CW-U and CW-L

Ken - NM9P

Yes! Thanks, Steve!

Paul Christensen, W9AC

Steve,

That results in extremely thin recovered audio with the wider CW filter settings -- and would undermine the goal.  I think you had it right the first time when you state: "-500 Hz to 1500 Hz" for a 500 Hz sidetone with a 2 kHz filter.  The opposite sideband limit should track the sidetone frequency on the opposite side of zero.  The reason for going below the zero point is that the full audio bandwidth of the CW waveform is heard and yet single-signal reception is maintained as long as there's sidetone tracking.  When the lower edge is set to zero, the low-end audio response is lost and recovered audio is very "thin."   

BTW - Rick Campbell, KK7B, discusses this effect in one of his R2 receiver articles.  Quite a bit of effort was spent in maximizing analog opposite rejection while providing a flat audio response down to below 100 Hz.  Let me know, and I'll dig it up if its of interest. 

The primary reason for the wider CW bandwidth is to judge the overall quality of the received CW waveform and at the same time, get a feel for what's happening in and around the center frequency.  For others, narrow CW filter settings can be fatiguing over time.

I realize that CW ops who only use narrow filters won't get this concept.   At the narrow settings of 500 Hz and lower, the filters should be left just as they are since single-signal reception is guaranteed.  Hope this helps.

Paul,W9AC

Ken - NM9P

I'm not sure I get it, and will need to read the article.  But if there are operators who like it this way, I would suggest that this could be a click-box option in the CW set up tab?  That way it could be customized to fit each op's preference.  (I am a multiple options kind of guy...)  <grin>

Steve N4LQ

Agree. The bandpass should go down to zero and maybe cross over to the other sideband just a bit when using wide filters. Of course if we just had the controls we could set it wherever we like. 

Steve-N5AC

I'm not certain we're talking about the same thing.  We have to decide whether we are talking about audio or RF.  When you put your pitch at 600Hz, we essentially shift the filter you have up by that about.  So a 200Hz filter is really 500-700Hz.  If your pitch is at 500Hz and we set the RF filter to -500 to 1500Hz, you will have sidebands from -500 to 500 mixing in the audio space.  In other words a tone at -200 sounds the same as a tone at 200.  The rule, based on what I THINK I said was that in CW-U the low edge of the filter cannot go more than <pitch>Hz below the carrier indicator in the display (this frequency in the demodulator is 0Hz).  Is this correct or am I still lost in space?

Paul Christensen, W9AC

The last thing I want is to come across like I'm asking Flex to design my own radio.  It's probably best to experiment with some of these choices.  As long as the skirts can be independently varied as they are now, that's what's most important.

W5XZ - dan

I think you got it, steve.  Some of us, who listen with wider filter widths on quiet, or 'dead' bands, prefer true 'single signal' reception, regardless of pitch setting or filter width. I don't want to hear signals on the other side of zero beat; or, have an irritatingly hi pitch setting, either.

thanks for taking another look. w5xz, dan
 

W5XZ - dan

now that I think about it a bit, for old timers, this is probably a carry over from old analog ssb rigs, with no cw filters, or fancy vbt / pbt, etc...

still, being able to park the rx on a known active CW freq with a wide rx passband, and just listen while you move around the shack still has some appeal...

73, w5xz, dan

Steve N4LQ

You got it right now Steve. The way it's set up now is really DSB Double sideband on CW. It should be SSB. I should never be able to hear the image of the other sideband. Even so, having passband tuning like PSDR was handy.

Steve N4LQ

I like to open the filter up as wide as possible and listen to as much of the cw band as my hearing range allows. The dynamic range on these rigs is amazing. My brain is able to pick out and lock on to one of multitudes of tones. The signals don't mush together as they do on conventional radios.

Paul Christensen, W9AC

Setting the lower limit to zero for wide CW bandwidths is better than it currently is.  For those of us who want to pull it across zero to match the side-tone frequency, it just takes a tug with the mouse. 

Paul, W9AC

Steve N4LQ

Paul
Yes but if you're using LSB or USB tugging the mouse won't drag the sideband past zero for some reason. Seems to me like you should be able to expand it either way or both. Isn't that how Psdr worked?

W5XZ - dan

me too! sort of a mental APF, i guess?

Steve N4LQ

An I use a tunable notch filter on the xyl

W5XZ - dan

ha! Cindy could verify I have one of those too!!

Bruce

N5FD  Bruce
 Thanks Paul for the last  paragraph "I realize that CW ops won't get this concept............should be left just as they at since single-signal reception is guaranteed"
And yes this helps.  Although The concept might be understood but not applicable to one who never uses a filter greater than 100 Hz. and has never connected his mike.
Thanks again Paul.
73,
Bruce N5FD

p.s. Love my Flex and it's filters best in my 55 years as a CW op in USN and amateur radio.

Charles - K5UA

Steve......I can imagine this is getting confusing. Your example of a wide filter with one edge below zero beat allowing a tone at -200 hz to sound the same as a tone at +200 hz is EXACTLY what we DO NOT want to happen. I (we) NEVER want to hear signals from the opposite side band. In a CW contest, that's another station's run frequency. I don't want to hear him or any stations calling him. I think you have it right if you apply the SSB analogy of never letting the edge of the filter go below zero beat (to the opposite sideband). If someone wants to listen to what is below zero beat, engage the RIT and go there. Or, let the op grab the lower edge of the filter skirt and drag it below zero beat. Keep the filter symmetrical on the CW offset until the filter width approaches twice the CW offset, then any additional filter width gets added to the high side of the filter, just like in SSB mode. Vice-versa for LSB CW mode. Good thread.

Paul Christensen, W9AC

Charles, that last full paragraph of yours is spot-on and should be underscored.  You did a much better job of explaining it than I did.   

Paul, W9AC

Steve N4LQ

This is a classic case of the company listening to their customers! 

Mike__K5JR

Steve, I'll throw my comments in here as well.  When using a "filter" that would exceed 1/2 the pitch frequency toward the zero beat side, I'd cap it at that and allow the non zero beat side to extend either the rest of the selected filter bandwidth or 1/2 the selected filter bandwidth.  In the former, the filter width away from zero beat would vary based on the pitch while in the latter, it is always equal to 1/2 the selected width, or "1500" for a 3.0K.  It's not really a 3.0k filter at that point, but a 1.5k plus 1/2 the pitch.  Yes, most contesters "running" will use a wider filter to not miss someone calling way off frequency, but you don't want to hear the station operating next to you that may be 1 KHz away from your frequency.

Mike / W5JR

Mike__K5JR

This also addresses the sound of the "filter" and removes the undesirable "bassy" sound from it.

And I've just played around more with pitches of 700, 500 and 400.  I like that a filter of 1.0 is 1.0.  That is 1/2 the pitch from spot towards zero beat and the remainder away.  So for the three pitches, that's 350 Hz lower than spot and 650 Hz above spot, 250 Hz lower than spot and 750 Hz above, and finally 200 Hz below spot and 800 Hz above.

Mike / W5JR

Paul Christensen, W9AC

Not sure how it was in PSDR since the 6700 is my first Flex.  In either USB or LSB, there's no need to set the filter below zero.  In CW the issue is affected by the offset and we must extend beyond zero.  The offset is a function of the selected side-tone frequency. 

K5UA's formula is the right one.  K5UA's Axiom: "Keep the filter symmetrical on the CW offset until the filter width approaches twice the CW offset, then any additional filter width gets added to the high side of the filter, just like in SSB mode. Vice-versa for LSB CW mode."

As an example of the need to go below the zero point for wide CW filter settings (wide is defined as more than 2x the CW offset),  let's select the 3 kHz CW filter in SSDR.  Presently the 3K CW is a symmetrical passband at -1500 Hz and +1500 Hz.  Now, as proposed by some, let's take the opposite sideband (determined by the new CW-L/CW-U menu feature) and bring it to zero.  For CW-U, keep the right side untouched at +1500 Hz, now bring the lower side from -1500 to exactly 0.  Now listen.   Pretty awful, isn't it.  There's no way anyone can listen to that audio passband.  If that becomes the default setting, it will generate a lot of complaints.  This is not the same as bringing the lower limit  to zero in USB or LSB.  Hate to mix audio into the discussion, but the audio passband is a function of the DSP passband. 

The right way:  For the 3 kHz CW filter and a 600 Hz offset chosen in SSDR, 2x the offset is a 1.2K filter.  Since half of the 3K filter is 1500 Hz, we have well exceeded the symmetry threshold.  For a 600 Hz offset, the symmetry threshold is a 1.2K filter (-600/+600).  Since 3K is more than 2x the offset (1.2K), then we keep the offset amount intact on the undesired sideband.  That's -600 Hz for CW-U.  Since 3K is more than 2x the offset, the additional bandwidth most go to the high side.  We end up with -600 Hz and +2400 Hz for a 3K filter. The bandwidth is 3K and single-signal reception is intact.  We get exactly what we want from the wider CW filter: (1) the desired wide bandwidth with full passband audio; and (2) single-signal reception.   

Probably the best way to understand this is to use my example above and move the filter skirts while listening to a CW QSO.  One experiment is worth a thousand words.

Paul, W9AC

Steve N4LQ

Last night I was listening to the CW portion of 40m with the filter at 3 khz. Heard a guy calling CQ right on frequency. Started to call him until I realized he was on the wrong side of zero-beat! That couldn't happen with any other receiver. This needs fixin. 

Charles - K5UA

Thanks Paul, two days ago I couldn't spell Axiom, now I ARE one!!! hi hi

Charles - K5UA

All levity aside, after all the manhours spent designing narrow analog crystal filters over the years, and after all the manhours recently spent refining skirts on digital filters in SDR radios, I have trouble envisioning anyone listening to CW in a filter wider than 500 hz. There could be quite a few stations dispersed in a 500 hz bandwidth on a busy frequency, not to mention a 1500 hz bandwidth. My mission to to isolate the desired CW signal for maximum signal to noise ratio amd maximum readibility. I don't understand the efficacy of using wide CW bandwidths, which will degrade signal to noise of the desired signal by introducing more band noise, or worse, add competiting signals in the bandwidth which will add to the processing load on your brain. To use a SSB analogy, you would not use a 6 khz filter in a crowded band to listen to a 3 khz station, would you? Then why would you use a wide filter in the CW analogy? I just don't see the practical application of wide CW filters in a panadapter capable radio where you can SEE adjacent CW signals without having to suffer the misery of hearing them.

Steve N4LQ

Charles
Close your eyes and listen to a 25 khz wide chunk of the CW spectrum and let the melodious calliope emanating from the ethereal heavens  permeate your soul as it wafts through your mind. Gradually your understand will come to light and you will come to despise the raucous racket of an unnecessarily narrow filter and learn to use such a device only when necessary.

Paul Christensen, W9AC

it's often relaxing to rag-chew using a wider CW bandwidth to get a feel for what's around you, then narrow-up as needed to fight QRM.  True, we can see them now too. 

I've had several QSOs where the calling station is outside of my narrow CW bandwidth and missed them only to be asked later why I was ignoring the calls.  

CW net control ops use wide CW bandwidths since folks are calling on everything from old crystal boatanchors to SDR transceivers.  Incoming frequencies can vary many kHz.

Paul, W9AC

Charles - K5UA

You are absolutely correct Steve, there is a harmonically displeasing penalty of using an unnecessarily narrow CW filter, the key word being unnecessarily.  And you are correct again, it should only be used when necessary. This is why I typically use a 400 hz filter as my standard CW filter on an uncrowded band, and 100 hz on a crowded band.  

My definition of torture is not water-boarding, it would be listening to CW on a 1500 hz filter under any band conditions, crowded or dead. 

For me, there is just too much high frequency hiss in the bandpass of a wide CW filter (i.e., greater than 500 hz).  It may not bother someone if they have a high frequency hearing loss (a biologic filter), but it drives me to distraction to hear all that high frequency hiss mixed in with what I am trying to hear at my CW offset of 400 hz.

Fortunately, once again.... isn't it great we have a   F..L..E..X..I..B..L..E    radio?

Ken - NM9P

On the other hand... As one who usually listens to CW with nothing less than 250 Hz filters on the 6500, there is a different, less tiresome effect when listening to a strong CW signal with AGC on MED or SLOW and the AGC-T turned down low so that there is NO noise other than the CW signal.  It is similar to listening to ESSB on the low end of 80 with strong signals and low RF gain.... Quiet and relaxing with no noise and high fidelity signals that stand out against the noise.  Yet in a crowded band with more noise, or with higher AGC-T the wide filter would bust my brain!

In any case, I agree that Flexibility is key, which we already have with the ability to grab and slide the filter.  I do agree with others that wider filters should default to settings that keep the low end above zero beat.  But for those who like to keep the filters symmetrical even below the zero beat, they should be able to drag them there. 

Clay N9IO

I just now found this month old thread. Had to comment that this has seriously been an interesting read. Paul your knowledge base has been blowing my mind since we were kids. 73' de Clay N9IO