Elecraft

K3s - 80W on 160

Classic

List

Threaded

31 messages Options

Alan Bloom

Re: IMD and CW

For a CW signal, the nonlinearity of a typical power amplifier should
have the effect of shortening the rise and fall times. That does indeed
widen the transmitted bandwidth. But intuitively, it seems like the
distortion would have to be really bad to shorten the rise/fall times by
much. You could always compensate for it by adjusting the K3 for softer
key shaping.

Alan N1AL

On 12/22/2016 10:22 PM, Vic Rosenthal wrote:

> I didn't exactly do this, but I did investigate the effects of distortion on CW signals. I disconnected my P3 from the K3 and set the IF to 7 MHz. so it would act like a standalone spectrum analyzer.
> I was interested in whether a class-C amplifier I had built would sharpen the keying enough to create noticeable clicks. The amplifier used minimal fixed bias to allow a small amount of resting plate current in transmit mode, with full class-C operating bias provided by a grid resistor.
> I noted a broadening of the base of the signal on the P3, indicating that the signal was not as clean as the K3 alone. But listening with another receiver on either side of the S9+30 dB signal didn't reveal audible clicks. I also asked other operators to listen and there were no reports of clicks.
> I am sure that attempting to use that amplifier (a pair of 813s) on SSB would have gotten me ridden out of town on a rail!
>
> Vic 4X6GP
>
>> On 22 Dec 2016, at 23:10, K9MA <[hidden email]> wrote:
>>
>>> On 12/22/2016 11:40, Jim Brown wrote:
>>> IMD also affects CW. What we call CW is not "continuous wave," it is a continuous wave that is 100% modulated by a rectangular wave. That rectangular wave is rich in harmonics, and excites IMD. We hear it as clicks.
>> It would be interesting to see how much IMD it takes to significantly increase key clicks of a "clean" keyed CW signal, like that of the K3. I'd expect the effect to be small, compared to that on an SSB signal for the same level of IMD, but I could be wrong. I suppose a simple test would be to have a local friend with a P3 watch the signal as I turned down the supply voltage on my K3. Has anyone tried anything like that?
>>
>> 73,
>>
>> Scott K9MA
>>
>>
>> --
>> Scott K9MA
>>
>> [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]
> ______________________________________________________________
> 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]

K9MA

Re: IMD and CW

On 12/23/2016 12:11, Alan Bloom wrote:
> For a CW signal, the nonlinearity of a typical power amplifier should
> have the effect of shortening the rise and fall times. That does
> indeed widen the transmitted bandwidth. But intuitively, it seems
> like the distortion would have to be really bad to shorten the
> rise/fall times by much. You could always compensate for it by
> adjusting the K3 for softer key shaping.
>
> Alan N1AL
I think it's a little more complicated than that: The K3 shapes the CW
waveform very carefully, to minimize bandwidth. (I think it's
equivalent to passing the signal through a narrow bandpass filter.)
Obviously, any distortion downstream is going to compromise that careful
shaping. However, based on my intuition and Vic's results, I suspect it
would take a truly terrible amount of distortion to significantly
increase the CW bandwidth. Given that the K3 is so much cleaner than
many other popular radios, I doubt we need to worry.

73,

Scott K9MA

--
Scott K9MA

[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]

k6dgw

Re: IMD and CW

In reply to this post by Vic Rosenthal

Not CW, and this probably won't work in 4X-land, but if you want to see
the difference a Class-C amp makes, set your P3 span to about 6 KHz, and
tune in WWV on its various frequencies. During the tone-on minutes, the
2.5 and 20 MHz signals will look as you expect an AM signal with a tone
modulating it ... two sidebands on either side of the carrier. If you
look very close, you can see the 100 Hz sub-carrier, and if you narrow
things down enough, you can see the sidebands generated around it by the
IRIG-H time code. These two transmitters are low level modulated with
linear amplification.

The 5, 10, and 15 MHz transmitters are plate-modulated Class-C and look
markedly different. The two sidebands are there, but so are their 2nd
and 3rd harmonics [sometimes when WWV is really strong, I can make out
the 4th harmonic]. The 100 Hz sub-carrier and the IRIG-H sidebands are
there as you would expect, but they also surround the audio sidebands
and harmonics.

WWV builds one baseband with all the tones, ticks, burps, codes, and
voice and it modulates all 5 transmitters. The difference between the
TX is really striking and a bit disappointing. I've always assumed WWV
was the "gold-standard."

Fred K6DGW
Sparks NV DM09dn
Washoe County

--Northern California Contest Club
--CU in the Cal QSO Party
--7-8 Oct 2017

On 12/22/2016 10:22 PM, Vic Rosenthal wrote:
> I didn't exactly do this, but I did investigate the effects of distortion on CW signals. I disconnected my P3 from the K3 and set the IF to 7 MHz. so it would act like a standalone spectrum analyzer.
> I was interested in whether a class-C amplifier I had built would sharpen the keying enough to create noticeable clicks. The amplifier used minimal fixed bias to allow a small amount of resting plate current in transmit mode, with full class-C operating bias provided by a grid resistor.
> I noted a broadening of the base of the signal on the P3, indicating that the signal was not as clean as the K3 alone. But listening with another receiver on either side of the S9+30 dB signal didn't reveal audible clicks. I also asked other operators to listen and there were no reports of clicks.
> I am sure that attempting to use that amplifier (a pair of 813s) on SSB would have gotten me ridden out of town on a rail!
>
> Vic 4X6GP
>

______________________________________________________________
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]

K9MA

Re: IMD and CW

Very interesting, Fred. It makes sense, though, that plate modulated
class-C amplifiers aren't terribly linear. On the other hand, that kind
of distortion probably doesn't affect WWV's mission.

I wonder if the AM broadcast industry achieves lower distortion.
Feedback comes to mind.

73,

Scott K9MA

On 12/23/2016 15:53, Fred Jensen wrote:

> Not CW, and this probably won't work in 4X-land, but if you want to
> see the difference a Class-C amp makes, set your P3 span to about 6
> KHz, and tune in WWV on its various frequencies. During the tone-on
> minutes, the 2.5 and 20 MHz signals will look as you expect an AM
> signal with a tone modulating it ... two sidebands on either side of
> the carrier. If you look very close, you can see the 100 Hz
> sub-carrier, and if you narrow things down enough, you can see the
> sidebands generated around it by the IRIG-H time code. These two
> transmitters are low level modulated with linear amplification.
>
> The 5, 10, and 15 MHz transmitters are plate-modulated Class-C and
> look markedly different. The two sidebands are there, but so are
> their 2nd and 3rd harmonics [sometimes when WWV is really strong, I
> can make out the 4th harmonic]. The 100 Hz sub-carrier and the IRIG-H
> sidebands are there as you would expect, but they also surround the
> audio sidebands and harmonics.
>
> WWV builds one baseband with all the tones, ticks, burps, codes, and
> voice and it modulates all 5 transmitters. The difference between the
> TX is really striking and a bit disappointing. I've always assumed
> WWV was the "gold-standard."
>
> Fred K6DGW
> Sparks NV DM09dn
> Washoe County
>
> --Northern California Contest Club
> --CU in the Cal QSO Party
> --7-8 Oct 2017
>
> On 12/22/2016 10:22 PM, Vic Rosenthal wrote:
>> I didn't exactly do this, but I did investigate the effects of
>> distortion on CW signals. I disconnected my P3 from the K3 and set
>> the IF to 7 MHz. so it would act like a standalone spectrum analyzer.
>> I was interested in whether a class-C amplifier I had built would
>> sharpen the keying enough to create noticeable clicks. The amplifier
>> used minimal fixed bias to allow a small amount of resting plate
>> current in transmit mode, with full class-C operating bias provided
>> by a grid resistor.
>> I noted a broadening of the base of the signal on the P3, indicating
>> that the signal was not as clean as the K3 alone. But listening with
>> another receiver on either side of the S9+30 dB signal didn't reveal
>> audible clicks. I also asked other operators to listen and there were
>> no reports of clicks.
>> I am sure that attempting to use that amplifier (a pair of 813s) on
>> SSB would have gotten me ridden out of town on a rail!
>>
>> Vic 4X6GP
>>
>
> ______________________________________________________________
> 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]

--
Scott K9MA

[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]

k6dgw

Re: IMD and CW

Google "ampliphase" ... OKA: "The Transmitter all broadcast engineers
hate." All linear phase modulation until AM magically squirts out the top.

Factoid: When WWV was brand new, it transmitted farm reports for the
Agriculture Department ... using Morse code. I'm guessing the number of
farmers fluent in Morse might have been two ... three tops. I remember
WWV using Morse, but by then it was Eastern Standard Time, not farm reports.

Fred K6DGW
Sparks NV DM09dn
Washoe County

--Northern California Contest Club
--CU in the Cal QSO Party
--7-8 Oct 2017On 12/23/2016 2:39 PM, K9MA wrote:

> Very interesting, Fred. It makes sense, though, that plate modulated
> class-C amplifiers aren't terribly linear. On the other hand, that
> kind of distortion probably doesn't affect WWV's mission.
>
> I wonder if the AM broadcast industry achieves lower distortion.
> Feedback comes to mind.
>
> 73,
>
> Scott K9MA

Alan Bloom

Re: IMD and CW

In reply to this post by K9MA

I believe the K3/K3S use a raised sinusoid key shaping. It is a good
compromise to reduce the bandwidth without excessive ringing/backwave.

In the bad old days, CW key shaping was typically a simple R-C filter on
the key line. That resulted in an exponential rise and fall, which
resulted in more key clicks for a given rise/fall time than a raised
sinusoid.

Alan N1AL

On 12/23/2016 01:30 PM, K9MA wrote:

> On 12/23/2016 12:11, Alan Bloom wrote:
>> For a CW signal, the nonlinearity of a typical power amplifier should
>> have the effect of shortening the rise and fall times. That does
>> indeed widen the transmitted bandwidth. But intuitively, it seems
>> like the distortion would have to be really bad to shorten the
>> rise/fall times by much. You could always compensate for it by
>> adjusting the K3 for softer key shaping.
>>
>> Alan N1AL
> I think it's a little more complicated than that: The K3 shapes the CW
> waveform very carefully, to minimize bandwidth. (I think it's
> equivalent to passing the signal through a narrow bandpass filter.)
> Obviously, any distortion downstream is going to compromise that careful
> shaping. However, based on my intuition and Vic's results, I suspect it
> would take a truly terrible amount of distortion to significantly
> increase the CW bandwidth. Given that the K3 is so much cleaner than
> many other popular radios, I doubt we need to worry.
>
> 73,
>
> Scott K9MA
>

k6dgw

Re: IMD and CW

In reply to this post by K9MA

This has drifted fairly far from the original. Thus encouraged ...

I wondered about that and being retired I pursued it, ultimately with
the tech folks [well, one folk] at WWV who repeatedly assured me that
they were on-frequency and that their time information was correct which
of course was never the issue. Somewhere in all the words I read about
the station, I did find a reference to plate modulated Class-C
transmitters but I have lost it's QTH on this disk drive.

Most plate modulators ran Class-B or -AB, and were subject to cross-over
non-linearities. The 5, 10, and 15 MHz signals look very much the same
on the spectrum display which [weakly] suggests the unexpected
distortion products may arise somewhere in the baseband chain. The 2.5
and 20 MHz transmitters, being low-level modulated, may tap that chain
before the distortion is introduced.

ARC-5's, when cathode-keyed, were notorious for key clicks, almost as
bad as the Yeasu rigs of recent eras. [:-) Of course, for my K3, the
"carrier-balance" and "opposite sideband suppression" is perfect. I
think, but don't know, that the K3 shapes the CW with a raised-cosine
filter. With strong signals, it *is* possible to identify a K3 by its
CW spectrum, particularly in the WF.

Fred K6DGW
Sparks NV DM09dn
Washoe County

--Northern California Contest Club
--CU in the Cal QSO Party
--7-8 Oct 2017

On 12/23/2016 3:46 PM, Ron D'Eau Claire wrote:

> The fact that the RF amplifier is running Class C should not be an issue if
> it is plate-modulated. Plate modulation is the "gold standard" because only
> the RF carrier is being amplified in the Class C stage. Beyond that its
> non-linearity is employed to mix the modulating signals with the RF carrier.
> While the process is extremely linear, it takes a lot of audio level power -
> at least 1/2 of the RF carrier power.
>
> Fred's observation makes me think WWV may be using some sort of "compromise"
> system such as grid modulation rather than plate modulation. Much lower
> audio (baseband) power required, but it is not as linear either.
>
> 73, Ron AC7AC
>

Jan Erik Holm-2

Re: IMD and CW

In reply to this post by Alan Bloom

On 2016-12-23 19:11, Alan Bloom wrote:
>
> You could always compensate for it by adjusting the K3 for softer
> key shaping.
>
> Alan N1AL
>
How do you do that ?

/ 73 Jim
______________________________________________________________
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]

Alan Bloom

Re: IMD and CW

Apparently you can't. I must have been thinking of the weight adjustment.

Alan N1AL

On 12/23/2016 09:22 PM, Jan Erik Holm wrote:

> On 2016-12-23 19:11, Alan Bloom wrote:
>>
>> You could always compensate for it by adjusting the K3 for softer
>> key shaping.
>>
>> Alan N1AL
>>
> How do you do that ?
>
> / 73 Jim

Jan Erik Holm-2

Re: IMD and CW

Yes was thinking it was something new.
Thank´s Alan

73 merry xmas / Jim SM2EKM
----------------------------------------
On 2016-12-24 06:23, Alan Bloom wrote:

> Apparently you can't. I must have been thinking of the weight adjustment.
>
> Alan N1AL
>
>
> On 12/23/2016 09:22 PM, Jan Erik Holm wrote:
>> On 2016-12-23 19:11, Alan Bloom wrote:
>>>
>>> You could always compensate for it by adjusting the K3 for softer
>>> key shaping.
>>>
>>> Alan N1AL
>>>
>> How do you do that ?
>>
>> / 73 Jim
>

Jim Brown-10

Re: IMD and CW

In reply to this post by Alan Bloom

On Fri,12/23/2016 10:11 AM, Alan Bloom wrote:
> You could always compensate for it by adjusting the K3 for softer key
> shaping.

Wayne carefully shaped the keying waveform to minimize occupied
bandwidth while maintaining good clarity at high keying speeds, and he
purposely did not allow users to screw it up. :)

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]