Elecraft › [K3]

Fwd: New Sherwood report

Classic

List

Threaded

82 messages Options

12345

alorona

Re: On the K3's CW keying envelope shaping, rise/fall times, and CW ALC

Several years ago when the K2 modification was created by Wayne to implement a
raised cosine keying filter I installed it and both ran simulations and made
measurements of the resulting waveform, which agreed very closely. In fact, I
liked the result so much that I duplicated the circuit and modified my Omni 6
with it. I just sold that rig and feel good that the next owner will put a very
clean-keying rig out there in the future.

In private conversations with several good contesters over the years I have the
general impression that although none of them would intentionally modify their
rigs to splatter (neither on CW nor SSB), they regard a clicking/splattering
transmitter as creating the happy side effect of a virtual guard band on either
side of them in a contest.

I also notice that whenever I am running with the K2 or K3 other contesters tend
to crowd me just a tad closer than I'm used to on one side or the other. It
annoys me. An unhappy side effect of exceptionally clean keying? Naw, more
likely just my weak, little-pistol signal!

Al W6LX
______________________________________________________________
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

Re: Fwd: New Sherwood report

In reply to this post by Don Wilhelm-4

> Another way of investigating is with a spectrum analyzer - and for
> investigation of a transmitted signal on one band, it does not have to
> be expensive (but must be homebrewed to be inexpensive).

Another solution for simple spectrum analysis is SDR-IQ and the CP-1
directional coupler are very good for looking at transmitter outputs
while the SDR-IQ and an active antenna like the one from Clifton
Laboratories (K8OZA) are good for "off air" use.

While this may not be as inexpensive as a homebrewed solution, it
can be very cost effective if one already owns the SDR-IQ for other
purposes.

73,

... Joe, W4TV

On 12/3/2010 11:41 AM, Don Wilhelm wrote:

> Paul and all,
>
> Another way of investigating is with a spectrum analyzer - and for
> investigation of a transmitted signal on one band, it does not have to
> be expensive (but must be homebrewed to be inexpensive). See the
> implementation by G4AON at http://www.astromag.co.uk/ssa/
>
> It is quite a nice narrowband spectrum analyzer - you can see the
> display using Spectrogram or Spectrum Lab or most any other audio
> spectrum analyzer running on the shack's soundcard equipped PC.
>
> You are correct, a 'scope is the most useful tool for measuring timing
> and amplitude.
>
> 73,
> Don W3FPR
>
> On 12/3/2010 10:55 AM, Paul Christensen wrote:
>>> Why, yes IMO the K3 clicks too much.
>> Jim,
>>
>> If you're experiencing clicks with your K3, you really need to investigate
>> with a scope. At the risk of pontificating, every station should have a
>> scope to monitor the transmit waveform. It can also be used to effectively
>> monitor T/R sequencing times when using an external amp. I suspect many
>> stations are hot-switching their amps and don't even know it until a T/R
>> relay's contacts fuse and the relay fails. Cost is no longer an excuse.
>> Anyone who can afford a K3 or FTdx5K can own a scope. Good quality used
>> scopes can be purchased on the surplus market between USD $100-200.
>>
>> If I may make a plug for N8LP, he's got a forthcoming product that addresses
>> my pontification nicely:
>>
>> http://www.telepostinc.com/LP-500.html
>>
>> Cost will be more than a used oscilloscope, but it looks like his product is
>> concentrated on what we need for monitoring rather than general bench work.
>> Design and cosmetics appear to be commensurate with the Elecraft K3.
>>
>> Paul, W9AC
>>
> ______________________________________________________________
> 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

Guy, K2AV

Re: On the K3's CW keying envelope shaping, rise/fall times, and CW ALC

In reply to this post by alorona

It ain't you. They squeeze in to a QRO signal on a 40m 5 element
wide-spaced quad on a 220 foot catenary at NY4A. It has to be the
bandwidth. When I used an MP out there, people above and below used
to complain.

On Fri, Dec 3, 2010 at 12:51 PM, Al Lorona <[hidden email]> wrote:

> Several years ago when the K2 modification was created by Wayne to implement a
> raised cosine keying filter I installed it and both ran simulations and made
> measurements of the resulting waveform, which agreed very closely. In fact, I
> liked the result so much that I duplicated the circuit and modified my Omni 6
> with it. I just sold that rig and feel good that the next owner will put a very
> clean-keying rig out there in the future.
>
> In private conversations with several good contesters over the years I have the
> general impression that although none of them would intentionally modify their
> rigs to splatter (neither on CW nor SSB), they regard a clicking/splattering
> transmitter as creating the happy side effect of a virtual guard band on either
> side of them in a contest.
>
> I also notice that whenever I am running with the K2 or K3 other contesters tend
> to crowd me just a tad closer than I'm used to on one side or the other. It
> annoys me. An unhappy side effect of exceptionally clean keying? Naw, more
> likely just my weak, little-pistol signal!
>
> Al W6LX
> ______________________________________________________________
> 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

Re: Fwd: New Sherwood report

In reply to this post by Jan Erik Holm

> Waveform? When you have a 1 ms raise/fall time no way you
> can shape the waveform so it doesnt click.

That's not the case. If the first and second derivatives of
the power output are minimized, the waveform does not click.
There are many studies about bandwidth vs. rate of change in
digital modulation that will provide the information to show
that principle.

In the specific case of CW as long as the transitions from
the "rise" to the steady "on" state and from the steady "on"
state to the fall are properly shaped (first/second derivative
minimized) the apparent clicks will be minimized.

The traditional "feedback" ALC system almost guarantees clicks
on the *trailing edge* of the CW element. Traditional ALC
attempts to maintain the output as the key is opened and will
increase system gain as the driving waveform starts to decay.
The output level will only begin to fall when the ALC gain
has reached maximum - at which point there will be a major discontinuity
in the waveform. A properly designed feedback
ALC system would be incorporate sample and hold to maintain
steady system gain during the "ramp up" and "ramp down" parts
of the CW waveform to prevent the "corner sharpening."

Again, rise time controls the basic CW bandwidth but "clicks"
are a feature of discontinuities in the CW envelope. Please
do not confuse the two.

73,

... Joe, W4TV

On 12/3/2010 9:46 AM, Jan Erik Holm wrote:

> Waveform? When you have a 1 ms raise/fall time no way you
> can shape the waveform so it doesnt click.
>
> /Jim SM2EKM
> ---------------
> On 2010-12-03 15:29, Guy Olinger K2AV wrote:
>> All the click complaints for years and still Yaesu doesn't take care
>> of the waveform. It's not like there aren't tons of reads on the
>> subject, no excuse at all since Google. It's more like they just
>> don't care, or are simply convinced it doesn't matter. After all,
>> isn't CW obsolete, and nobody uses it anymore, right?
>>
>> 73, Guy.
>>
>> On Fri, Dec 3, 2010 at 9:08 AM, Paul Christensen<[hidden email]> wrote:
>>> 4 ms. Given that the FTdx5K rise/fall is roughly double the time of the
>>> K3 -- and that the K3 at 2 ms consumes significantly less bandwidth than the
>>> FTdx5K at 4 ms, the CW transmit quality of the K3 is far superior. I would
>>> like to hear QSK on the Yaesu, compare it with the K3's QRQ mode, and note
>>> any audible artifacts in the headphones -- especially for signs of clicks
>>> and DC thumps.
>>>
>>> Referring to the December, 2010 QST Product Review on p.45, take a look at
>>> the leading edge of the second pulse (lower trace) in Figure 1. The sharp
>>> rise and sharp edge at the top of the waveform is what is consuming the
>>> bandwidth. Based on that waveform, it was absolutely generating clicks at
>>> some distance from the Fc. That pattern is characteristic of a deficient
>>> ALC system.
>>>
>>> A deficiency in the ALC system then takes us into SSB Tx IMD. The published
>>> FTdx5K Tx IMD numbers look great with the ARRL's steady tone method but
>>> arguably, that's probably more relevant to data modes. The question is
>>> "what is the FTdx5K's ALC doing to undermine the rig's otherwise excellent
>>> SSB Tx IMD numbers, especially in Class A?" Really, the published Tx IMD
>>> figures are meaningless unless dynamic testing is conducted to simulate
>>> rapid changes in power associated with voice modes.
>>>
>>> Paul, W9AC
>>>
>>> ----- Original Message -----
>>> From: "Jan Erik Holm"<[hidden email]>
>>> To:<[hidden email]>
>>> Sent: Friday, December 03, 2010 8:27 AM
>>> Subject: Re: [Elecraft] Fwd: New Sherwood report
>>>
>>>
>>>> Yes but to what rise/fall time was the radio set? AFAIK in
>>>> the FT5000 it can be changed.
>>>>
>>>> /Jim SM2EKM

Jan Erik Holm

Re: Fwd: New Sherwood report

In reply to this post by P.B. Christensen

No I do not experience key clicks with my own K3. I don´t know
what it sounds like and it´s no idea to ask anyone either.

It is all the other K3´s that are on the air that has "mild keyclicks"

Yes there are hot switching amps and QSK amps that aren´t
correct, I usually can detect those.

No it´s plain and simple the K3 would need to "soften up" the
keying slightly.

The only modern radio I have ever herd on the air that was click
free was a TenTec ORION that was set up right, I think he had it
at 7 ms, keying was just beautiful and not to soft. So my friends
it can be done!!

/ Jim SM2EKM
------------------
On 2010-12-03 16:55, Paul Christensen wrote:

>> Why, yes IMO the K3 clicks too much.
>
> Jim,
>
> If you're experiencing clicks with your K3, you really need to investigate
> with a scope. At the risk of pontificating, every station should have a
> scope to monitor the transmit waveform. It can also be used to effectively
> monitor T/R sequencing times when using an external amp. I suspect many
> stations are hot-switching their amps and don't even know it until a T/R
> relay's contacts fuse and the relay fails. Cost is no longer an excuse.
> Anyone who can afford a K3 or FTdx5K can own a scope. Good quality used
> scopes can be purchased on the surplus market between USD $100-200.
>
> If I may make a plug for N8LP, he's got a forthcoming product that addresses
> my pontification nicely:
>
> http://www.telepostinc.com/LP-500.html
>
> Cost will be more than a used oscilloscope, but it looks like his product is
> concentrated on what we need for monitoring rather than general bench work.
> Design and cosmetics appear to be commensurate with the Elecraft K3.
>
> Paul, W9AC
>

Jan Erik Holm

Re: Fwd: New Sherwood report

In reply to this post by Jan Erik Holm

No but if you read what I wrote I didn´t say that either.
BUT rise/fall time is the ONLY thing we these days can
change on our radios (except the K3)

Also most people does not know how to listen for key
clicks. I know a lot of people that doesn´t even find
key clicks on a FT-1000MP, yeah right....go figure.

/ Jim SM2EKM
------------------
On 2010-12-03 17:47, Tommy Alderman wrote:

> I think you are incorrect! I used IC-781's, Corsair II, and Corsair 6+ for
> many year at CW speeds over 70 wpm, constantly having my QSO partners look
> for key clicks and they were not present! Rise time alone, does not cause
> key clicks!
>
> And just for the record Elecraft, I do NOT want you to change this CW rise
> time on the K3!
>
> Tom - W4BQF
>
>
> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf Of Jan Erik Holm
> Sent: Friday, December 03, 2010 9:47 AM
> To: [hidden email]
> Subject: Re: [Elecraft] Fwd: New Sherwood report
>
> Waveform? When you have a 1 ms raise/fall time no way you
> can shape the waveform so it doesnt click.
>
> /Jim SM2EKM
> [||]
>
>

Jim Brown-10

Re: On the K3's CW keying envelope shaping, rise/fall times, and CW ALC

In reply to this post by wayne burdick

On 12/3/2010 8:56 AM, Wayne Burdick wrote:
> K3s
> are at or near the top in every CW contest already, thanks to the
> skill of our customers rather than bandwidth-hogging signals.

Well said, Wayne.

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

Jan Erik Holm

Re: Fwd: New Sherwood report

In reply to this post by Joe Subich, W4TV-4

Yes 99% of all CW stations clicks at the "break" side.

I didn´t think about the ALC but that is right, thanks.

However by increasing the envelope fall time we will
"help" a bad ALC system if you understand what I mean.

Anyway I better can this before HHQ puts the lid on.

/SM2EKM QRT
------------------
On 2010-12-03 19:18, Joe Subich, W4TV wrote:

>
> > Waveform? When you have a 1 ms raise/fall time no way you
> > can shape the waveform so it doesnt click.
>
> That's not the case. If the first and second derivatives of
> the power output are minimized, the waveform does not click.
> There are many studies about bandwidth vs. rate of change in
> digital modulation that will provide the information to show
> that principle.
>
> In the specific case of CW as long as the transitions from
> the "rise" to the steady "on" state and from the steady "on"
> state to the fall are properly shaped (first/second derivative
> minimized) the apparent clicks will be minimized.
>
> The traditional "feedback" ALC system almost guarantees clicks
> on the *trailing edge* of the CW element. Traditional ALC
> attempts to maintain the output as the key is opened and will
> increase system gain as the driving waveform starts to decay.
> The output level will only begin to fall when the ALC gain
> has reached maximum - at which point there will be a major discontinuity
> in the waveform. A properly designed feedback
> ALC system would be incorporate sample and hold to maintain
> steady system gain during the "ramp up" and "ramp down" parts
> of the CW waveform to prevent the "corner sharpening."
>
> Again, rise time controls the basic CW bandwidth but "clicks"
> are a feature of discontinuities in the CW envelope. Please
> do not confuse the two.
>
> 73,
>
> ... Joe, W4TV
>
> On 12/3/2010 9:46 AM, Jan Erik Holm wrote:
>> Waveform? When you have a 1 ms raise/fall time no way you
>> can shape the waveform so it doesnt click.
>>
>> /Jim SM2EKM
>> ---------------
>> On 2010-12-03 15:29, Guy Olinger K2AV wrote:
>>> All the click complaints for years and still Yaesu doesn't take care
>>> of the waveform. It's not like there aren't tons of reads on the
>>> subject, no excuse at all since Google. It's more like they just
>>> don't care, or are simply convinced it doesn't matter. After all,
>>> isn't CW obsolete, and nobody uses it anymore, right?
>>>
>>> 73, Guy.
>>>
>>> On Fri, Dec 3, 2010 at 9:08 AM, Paul Christensen<[hidden email]> wrote:
>>>> 4 ms. Given that the FTdx5K rise/fall is roughly double the time of the
>>>> K3 -- and that the K3 at 2 ms consumes significantly less bandwidth
>>>> than the
>>>> FTdx5K at 4 ms, the CW transmit quality of the K3 is far superior. I
>>>> would
>>>> like to hear QSK on the Yaesu, compare it with the K3's QRQ mode,
>>>> and note
>>>> any audible artifacts in the headphones -- especially for signs of
>>>> clicks
>>>> and DC thumps.
>>>>
>>>> Referring to the December, 2010 QST Product Review on p.45, take a
>>>> look at
>>>> the leading edge of the second pulse (lower trace) in Figure 1. The
>>>> sharp
>>>> rise and sharp edge at the top of the waveform is what is consuming the
>>>> bandwidth. Based on that waveform, it was absolutely generating
>>>> clicks at
>>>> some distance from the Fc. That pattern is characteristic of a
>>>> deficient
>>>> ALC system.
>>>>
>>>> A deficiency in the ALC system then takes us into SSB Tx IMD. The
>>>> published
>>>> FTdx5K Tx IMD numbers look great with the ARRL's steady tone method but
>>>> arguably, that's probably more relevant to data modes. The question is
>>>> "what is the FTdx5K's ALC doing to undermine the rig's otherwise
>>>> excellent
>>>> SSB Tx IMD numbers, especially in Class A?" Really, the published Tx
>>>> IMD
>>>> figures are meaningless unless dynamic testing is conducted to simulate
>>>> rapid changes in power associated with voice modes.
>>>>
>>>> Paul, W9AC
>>>>
>>>> ----- Original Message -----
>>>> From: "Jan Erik Holm"<[hidden email]>
>>>> To:<[hidden email]>
>>>> Sent: Friday, December 03, 2010 8:27 AM
>>>> Subject: Re: [Elecraft] Fwd: New Sherwood report
>>>>
>>>>
>>>>> Yes but to what rise/fall time was the radio set? AFAIK in
>>>>> the FT5000 it can be changed.
>>>>>
>>>>> /Jim SM2EKM
>
>

Jan Erik Holm

Re: On the K3's CW keying envelope shaping, rise/fall times, and CW ALC

In reply to this post by wayne burdick

Good my 5 ms measurement was the closest so far.

Thank´s for the explanation on the construction,
appreciate it.

After all, why should I bother, if my K3 clicks it
doesn´t disturb me.

/Jim SM2EKM
----------------
On 2010-12-03 17:56, Wayne Burdick wrote:

> There will never be a K3 menu option that allows for clicky CW. K3s
> are at or near the top in every CW contest already, thanks to the
> skill of our customers rather than bandwidth-hogging signals.
>
> There are two factors that control how clean and click-free a CW
> keying envelope will be: explicit shaping by DSP or analog circuitry,
> and incidental shaping due to ALC or transmit chain effects. The K3's
> CW is extremely clean and narrow-banded and click-free because we took
> both factors into consideration. (Note the this is also true of the K2.)
>
> DSP Shaping:
>
> The rising and falling edges of the CW waveform are shaped by the DSP
> using an optimal raised-cosine envelope. Rise and fall times are
> approximately 4 ms, varying only slightly over the entire available
> power output range of 0.1 to 110 W. We experimented with other
> sigmoidal envelope shapes, but the raised cosine was the best overall.
>
> ALC:
>
> The K3 one of very few high-end transceivers that use open-loop
> application of keying envelope shaping. This ensures that the applied
> shape is not compromised by ALC action. Most transceivers, including
> some recently discussed, use fast ALC to control power output level.
> Even if they start with DSP or analog shaping, the ALC jumps in to
> limit the peak amplitude of the transmit waveform, resulting in
> envelope clipping and thus wideband clicks.
>
> So how does the K3's CW ALC work? First, we use a TX gain calibration
> procedure to store per-band gain constants. This information is used
> to preset transmit gain as you rotate the POWER control. When you hit
> the key, we start off just below this target level (about 0.5 to 1
> dB), then use a slow ALC loop to adjust gain to hit the exact level
> requested. Generally the power stabilizes in one or two dits. Shaping
> is excellent on every code element.
>
> We use similar techniques to ensure virtually perfect envelope shaping
> in data modes. In voice modes, we use two-stage ALC -- fast pre-
> crystal-filter ALC in the DSP, slow firmware-based ALC *after* the
> crystal filter -- to ensure that speech signals are completely free of
> splatter. This technique also results in extremely clean and effective
> speech compression.
>
> 73,
> Wayne
> N6KR
>

Joe Subich, W4TV-4

Re: Fwd: New Sherwood report

In reply to this post by Jan Erik Holm

> However by increasing the envelope fall time we will
> "help" a bad ALC system if you understand what I mean.

Changing (increasing) the envelope fall time will not
eliminate AGC generated clicks on break ... it will only
delay their timing be a small fraction of a millisecond.
ALC will continue to hold the steady state "on" level
until it runs out of gain and then the sudden change
from a constant level to a N dB/ms slope will still be
a discontinuity in the envelope with the resulting click.

The change in ALC generated clicks with changing envelope
decay is minimal.

73,

... Joe, W4TV

On 12/3/2010 2:06 PM, Jan Erik Holm wrote:

> Yes 99% of all CW stations clicks at the "break" side.
>
> I didn´t think about the ALC but that is right, thanks.
>
> However by increasing the envelope fall time we will
> "help" a bad ALC system if you understand what I mean.
>
> Anyway I better can this before HHQ puts the lid on.
>
> /SM2EKM QRT
> ------------------
> On 2010-12-03 19:18, Joe Subich, W4TV wrote:
>>
>> > Waveform? When you have a 1 ms raise/fall time no way you
>> > can shape the waveform so it doesnt click.
>>
>> That's not the case. If the first and second derivatives of
>> the power output are minimized, the waveform does not click.
>> There are many studies about bandwidth vs. rate of change in
>> digital modulation that will provide the information to show
>> that principle.
>>
>> In the specific case of CW as long as the transitions from
>> the "rise" to the steady "on" state and from the steady "on"
>> state to the fall are properly shaped (first/second derivative
>> minimized) the apparent clicks will be minimized.
>>
>> The traditional "feedback" ALC system almost guarantees clicks
>> on the *trailing edge* of the CW element. Traditional ALC
>> attempts to maintain the output as the key is opened and will
>> increase system gain as the driving waveform starts to decay.
>> The output level will only begin to fall when the ALC gain
>> has reached maximum - at which point there will be a major discontinuity
>> in the waveform. A properly designed feedback
>> ALC system would be incorporate sample and hold to maintain
>> steady system gain during the "ramp up" and "ramp down" parts
>> of the CW waveform to prevent the "corner sharpening."
>>
>> Again, rise time controls the basic CW bandwidth but "clicks"
>> are a feature of discontinuities in the CW envelope. Please
>> do not confuse the two.
>>
>> 73,
>>
>> ... Joe, W4TV
>>
>> On 12/3/2010 9:46 AM, Jan Erik Holm wrote:
>>> Waveform? When you have a 1 ms raise/fall time no way you
>>> can shape the waveform so it doesnt click.
>>>
>>> /Jim SM2EKM
>>> ---------------
>>> On 2010-12-03 15:29, Guy Olinger K2AV wrote:
>>>> All the click complaints for years and still Yaesu doesn't take care
>>>> of the waveform. It's not like there aren't tons of reads on the
>>>> subject, no excuse at all since Google. It's more like they just
>>>> don't care, or are simply convinced it doesn't matter. After all,
>>>> isn't CW obsolete, and nobody uses it anymore, right?
>>>>
>>>> 73, Guy.
>>>>
>>>> On Fri, Dec 3, 2010 at 9:08 AM, Paul Christensen<[hidden email]> wrote:
>>>>> 4 ms. Given that the FTdx5K rise/fall is roughly double the time of the
>>>>> K3 -- and that the K3 at 2 ms consumes significantly less bandwidth
>>>>> than the
>>>>> FTdx5K at 4 ms, the CW transmit quality of the K3 is far superior. I
>>>>> would
>>>>> like to hear QSK on the Yaesu, compare it with the K3's QRQ mode,
>>>>> and note
>>>>> any audible artifacts in the headphones -- especially for signs of
>>>>> clicks
>>>>> and DC thumps.
>>>>>
>>>>> Referring to the December, 2010 QST Product Review on p.45, take a
>>>>> look at
>>>>> the leading edge of the second pulse (lower trace) in Figure 1. The
>>>>> sharp
>>>>> rise and sharp edge at the top of the waveform is what is consuming the
>>>>> bandwidth. Based on that waveform, it was absolutely generating
>>>>> clicks at
>>>>> some distance from the Fc. That pattern is characteristic of a
>>>>> deficient
>>>>> ALC system.
>>>>>
>>>>> A deficiency in the ALC system then takes us into SSB Tx IMD. The
>>>>> published
>>>>> FTdx5K Tx IMD numbers look great with the ARRL's steady tone method but
>>>>> arguably, that's probably more relevant to data modes. The question is
>>>>> "what is the FTdx5K's ALC doing to undermine the rig's otherwise
>>>>> excellent
>>>>> SSB Tx IMD numbers, especially in Class A?" Really, the published Tx
>>>>> IMD
>>>>> figures are meaningless unless dynamic testing is conducted to simulate
>>>>> rapid changes in power associated with voice modes.
>>>>>
>>>>> Paul, W9AC
>>>>>
>>>>> ----- Original Message -----
>>>>> From: "Jan Erik Holm"<[hidden email]>
>>>>> To:<[hidden email]>
>>>>> Sent: Friday, December 03, 2010 8:27 AM
>>>>> Subject: Re: [Elecraft] Fwd: New Sherwood report
>>>>>
>>>>>
>>>>>> Yes but to what rise/fall time was the radio set? AFAIK in
>>>>>> the FT5000 it can be changed.
>>>>>>
>>>>>> /Jim SM2EKM
>>
>>
> ______________________________________________________________
> 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
>

Kok Chen

Re: Fwd: New Sherwood report

In reply to this post by Jan Erik Holm

On Dec 3, 2010, at 11:06 AM, Jan Erik Holm wrote:

> Yes 99% of all CW stations clicks at the "break" side.

That makes perfect sense, Jan.

If they are using simple IIR filters, the slope discontinuity is worse at the onset of switching than when it is at the end of the leading edge or trailing edge (visualize the RC constant diagrams that we learn in school :-).

Now, the onset of switching states happens to occur at low power (zero power) on the rising edge of a CW pulse, so even if it is dirty, you are not putting out energy. But it the worst part of an RC filter occurs at the highest power location at the trailing edge of a CW pulse.

If the click spectra is mostly coming from slope discontinuities, the energy from the clicks are going to be much greater when the key is breaking than when the key is making.

This is why a couple of us has said that it is not the "rise time" (or fall time) that is important, it is the n-th order discontinuities (slope discontinuity contributes more than higher order ones obviously, when you look at it as a Fourier series).

Someone else can probably explain better than I can. I can't do it without using equations. English is my third language :-).

73
Chen, W7AY

______________________________________________________________
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

Jan Erik Holm

Re: Fwd: New Sherwood report

In reply to this post by Joe Subich, W4TV-4

I see. Was thinking along terms if change was slow
it would make it easier for an AGC with bad dynamic range,
i e a slow AGC loop would create less IMD so to speak,
well slow at least not fast enough.

Interesting but I´m not expert enough, have to think about
it more.

/ Jim SM2EKM
--------------
On 2010-12-03 20:31, Joe Subich, W4TV wrote:

>
> > However by increasing the envelope fall time we will
> > "help" a bad ALC system if you understand what I mean.
>
> Changing (increasing) the envelope fall time will not
> eliminate AGC generated clicks on break ... it will only
> delay their timing be a small fraction of a millisecond.
> ALC will continue to hold the steady state "on" level
> until it runs out of gain and then the sudden change
> from a constant level to a N dB/ms slope will still be
> a discontinuity in the envelope with the resulting click.
>
> The change in ALC generated clicks with changing envelope
> decay is minimal.
>
> 73,
>
> ... Joe, W4TV
>
> On 12/3/2010 2:06 PM, Jan Erik Holm wrote:
>> Yes 99% of all CW stations clicks at the "break" side.
>>
>> I didn´t think about the ALC but that is right, thanks.
>>
>> However by increasing the envelope fall time we will
>> "help" a bad ALC system if you understand what I mean.
>>
>> Anyway I better can this before HHQ puts the lid on.
>>
>> /SM2EKM QRT
>> ------------------
>> On 2010-12-03 19:18, Joe Subich, W4TV wrote:
>>>
>>> > Waveform? When you have a 1 ms raise/fall time no way you
>>> > can shape the waveform so it doesnt click.
>>>
>>> That's not the case. If the first and second derivatives of
>>> the power output are minimized, the waveform does not click.
>>> There are many studies about bandwidth vs. rate of change in
>>> digital modulation that will provide the information to show
>>> that principle.
>>>
>>> In the specific case of CW as long as the transitions from
>>> the "rise" to the steady "on" state and from the steady "on"
>>> state to the fall are properly shaped (first/second derivative
>>> minimized) the apparent clicks will be minimized.
>>>
>>> The traditional "feedback" ALC system almost guarantees clicks
>>> on the *trailing edge* of the CW element. Traditional ALC
>>> attempts to maintain the output as the key is opened and will
>>> increase system gain as the driving waveform starts to decay.
>>> The output level will only begin to fall when the ALC gain
>>> has reached maximum - at which point there will be a major discontinuity
>>> in the waveform. A properly designed feedback
>>> ALC system would be incorporate sample and hold to maintain
>>> steady system gain during the "ramp up" and "ramp down" parts
>>> of the CW waveform to prevent the "corner sharpening."
>>>
>>> Again, rise time controls the basic CW bandwidth but "clicks"
>>> are a feature of discontinuities in the CW envelope. Please
>>> do not confuse the two.
>>>
>>> 73,
>>>
>>> ... Joe, W4TV
>>>
>>> On 12/3/2010 9:46 AM, Jan Erik Holm wrote:
>>>> Waveform? When you have a 1 ms raise/fall time no way you
>>>> can shape the waveform so it doesnt click.
>>>>
>>>> /Jim SM2EKM
>>>> ---------------
>>>> On 2010-12-03 15:29, Guy Olinger K2AV wrote:
>>>>> All the click complaints for years and still Yaesu doesn't take care
>>>>> of the waveform. It's not like there aren't tons of reads on the
>>>>> subject, no excuse at all since Google. It's more like they just
>>>>> don't care, or are simply convinced it doesn't matter. After all,
>>>>> isn't CW obsolete, and nobody uses it anymore, right?
>>>>>
>>>>> 73, Guy.
>>>>>
>>>>> On Fri, Dec 3, 2010 at 9:08 AM, Paul Christensen<[hidden email]> wrote:
>>>>>> 4 ms. Given that the FTdx5K rise/fall is roughly double the time
>>>>>> of the
>>>>>> K3 -- and that the K3 at 2 ms consumes significantly less bandwidth
>>>>>> than the
>>>>>> FTdx5K at 4 ms, the CW transmit quality of the K3 is far superior. I
>>>>>> would
>>>>>> like to hear QSK on the Yaesu, compare it with the K3's QRQ mode,
>>>>>> and note
>>>>>> any audible artifacts in the headphones -- especially for signs of
>>>>>> clicks
>>>>>> and DC thumps.
>>>>>>
>>>>>> Referring to the December, 2010 QST Product Review on p.45, take a
>>>>>> look at
>>>>>> the leading edge of the second pulse (lower trace) in Figure 1. The
>>>>>> sharp
>>>>>> rise and sharp edge at the top of the waveform is what is
>>>>>> consuming the
>>>>>> bandwidth. Based on that waveform, it was absolutely generating
>>>>>> clicks at
>>>>>> some distance from the Fc. That pattern is characteristic of a
>>>>>> deficient
>>>>>> ALC system.
>>>>>>
>>>>>> A deficiency in the ALC system then takes us into SSB Tx IMD. The
>>>>>> published
>>>>>> FTdx5K Tx IMD numbers look great with the ARRL's steady tone
>>>>>> method but
>>>>>> arguably, that's probably more relevant to data modes. The
>>>>>> question is
>>>>>> "what is the FTdx5K's ALC doing to undermine the rig's otherwise
>>>>>> excellent
>>>>>> SSB Tx IMD numbers, especially in Class A?" Really, the published Tx
>>>>>> IMD
>>>>>> figures are meaningless unless dynamic testing is conducted to
>>>>>> simulate
>>>>>> rapid changes in power associated with voice modes.
>>>>>>
>>>>>> Paul, W9AC
>>>>>>
>>>>>> ----- Original Message -----
>>>>>> From: "Jan Erik Holm"<[hidden email]>
>>>>>> To:<[hidden email]>
>>>>>> Sent: Friday, December 03, 2010 8:27 AM
>>>>>> Subject: Re: [Elecraft] Fwd: New Sherwood report
>>>>>>
>>>>>>
>>>>>>> Yes but to what rise/fall time was the radio set? AFAIK in
>>>>>>> the FT5000 it can be changed.
>>>>>>>
>>>>>>> /Jim SM2EKM
>>>

Jan Erik Holm

Re: Fwd: New Sherwood report

In reply to this post by Kok Chen

Yes logical at least to me. Well English is my second language
but still makes it difficult.

Yes but if the fall time is longer the discontinuity will be
less abrupt and by so make it easier for a bad regulating device,
if you understand what I mean.
We want to shape the first part of the envelope on the fall side,
not being to sharp of a knee, i e the first 1 ms from full power
is the most important part. If you increase the over all fall time
that transition period will also be longer, i e make it easier for
a system with not enough dynamic range.

Oh well once upon a moon I could do Fourier analysis on this but it
was 35 years ago and I just have forgot it all. The teacher I had
was a ham but he is an SK now.

/ Jim SM2EKM
--------------
On 2010-12-03 20:36, Kok Chen wrote:

>
> On Dec 3, 2010, at 11:06 AM, Jan Erik Holm wrote:
>
>> Yes 99% of all CW stations clicks at the "break" side.
>
> That makes perfect sense, Jan.
>
> If they are using simple IIR filters, the slope discontinuity is worse at the onset of switching than when it is at the end of the leading edge or trailing edge (visualize the RC constant diagrams that we learn in school :-).
>
> Now, the onset of switching states happens to occur at low power (zero power) on the rising edge of a CW pulse, so even if it is dirty, you are not putting out energy. But it the worst part of an RC filter occurs at the highest power location at the trailing edge of a CW pulse.
>
> If the click spectra is mostly coming from slope discontinuities, the energy from the clicks are going to be much greater when the key is breaking than when the key is making.
>
> This is why a couple of us has said that it is not the "rise time" (or fall time) that is important, it is the n-th order discontinuities (slope discontinuity contributes more than higher order ones obviously, when you look at it as a Fourier series).
>
> Someone else can probably explain better than I can. I can't do it without using equations. English is my third language :-).
>
> 73
> Chen, W7AY
>
>

Corboy - Poteet

[K3] Key Clicks

In reply to this post by Kok Chen

OK, this has been bashed about pretty good. So my question is: if someone puts a spectrum analyzer on the CW output of a K3, are key clicks detectable? On make, on break? Close in, far out, in between? If they exist, what is the level of the click relative to the signal? If they exist, are the clicks really a problem? I've not heard any K3 with detectable clicks (but I haven't been trying to find them either).

Mike W5FTD

>
> On Dec 3, 2010, at 11:06 AM, Jan Erik Holm wrote:
>
>> Yes 99% of all CW stations clicks at the "break" side.
>>
> That makes perfect sense, Jan.
>
> If they are using simple IIR filters, the slope discontinuity is
> worse at the onset of switching than when it is at the end of the
> leading edge or trailing edge (visualize the RC constant diagrams
> that we learn in school :-).
>
> Now, the onset of switching states happens to occur at low power
> (zero power) on the rising edge of a CW pulse, so even if it is
> dirty, you are not putting out energy. But it the worst part of an
> RC filter occurs at the highest power location at the trailing edge
> of a CW pulse.
>
> If the click spectra is mostly coming from slope discontinuities,
> the energy from the clicks are going to be much greater when the
> key is breaking than when the key is making.
>
> This is why a couple of us has said that it is not the "rise time"
> (or fall time) that is important, it is the n-th order
> discontinuities (slope discontinuity contributes more than higher
> order ones obviously, when you look at it as a Fourier series).
>
> Someone else can probably explain better than I can. I can't do it
> without using equations. English is my third language :-).
>
> 73
> Chen, W7AY

P.B. Christensen

Re: Fwd: New Sherwood report

In reply to this post by Jan Erik Holm

> Yes 99% of all CW stations clicks at the "break" side.

That's an unusually high percentage from what I recall seeing in the QST
Product Reviews during the past twenty years. So, I looked at the CW
waveforms of the more notorious transceivers. It seems to be an even mix
between leading edge and trailing edge issues.

In looking at the Ten Tec Omni Six Plus, it's the leading edge with the
sharp slope and discontinuity that creates clicks. I drafted a fix for that
back in 1998 and it appears on the N1EU website.

The ALC problem with the FT-1000MP series is more problematic on the
trailing edge, although from the 1996 QST review, the initial keyed element
has a severe problem at the leading edge, then followed by a problem at the
trailing edge with successive keying. That occurs as the ALC is trying to
stabilize.

On several other transceivers, I noticed significant dit shortening and
discontinuities on the initial keyed element, then the rest in the series
look fine, again as the ALC stabilizes. The FT-1000D shows a significant
problem on the leading edge, with dit shortening occurring on the first dit.
Dit shortening is not directly attributed to ALC. The root cause of that is
unrelated. The K3 has a slight bit of dit shortening, and about 3 msec of
compensation from an external keyer helps to restore the envelope to the key
closure time, although deciding exactly how the contact closure and RF pulse
time should match can result in a spirited debate, owing to the required
rise/fall time.

I would tend to throw out problems that only exist with the initial keyed
element so long as the remaining series looks fine. These transceivers
consume a lot of bandwidth very briefly, then the bandwidth dissipates. My
TS-480 comes to mind as it has a sharp leading edge with a slight power
spike on the leading edge of the initial element then disappears with
continued sending until there's a long pause. So, based on what I am
seeing, the ALC problem is pretty well evenly distributed between the
leading and trailing edge issues.

Paul, W9AC

______________________________________________________________
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

Doug Alspaugh

Re: [K3] Key Clicks

In reply to this post by Corboy - Poteet

If you’re a ARRL member you can see a scope output in the K3 review on line.
They called it out standing.

73 Doug N3QW

-----Original Message-----
From: Corboy - Poteet
Sent: Friday, December 03, 2010 3:19 PM
To: Elecraft reflector
Subject: [Elecraft] [K3] Key Clicks

OK, this has been bashed about pretty good. So my question is: if someone
puts a spectrum analyzer on the CW output of a K3, are key clicks
detectable? On make, on break? Close in, far out, in between? If they
exist, what is the level of the click relative to the signal? If they
exist, are the clicks really a problem? I've not heard any K3 with
detectable clicks (but I haven't been trying to find them either).

Mike W5FTD

OK, this has been bashed
______________________________________________________________
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

Kok Chen

Re: [K3] Key Clicks

In reply to this post by Corboy - Poteet

On Dec 3, 2010, at 12:19 PM, Corboy - Poteet wrote:

> So my question is: if someone puts a spectrum analyzer on the CW output of a K3, are key clicks detectable? On make, on break? Close in, far out, in between?

If you have one of these new fangled digital 'scope that takes a USB memory stick, you can record the time sequence and do your own FFT to get a good approximation of the spectrum. There is no need for a spectrum analyzer.

Take the data from one complete pulse, and prefill and post fill with zeros (so you can take a long FFT for good frequency resolution), and you get the envelope of the keyclick spectrum.

Take the leading edge half of the pulse only, but postfill the data with an identical high voltage of the on state of the pulse, and you get the contribution from only the "make" part of the keying waveform.

If you prefill the data with the same high voltage right up to where the key starts breaking, you get the contribution from the trailing edge.

FFT lengths and sampling periods will determine how much and how far away you can see.

For many applications such as this, you can even do without a scope. Just connect a sound card to a second receiver that is tapped off a directional couple in line from the test transmitter to a dummy load. But you will need to use a Softrock or LP-PAN if you want to see what the keyclicks look like 20 kHz away.

Or, just transmit at 8215 kHz and connect the LP-PAN to the directional coupler tap and you won't need a second receiver :-).

73
Chen, W7AY

______________________________________________________________
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

Guy, K2AV

Re: [K3] Key Clicks

In reply to this post by Doug Alspaugh

I have never heard a K3 barefoot with key clicks.

I have heard one where the owner was attempting to operate QSK with a
famous brand amp with slow frame relay T/R transfer and was hot
switching the amp all the way. The COMBO had key clicks. Really bad
key clicks. Probably frying stuff in the amp too. I'm sure you could
generate clicks with a K3 (or any rig) by overdriving a small amp and
using ALC to set the level. God would have key clicks if treated that
way.

K3's do not have key clicks. There certainly are some Yaesu's with
key clicks. They don't really seem to care that much about CW
bandwidth. If they did care they would be as narrow as a K3.

73, Guy.

On Fri, Dec 3, 2010 at 5:00 PM, Doug Alspaugh <[hidden email]> wrote:

> If you’re a ARRL member you can see a scope output in the K3 review on line.
> They called it out standing.
>
>
>
> 73 Doug N3QW
>
>
>
> -----Original Message-----
> From: Corboy - Poteet
> Sent: Friday, December 03, 2010 3:19 PM
> To: Elecraft reflector
> Subject: [Elecraft] [K3] Key Clicks
>
> OK, this has been bashed about pretty good. So my question is: if someone
> puts a spectrum analyzer on the CW output of a K3, are key clicks
> detectable? On make, on break? Close in, far out, in between? If they
> exist, what is the level of the click relative to the signal? If they
> exist, are the clicks really a problem? I've not heard any K3 with
> detectable clicks (but I haven't been trying to find them either).
>
> Mike W5FTD
>
>
>
>>
>> On Dec 3, 2010, at 11:06 AM, Jan Erik Holm wrote:
>>
>>> Yes 99% of all CW stations clicks at the "break" side.
>>>
>> That makes perfect sense, Jan.
>>
>> If they are using simple IIR filters, the slope discontinuity is
>> worse at the onset of switching than when it is at the end of the
>> leading edge or trailing edge (visualize the RC constant diagrams
>> that we learn in school :-).
>>
>> Now, the onset of switching states happens to occur at low power
>> (zero power) on the rising edge of a CW pulse, so even if it is
>> dirty, you are not putting out energy. But it the worst part of an
>> RC filter occurs at the highest power location at the trailing edge
>> of a CW pulse.
>>
>> If the click spectra is mostly coming from slope discontinuities,
>> the energy from the clicks are going to be much greater when the
>> key is breaking than when the key is making.
>>
>> This is why a couple of us has said that it is not the "rise time"
>> (or fall time) that is important, it is the n-th order
>> discontinuities (slope discontinuity contributes more than higher
>> order ones obviously, when you look at it as a Fourier series).
>>
>> Someone else can probably explain better than I can. I can't do it
>> without using equations. English is my third language :-).
>>
>> 73
>> Chen, W7AY
> OK, this has been bashed
> ______________________________________________________________
> 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

Vic K2VCO

Re: Fwd: New Sherwood report

In reply to this post by P.B. Christensen

By the way, if you have a P3 you can see the spikes on make, break, or both on the
waterfall. You can spot the rigs that are using semi-QSK that have a big spike on the
first dit and then quiet down, too.

On 12/3/2010 1:35 PM, Paul Christensen wrote:

> I would tend to throw out problems that only exist with the initial keyed
> element so long as the remaining series looks fine. These transceivers
> consume a lot of bandwidth very briefly, then the bandwidth dissipates. My
> TS-480 comes to mind as it has a sharp leading edge with a slight power
> spike on the leading edge of the initial element then disappears with
> continued sending until there's a long pause. So, based on what I am
> seeing, the ALC problem is pretty well evenly distributed between the
> leading and trailing edge issues.
>
> Paul, W9AC

--
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco/
______________________________________________________________
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

P.B. Christensen

Re: Fwd: New Sherwood report

> By the way, if you have a P3 you can see the spikes on make, break, or
> both on the
> waterfall. You can spot the rigs that are using semi-QSK that have a big
> spike on the
> first dit and then quiet down, too.

Same with relative SSB IMD monitoring. I've been looking at the area near
the SSB carrier set point since IMD is not confused with Tx bandwidth. Wow,
some huge differences between stations. And several ESSB ops with
emphasized low end are not necessarily a problem.

With a bit of practice you can clearly see the grunge below the set point
and even call a rough value. I've seen rough values vary from -20 dBc to
better than -50 dBc. On 20m last week, an ESSB op was using a TS-950SDX and
Alpha 89. Using the SDR-IQ, he had one of the cleanest spectrums I've seen.
By contrast, an op using an FT-1000D and ACOM amp showed about -30 dBc I
want to emphasize that these are not actual IMD numbers -- only a relative
IMD indication.

Paul, W9AC

______________________________________________________________
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

12345