CW-U and CW-L

Paul Christensen, W9AC

V1.3 now has the ability to select either CW-upper sideband or CW-lower sideband in the setup menu.  For CW fans, it helps to have an increase in pitch as the slice frequency increases.  I take the opposite sideband and set it to the CW pitch frequency.  When set this way with wider filter bandwidths, the opposite sideband stays suppressed for really excellent single-signal reception. 

Although we've been able to independently adjust the lower and upper filter skirts to null an interfering signal either above or below the operating frequency, now the pitch shift can move in the direction of the operating frequency.

At first, I thought placement would be better in the slice flag, but since the upper and lower skirts are manually set anyway, I see no detriment in leaving it in the setup menu.  I have not tried it yet, but hopefully CWU and CWL are captured in the new profiles.  Well done!

Paul, W9AC

Steve N4LQ

Paul: Yes the CWL is nice but I don't follow your comment: "I take the opposite sideband and set it to the CW pitch frequency".
Let's say my BW is set for 3.0k. To get single signal I have to slide the passband over to the right...Same as before. I'm missing something.

Paul Christensen, W9AC

Right.  No difference with the new feature except we can now control pitch direction,  If I'm set to CW-L and set BW to 3 kHz, I'm now pulling the *upper* sideband back to 600 Hz and leaving it there.  Then, the lower sideband is adjusted to the desired Rx bandwidth.  It was just the opposite when it was fixed on CW-U. 

Steve N4LQ

But of course it goes right back to default if you click a preset BW. I've wondered why they didn't set those for single signal...Llike my Drake 2B )

Paul Christensen, W9AC

I agree.  I would like to see a setting that automatically pulls in the opposite sideband skirt at or near the CW offset frequency, or something user defined. 

In CW, it makes no sense to have perfectly symmetrical skirts at wide bandwidths.  We've then lost single-signal ability.  Using 600 Hz as an example, then for anything less than 600 Hz, we *would* want equal CW symmetry.   

Steve N4LQ

I think the K3 has this same odd characteristic but most receivers since 1957 were set to receive only one sideband at a time. Remember the HQ-170 had U, L and BOTH. This is BOTH. 

Paul Christensen, W9AC

Yes, the K3 requires the same game.  I have to move it from the default bandwidth settings to get a good balance between opposite sideband rejection and full audio.

Really, I'm just glad the sidebands are independently adjustable - even if it's manual.  That was something I had never experienced until purchasing the 6700. 

Paul, W9AC

Steve-N5AC

I'd like to better understand your comment, but I'm confused. As I'm sure you know, we let you pick the sideband and then the filters are centered on the pitch frequency in the sideband and the filter skirts reflect this. This just seemed like common sense to me. Are you saying that you would like the skirts to be sharp on one side and not as sharp on the other?

Paul Christensen, W9AC

In N4LQ's example, he's right in that when a wide CW filter is selected, (e.g., 3 KHz), the sidebands revert to being symmetrical after the filter button is selected.

Probably the best way to describe this is to go into CW mode, then select the 3 kHz filter, which is the widest preset CW bandwidth setting.  Now, tune across a strong CW carrier.  Approach the carrier from either direction,.  When you do, you will reach zero beat.  Then continuing through zero beat, you will hear the opposite sideband come up at the same amplitude as the desired sideband. 

Now, click the 3 KHz filter button again.  This time, pull the upper sideband to 600 Hz when in CW-L.  Try the same sweep though zero beat.  As you sweep through zero beat, you will no longer hear the opposite sideband but full audio balance is maintained.  The bandwidth has been reduced from 3 kHz to 2.1 kHz:   -1500 Hz LSB and +600 Hz USB (instead of +1500 Hz USB).  As the selected filter becomes narrower (i.e., below 600 Hz), then sidebands shoukd be symmetrical (as they are now) with no need to offset one against the other for single-signal reception.

Yes, we can do all of this manually and it works great.  But as N4LQ points out, every time a wide filter is selected in CW, it reverts back to a symmetrical passband, requiring a manual pulling in of the unwanted sideband. 

Hope this made a bit more sense.  A hands-on demo would speak many more words!

Paul;, W9AC

 

Steve N4LQ

Steve
Yes that's pretty much it. Most any receiver offsets the passband so when you tune through the signal and cross over the zero-beat point you quickly loose reception. The exception would be a direct conversion receiver where you can hear the cw tone on either sideband. Those are horrible because you can't easily tell which sidetone is correct. 
What seems even stranger is SSB modes are in fact set up correctly thus rejecting the opposite sideband. Why would a cw op want to hear the other sideband? 
The system is called "Single-Signal Recption" and it was developed after WWII. Receivers with just a single crystal filter allowed one to adjust the BFO offset away from zero beat and take advantage of the crystal filter which was centered at the IF's frequency. Signals on the opposite side of zero beat were rejected. 
I can't think of any superhet that didn't work this way until the K3 came along and now the Flex. 

Steve N4LQ

Let me add that when we click the 3k bandwidth button, that 3 khz bandwidth should all be on one side of zero. As bandwidth is decreased, only one side of the passband should move toward zero. 
This way we don't hear stations on the other side of zero beat which can be annoying and confusing. 

Paul Christensen, W9AC

> "The system is called "Single-Signal Recption" and it was developed after WWII."

Steve, I think single-signal designs went into mass production after WWII, but the concept was first popularized in the early-to-mid 1930s through a series of articles written by James Lamb, then QST's Technical Editor.  It was Lamb's single-signal crystal filter that went into the ca. 1935 HRO,  then later into the NC-101X.  Lamb had a series of articles on the subject through the 1930s.  They're works of pure genius.  

Paul, W9AC

Steve N4LQ

RIght on Paul...And I mentioned, SSB modes have it right! In fact, it won't let you hear both sidebands at once. CW is actually acting more like AM mode.
Paul - When I do as you suggest it does work however somehow my sidetone pitch slider ended up saying 100 instead of the 550 which I had previously set. That's just plain wrong since the pitch coming from my speaker is still 550 hz. 

Paul Christensen, W9AC

Same here.  The CW sidetone defaulted to 100 Hz after installing V1.3.  I then bumped it back up to 600 Hz where I like it.  Strange, I think it was at 600 Hz, but showed 100 Hz.  Anyway, fixed now. 

W5XZ - dan

FWIW, I had this issue a while back, if I follow what you guys are talking about.  On a quiet or dead band, I like wider cw filters.  1, 1.5, or even 3 khz, for example.  If you expect to get 'single signal' reception, you must set the pitch much higher than you might be accustomed to; effectively, 1/2 the filter bandwidth.  Since I prefer 800 hz, it's not a major problem, except at 3 khz..

or, am I off in the weeds?  73, w5xz, dan

Ken - NM9P

Let me try this in other words.   ...  There is no problem as long as the filter width is narrow enough that one side of the symetrical filter is less than the CW offset pitch, (i.e. 500Hz in my case)   But if you have a filter set wider than that, the placement of the CW not actually lets you hear on the other side of Zero beat......Requiring the user to shift the filter passband up or down in order to keep the total receive window on one side of the zero beat frequency.....

It would be nice if there was a function that automatically shifted things so that when the filter was too wide it automatically shifted the passband to keep the low end of the filter skirt above zero beat. (or below it, depending upon which "sideband" is selected in setup) 

Hope this doesn't confuse the issue any further.

Ken - NM9P

Steve N4LQ

Dan. That would work but the pitch of the residual band noise would soon drive one to madness.

Ken - NM9P

Actually, I like having the filter symetric to the signal, until the filter is so wide that it lets the filter skirt fall below Zero beat.  In any case, I hardly ever use anything wider than 500 Hz any more on CW, and usually nothing more than 100-250!  It is just so much quieter.  And with a 500 Hz offset, I am still good up to 1K bandwidth before it becomes a problem.  The only case I can see that would be an exception would be when trying to copy an SSB signal and send CW in return.  In this case, the addition of "CW TX while in SSB" mode would solve that one nicely!

Ken - NM9P

Steve N4LQ

Well stated Ken! 
Collins did this with gears in the 75A4. 

Paul Christensen, W9AC

Maybe time for N4LQ to demonstrate this in one of his excellent video presentations!

Paul, W9AC

Steve N4LQ

Naw...A thousand words tells a picture!

Steve N4LQ

Actually I can't even believe we are having this discussion. )

W5XZ - dan

i agree! but, 'it is what it is'....

Charles - K5UA

Hi Paul, Didn't know where to jump in on this topic, but what you are describing is what I want to do with IF shift....that is, shift the entire bandpass to one side or the other of the zero beat freq. The SSB example you describe is essentially the same concept, it puts the bandpass on one side of the zero beat so you don't hear anything from the other sideband. In your example, CW would work the same way as SSB works.....decreasing the CW bandwidth would only move the filter edge farthest away from the zero beat frequency, just as it does in SSB mode. When the fiter width is very narrow, then the ability to make fine adjustents using IF shift to precisely position the bandpass reative to the zero beat frequency would be outstanding. I didn't know this concept had a name. Thanks for the history lesson.

Steve N4LQ

It's called Passband Tuning, PBT or IF-SHIFT. PowerSDR had it so nothing new. The difference here is having it follow the bandwidth. 

Bob N7ZO

Agreed and good idea.  :-)

Charles - K5UA

One more comment about CW filters. In PowerSDR we had custom filter capability. Since my CW offset was 400 hz, my widest CW filter was 500 hz and I set the upper and lower roll-offs at 150 hz and 650 hz. This centered the filter on my cw offset. As I customized each subsequent filter all the way down to 25 hz, the upper and lower roll-offs were selected to center each filter on my 400 hz CW off-set. This provides the ideal narrow CW filter relationship for your preferred CW off-set. There is no opposite sideband reception with this technique UNLESS you define a filter width greater than twice your CW off-set. In my case with a 400 hz CW offset, defining. 1200 hz custom filter centered on 400 hz would put the lower edge of the filter 200 hz into the opposite sideband. And if you did this with a 3000 filter, you would be hearing signals 1100 kz into the opposite sideband. This would not be good in a crowded band. Until we have custom filter capability, switching from filter to filter will always reset the bandpass of the filter symmetrically centered on the CW offset. No problem for filters less than twice the CW offset in width, but it is a problem for any filter significantly wider than twice the CW offset. As Paul said, a demo would be worth a thousand words.

Steve-N5AC

OK thanks for all the great explanations.  I think the issue is that we didn't anticipate someone using a wide filter with CW.  I know some folks prefer this though.  Unless I hear otherwise, I have entered an issue in the system to suppress filters crossing DC and instead extend the filter on the other side.  So if you were at 500Hz pitch and specified a 2kHz filter, we would go from 0-2kHz instead of -500Hz to 1500Hz.  If I'm still not clued in, let me know.

Ken - NM9P

Thanks, I think that is good, Steve.  In addition to maintaining "single signal" reception, it will make working crossband CW/SSB possible without using a second slice, too.  Which I have occasionally done on 6 Meters.  (Until "CW active while in SSB mode" is implemented?)

Steve-N5AC

Yes.  Incidentally, I added CW is SSB modes as a possible enhancement (#687).

W5XZ - dan

that would be great, IMHO..

w5xz