just
reading a P3 -- and note that none of the P3 owners had paid for the
NIST-traceable calibration certificate.

Setup was CW station with K3newsyn, digital station with IC7300, only
20m separation from one CW antenna to the digital antenna. And SSB
station with K3newsyn, 200m away.

We did not have trouble, but always had at least 50 kHz separation.

Received levels ranged from S9+40 to S9+73, with S9+50 typical.

So received signals were

-33 dBm min, -23 dBm typical, 0dBm max

With +50 dBm transmit, that works out to path loss:
83 dB max, 73 dB typical, 50 dB min

(The 0 dBm received signal was between an OCFD and a G5RV, about 20m apart.)

It is interesting to hear of S9+65 from a neighbor. Even if they are
running 1.5 kW, seems like it must be only a few hundred meters
separation. If it's farther the details of distance, tx power, antenna
types would be interesting.

I realize this is not responsive to Eric's question about the K4, but
thought that additional real-world cosite path loss data points would be
of interest..

73 de n1dam
______________________________________________________________
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
Message delivered to [hidden email]

k6dgw

Re: K4 Question

During the BPL conflagration a number of years ago, most of the concern
seemed to be interference "to" amateur radio operations. I became a bit
concerned that my neighbor would go BPL and my 1200 HF watts would kill
his I'net. Technically and legally his problem, not mine of course, but
I still needed to get along with him.

We had a 69 KV line running through our property tying two hydroelectric
plants together, with a primary 12 KV distribution line underneath which
would have been the BPL feed. At closest approach my antenna and the
lines were about 165 m apart. I modeled them and my sloping-V in EZNEC
and found that the coupling between them was -32 to -35 dB on all bands
except 80 and 160, where it ran about -17 dB on both bands. On 80 and
160, the power lines were in my near-field.

So, I would conclude that Wayne's 30 dB, based on a different model, is
a pretty good estimate. Fortunately, BPL faded into the sunset and I
was never confronted with the problem.

73,
Fred ["Skip"] K6DGW
Sparks NV DM09dn
Washoe County

On 6/24/2020 12:12 PM, Greg Troxel wrote:

> Wayne Burdick <[hidden email]> writes:
>
>> I set mutual attenuation to 30 dB, a rough estimate of the path loss
>> using dipoles 500' apart at 7 MHz. This is a pretty wild guess,
>> though. Loss could be much higher if the antennas were oriented to
>> avoid coupling, and it'll vary with frequency, terrain, actual
>> distance, etc. Of course path loss could be lower with gain antennas
>> at either or both ends, aimed at the other. (A situation generally
>> avoided at FD.)
> I think 30 dB loss is an extremely conservative test and that very few
> people will see coupling that strong.
>
> At my club's 2019 Field Day, with help from several, I made measurements
> of received signal strength at several of our stations with various
> antennas, with a nominal 100W test carrier. Measurements are just
> reading a P3 -- and note that none of the P3 owners had paid for the
> NIST-traceable calibration certificate.
>
> Setup was CW station with K3newsyn, digital station with IC7300, only
> 20m separation from one CW antenna to the digital antenna. And SSB
> station with K3newsyn, 200m away.
>
> We did not have trouble, but always had at least 50 kHz separation.
>
> Received levels ranged from S9+40 to S9+73, with S9+50 typical.
>
> So received signals were
>
> -33 dBm min, -23 dBm typical, 0dBm max
>
> With +50 dBm transmit, that works out to path loss:
> 83 dB max, 73 dB typical, 50 dB min
>
> (The 0 dBm received signal was between an OCFD and a G5RV, about 20m apart.)
>
> It is interesting to hear of S9+65 from a neighbor. Even if they are
> running 1.5 kW, seems like it must be only a few hundred meters
> separation. If it's farther the details of distance, tx power, antenna
> types would be interesting.
>
> I realize this is not responsive to Eric's question about the K4, but
> thought that additional real-world cosite path loss data points would be
> of interest..
>
> 73 de n1dam
> ______________________________________________________________
> 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
> Message delivered to [hidden email]

Eric Norris-2

Re: K4 Question

In reply to this post by Greg Troxel-2

My neighbor is 714ft (218m) away, running an ic7300 at 100w into a ZS6BW
antenna at 40ft--aimed in my general direction.

73 Eric WD6DBM

On Wed, Jun 24, 2020, 12:12 PM Greg Troxel <[hidden email]> wrote:

> Wayne Burdick <[hidden email]> writes:
>
> > I set mutual attenuation to 30 dB, a rough estimate of the path loss
> > using dipoles 500' apart at 7 MHz. This is a pretty wild guess,
> > though. Loss could be much higher if the antennas were oriented to
> > avoid coupling, and it'll vary with frequency, terrain, actual
> > distance, etc. Of course path loss could be lower with gain antennas
> > at either or both ends, aimed at the other. (A situation generally
> > avoided at FD.)
>
> I think 30 dB loss is an extremely conservative test and that very few
> people will see coupling that strong.
>
> At my club's 2019 Field Day, with help from several, I made measurements
> of received signal strength at several of our stations with various
> antennas, with a nominal 100W test carrier. Measurements are just
> reading a P3 -- and note that none of the P3 owners had paid for the
> NIST-traceable calibration certificate.
>
> Setup was CW station with K3newsyn, digital station with IC7300, only
> 20m separation from one CW antenna to the digital antenna. And SSB
> station with K3newsyn, 200m away.
>
> We did not have trouble, but always had at least 50 kHz separation.
>
> Received levels ranged from S9+40 to S9+73, with S9+50 typical.
>
> So received signals were
>
> -33 dBm min, -23 dBm typical, 0dBm max
>
> With +50 dBm transmit, that works out to path loss:
> 83 dB max, 73 dB typical, 50 dB min
>
> (The 0 dBm received signal was between an OCFD and a G5RV, about 20m
> apart.)
>
> It is interesting to hear of S9+65 from a neighbor. Even if they are
> running 1.5 kW, seems like it must be only a few hundred meters
> separation. If it's farther the details of distance, tx power, antenna
> types would be interesting.
>
> I realize this is not responsive to Eric's question about the K4, but
> thought that additional real-world cosite path loss data points would be
> of interest..
>
> 73 de n1dam
>

K5WA

Re: K4 Question

In reply to this post by Eric Norris-2

Bob,

Thanks for taking the time to make this second recording and put up with us arm chair quarterbacks who are critiquing the results. 😉

I can’t wait to make my own recordings (assuming Gov. Newsome allows CA factories to open a bit more) and have critiques come my way. In the meantime, I’m drooling over the thought of having a K4 and getting my flame suit ready to go! 😉

Thanks es 73,

Bob K5WA

______________________________________________________________
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
Message delivered to [hidden email]