+++++++++++++++++ From: "Don Brown" To: "COLIN WHITMORE" , "Elecraft" Subject: Re: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 08:04:59 -0500 Hi No this is not normal for a properly aligned K2. If you have FL4 set up a= t 100 Hz or less (I set mine to 150) then you may hear a small drop in si= gnal but it should not go away. All of the other filters should not chang= e the signal strength at all. As the filters get narrower the noise and a= djacent signal go away not the desired signal. =20 If your radio does not perform this way then you need to realign the filt= ers using spectrogram and a noise source. The settings in the K2 manual w= ill get your radio working, but not to the peak performance. Spectrogram = or some other audio spectrum analyzer program with a broad band noise gen= erator as a signal source is the only easy way to get your radio aligned = perfectly. =20 A few times I just could not get the radio to set up so it worked properl= y. I could hear the tone on CW reverse but it would go away on CW normal.= I checked the alignment several times and compared the settings to the K= 2 manual and they were in the ball park but it just would not work the wa= y I new it should. I finally did a hard reset of the K2 (hold 4,5 and 6 o= n power up) and set up the filters with Spectrogram and everything fell i= nto line. You may need to do this as a last resort if you cannot get it t= o align. You may need to do the CAL PLL also if you do the reset. The res= et clears out all of the memory and loads the factory defaults. The procedure I outlined in my previous post is a performance check not a= n alignment technique. It tells you if you have it set up right but you s= till need to align the radio by Tom's method described on the website usi= ng Spectrogram and the noise generator. Don Brown KD5NDB 7 K2's and counting 3 K1's 4 KPA100 all Options ----- Original Message ----- From: COLIN WHITMORE Sent: Tuesday, October 15, 2002 11:55 PM To: Elecraft Subject: [Elecraft] K2 filter attenuation I've noticed that when trying to grab a weak signal I have to keep the fi= lter wide. If I try to narrow the filter selection to block out an adjace= nt signal there seems to be some (significant) attenuation to the filter.= That's not a problem for strong signals but for the very weak signals I = will totally lose the station. So if I want to work a weak signal I have = to put up with other stations well off the centered frequency. Is this no= rmal? I've checked the alignment (according to Don, KD5NDB) and the 40M b= irdie remains fairly consistent in pitch as I switch through the filters.= Any suggestions? Cheers, Colin N=D8YG +++++++++++++++++ From: "Don Brown" To: "Dan Barker" , "Elecraft" Subject: Re: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 09:47:39 -0500 Hi Another way to check if you are on frequency (assuming your filters are s= et up right) is to hold the CW REV button. The CW tone frequency should b= e the same on each side when switched. If one side is high and the other = side is low just adjust the VFO a little to make them the same. For some = people it is easier to hear a difference in two frequencies than to hear = the zero beat of the side-tone. Don Brown KD5NDB ----- Original Message ----- From: Dan Barker Sent: Wednesday, October 16, 2002 9:18 AM To: Elecraft Subject: FW: [Elecraft] K2 filter attenuation If you haven't zero-beat the weak sig correctly, it may be falling off th= e edge of the tighter filter. Always look at the simple solution first. Thr= ow a Spectrogram on the output of the rig on a weak one, and see if the CW spike (or bump, if it's really weak) is dead-denter in the passband on th= e tighter filter. You may just need to QSY a hundred or two Hz to bring the= m up. My rig is aligned fine, but my ear isn't. When they disappear, they a= re usually nearby. Now, if I could just hear whether to go up or down, not t= o lose them. Dan / WG4S / K2 #2456 No this is not normal for a properly aligned K2. [but perfectly normal for a poorly aligned ear!] +++++++++++++++++ To: cncwhitmore at msn.com Cc: elecraft at mailman.qth.net Date: Wed, 16 Oct 2002 08:10:33 -0700 Subject: Re: [Elecraft] K2 filter attenuation From: k6se at juno.com Colin, N0YGY wrote: "I've noticed that when trying to grab a weak signal I have to keep the filter wide. If I try to narrow the filter selection to block out an adjacent signal there seems to be some (significant) attenuation to the filter. That's not a problem for strong signals but for the very weak signals I will totally lose the station. So if I want to work a weak signal I have to put up with other stations well off the centered frequency. Is this normal? I've checked the alignment (according to Don, KD5NDB) and the 40M birdie remains fairly consistent in pitch as I switch through the filters. Any suggestions?" ========== Sounds like the classic BFO temperature drift to me, which causes the desired sigbal to drift out of the passband of narrow CW filter settings. Try running Spectrogram when the problem occurs. 73, de Earl, K6SE +++++++++++++++++ Reply-To: From: "Ron D'Eau Claire" To: "'Elecraft'" Subject: RE: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 08:42:54 -0700 I've noticed that when trying to grab a weak signal I have to keep the filter wide. If I try to narrow the filter selection to block out an adjacent signal there seems to be some (significant) attenuation to the filter. ... Colin N=D8YGY ---------------------------- Is it really attenuation, Colin? Sometimes I can read a very weak signal better at wider bandwidths, even with some QRM. That's because of what he filter does to the background noise.=20 When the band noise encounters a narrow filter, individual spikes of noise get stretched out - made broader. That's a normal consequence of any filter and it shows up dramatically in a narrow filter. It also happens to the signal itself on the on/off transitions at the end of each dit and dah. That makes the transitions at the end of every dit and dah less clear. The combination can make a weak signal much harder to copy at a narrow bandwidth and it is with a wider bandwidth.=20 The total effect is to make the signal sound weaker in relation to the noise and, unless he is sending quite slowly, makes the CW a little harder to copy even without the extra noise because the code element spacing has been shortened and less distinct. =20 Ron AC7AC K2 # 1289 +++++++++++++++ To: elecraft at mailman.qth.net Date: Wed, 16 Oct 2002 13:35:41 -0700 Subject: Re: [Elecraft] K2 filter attenuation From: k6se at juno.com Ron, AC7AC wrote: "Sometimes I can read a very weak signal better at wider bandwidths, even with some QRM. That's because of what the filter does to the background noise. When the band noise encounters a narrow filter, individual spikes of noise get stretched out - made broader. That's a normal consequence of any filter and it shows up dramatically in a narrow filter. It also happens to the signal itself on the on/off transitions at the end of each dit and dah. That makes the transitions at the end of every dit and dah less clear. The combination can make a weak signal much harder to copy at a narrow bandwidth than it is with a wider bandwidth. The total effect is to make the signal sound weaker in relation to the noise and, unless he is sending quite slowly, makes the CW a little harder to copy even without the extra noise because the code element spacing has been shortened and less distinct." ========== I "specialize" in weak signal DXing on 160m CW (the 2nd on the west coast to achieve 160m DXCC, 55th in the world - now at 182 countires on 160m), and I have never run into the phenomenon that Ron talks about (and there is a LOT of "band" noise there!). Simply because a narrow CW filter provides an appreciable improvement in the S/N ratio of the receiver, the narrower filter ALWAYS makes copy of the weak signal better (even with the K2, as long as the temperature is such that the signal is centered in the narrow filter passband). NEVER have I been able to copy a weak CW signal better with a wide filter than with a narrow filter. The opposite has ALWAYS been true while using many different receivers over many years. BTW, "ringing" when using a narrow filter is no problem if you have your receiver adjusted to listen to a low-pitched tone (400 Hz or less). Those who complain of "ringing" invariably are listening to a significantly higher tone. The lower pitch also helps to "filter out" high-pitched band noise and IF noise (as well as the "ringing"). If the DX station sends very slowly, this tends to make copy more difficult due to QRN and QSB. Most 160m DXers prefer to use around 20 WPM for DX contacts. Perhaps Ron runs his receiver with the AGC turned on while listening to weak CW signals. In that case, a loud static crash would obliterate reception until the AGC decayed. With the AGC turned off, this would not happen. 73, de Earl, K6SE ++++++++++++++++ Reply-To: From: "Ron D'Eau Claire" To: Subject: RE: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 14:34:29 -0700 Earl, K6SE wrote: Perhaps Ron runs his receiver with the AGC turned on while listening to weak CW signals. In that case, a loud static crash would obliterate reception until the AGC decayed. With the AGC turned off, this would not happen. ----------------------------------------------------------------- No, I typically listen with AGC off and manage the gain using the "RF" gain control. Probably the largest issue in the design of filters for receivers since they were first explored almost a century ago is the issue of their effect on many types of noise with sharp rise and fall times. Many of the names given various filter designs are the names of engineers who investigated the relative benefits of various shapes of filter response on signals. In general, the sharper the edge on a filter, the greater the effect on noise with high-frequency rise and fall times that produces more overall noise within the filter bandpass, but a meaningful analysis is the stuff of much research and complex mathematical modeling. In GENERAL, narrower bandwidths means a better signal-to-noise ratio. But when getting down into the signals that are in the noise, it is unlikely that anyone has built a filter that can compete with a well-trained human brain and ears. It is in that "twilight" zone that narrower is not always better. Often the brain can do better with a less highly filtered signal, in my experience. That's based on 50 years of digging signals "out of the mud" as a commercial radiotelegraph operator and as a radio amateur. Something else that experience has taught me is that no two operators approach a problem in exactly the same way. What seems to work better for one operator doesn't work as well for another. The technique varies just like any other expression of personal skills. I didn't bring this up to argue technique or to say that only one way of doing things is "right" and others are "wrong". I only made the suggestion because Colin reported that signals sounded weaker when his filters were narrow, and it was my experience that can be due to the way the filter reacts in the presence of certain noise when listening to very weak signals. Ron AC7AC K2 # 1289 ++++++++++++++++ From: "Wallace, Andy" To: elecraft at mailman.qth.net Subject: RE: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 17:43:40 -0400 > -----Original Message----- > From: Ron D'Eau Claire [mailto:rondec at easystreet.com] > Sent: Wednesday, October 16, 2002 5:34 PM > To: elecraft at mailman.qth.net > Subject: RE: [Elecraft] K2 filter attenuation > In GENERAL, narrower bandwidths means a better signal-to-noise ratio. > But when getting down into the signals that are in the noise, it is > unlikely that anyone has built a filter that can compete with a > well-trained human brain and ears. It is in that "twilight" zone that > narrower is not always better. Often the brain can do better > with a less highly filtered signal, in my experience. This is what I find at times, Ron. It depends on the conditions. But something to remember is that when you select a narrow filter, you cut the passband down such that static crashes, electrical noise, etc., as WELL as the signal, are all within that passband and therefore sound similar in frequency. Some filters also exhibit ringing, too. This all combines to make things muddy under CERTAIN conditions. With a wider filter setting, sometimes the brain can more easily pick a weak CW signal out of a combination of several signals, noise, and static that ranges from 0 to say 700 Hz in frequency. It's like picking out the solo violin from the orchestra. Believe me, I am SO happy to have the choice of THREE bandwidths in my humble K1, in a time when other manufacturers only include the SSB filter in their rigs. -Andy ++++++++++++++++ Date: Wed, 16 Oct 2002 15:06:35 -0700 From: Vic Rosenthal Organization: Transparent Software To: Elecraft Subject: Re: [Elecraft] K2 filter attenuation Earl, K6SE wrote: > Simply because a narrow CW filter provides an appreciable improvement in > the S/N ratio of the receiver, the narrower filter ALWAYS makes copy of > the weak signal better... Ron D'Eau Claire wrote: > > In GENERAL, narrower bandwidths means a better signal-to-noise ratio. > But when getting down into the signals that are in the noise, it is > unlikely that anyone has built a filter that can compete with a > well-trained human brain and ears. As a CW DXer, I've been following the comments of Ron and Earl with great interest. I used to be 100 percent in Earl's camp, thinking that it was always better to use as much selectivity as possible on weak signals. Lately I've been coming over to Ron's side. There are a lot of posssible issues here, such as 1) The characteristics of the interfering noise -- atmospherics, powerline noise, RF burbles, computer noise, ad. inf. 2) The shape of the filter. 3) Other receiver characteristics, perhaps in the audio and AGC sections. 4) The operator. For what it's worth, I've moved into the 'low selectivity' camp pretty much since I've been using my K2 regularly. Before that I used a TS850 with cascaded 400 Hz Inrad filters with the passband narrowed to about 200 Hz most of the time -- and I often switched in an outboard DSP or SCAF filter. Now the outboard filters are disconnected, I have my K2 set to 1.4 KHz and reduce it only for QRM, and I rarely use the K2 audio filter. I'm not sure what's changed. Vic K2VCO ++++++++++++++++ Reply-To: From: "Ron D'Eau Claire" To: Subject: RE: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 15:43:29 -0700 Believe me, I am SO happy to have the choice of THREE bandwidths in my humble K1, in a time when other manufacturers only include the SSB filter in their rigs. -Andy ------------------------------ Truer words were never spoken, Andy! I have always found the biggest problem to be with very strong signals in the pass band while listening to a relatively weaker one. It's like trying to hold a conversation at normal volume at a crowded party while some loudmouth right behind me is putting out way too many dB talking to someone else. My brain gets 'captured' by the loudmouth and I can't concentrate on the quieter voice, even though it is loud enough to hear. At times like that a good filter is invaluable! But I like to narrow down the bandwidth only as much as needed to kill the loud signal(s). That's just a personal preference, I know. Many operators tell me that they always work CW with the narrowest filter they have regardless. Ron AC7AC K2 # 1289 +++++++++++++++ From: "Rich Lentz" To: Subject: RE: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 19:25:26 -0500 Back to the original subject - I also hear and see (with spectrogram) a noticeable drop in signal (in excess of 10 db) on the narrowest filter. I am positive the filter is properly aligned, and this happens on normal, reverse, USB, and LSB. My filter numbers are almost the same as that in the manual, with a slight difference for a 750 Hz CW tone. The same decrease is also shown with the noise generator on the narrowest filter. I cant believe it is "tuning" or the filters are "miss tuned," I might believe there is an improper component or miss-wiring that I can't find. But I have checked each component at least 8 times (counting the four times that I checked every part when the K2 was built. The level matches the other filters when it is as wired as the other filters which makes me believe there is no incorrect components or miss-wiring. As the first question in this thread "Any suggestions?" Rich KE0X +++++++++++++ Date: Wed, 16 Oct 2002 17:49:30 -0700 From: Vic Rosenthal Organization: Transparent Software To: Rich Lentz Cc: elecraft at mailman.qth.net Subject: Re: [Elecraft] K2 filter attenuation Rich Lentz wrote: > > Back to the original subject - > > I also hear and see (with spectrogram) a noticeable drop in signal (in > excess of 10 db) on the narrowest filter. > As the first question in this thread "Any suggestions?" Triple check the number of turns on both windings of T7. The current K2 manual calls for the 3-4 winding to have 20t and the 1-2 winding to have 5t. I had a similar symptom in my K2 early on and found I had miscounted the turns on one of the windings. 73 Vic K2VCO +++++++++++++++ Date: Wed, 16 Oct 2002 18:22:14 -0700 From: Eric Swartz WA6HHQ - Elecraft Organization: Elecraft To: Rich Lentz Cc: elecraft at mailman.qth.net Subject: Re: [Elecraft] K2 filter attenuation Hi Rich, When you say 'narrowest filter', what bandwidth are you referring to? Is it below 400 Hz to 500 Hz? The K2 was designed for a proper impedance match to the CW crystal filters when they are adjusted in a range between 400 and 800 Hz wide. Narrower filter settings are not guaranteed and are going to vary from rig to rig. Wider settings will have an increase in ripple in their passband. (That's why we supply an additional SSB fixed width filter in the KSB2.) We have seen good results on most K2s down to 300 Hz. I guess we shouldn't be surprised that some builders will try and push the envelope even further :*) Settings below 400 Hz are outside the main K2 design parameters and certainly will have more attenuation in their passbands. In most cases though, since the noise figure of the RX is set by the front end of the receiver (pre-amp, mixer and post amp) this will not directly impact sensitivity. It will reduce overall RX gain though. 73, Eric WA6HHQ ++++++++++++++++ From: "Rich Lentz" To: "'Eric Swartz WA6HHQ - Elecraft'" Cc: Subject: RE: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 20:42:48 -0500 Eric: FL3 is set to 0.50 which I assume is 500 Hz and FL4 is set for 0.25 which I assume is 250 Hz. Anything narrower than 250 gives to much attenuation. As I said earlier The attenuation is similar, about 10 db for either a signal e.g. the 4.000 mhz clock or for a noise generator, or fairly white antenna noise. Originally had it set for 0.10 (100 Hz) as in the table on page 86, but the attenuation was just to much to be of any use. Have made the SSB flattening mods, including the latest one with the added resistor - but they shouldn't affect CW, should they? Rich KE0X +++++++++++++++++ Date: Wed, 16 Oct 2002 19:53:08 -0600 To: "Rich Lentz" , From: Larry East Subject: RE: [Elecraft] K2 filter attenuation At 07:25 PM 10/16/02 -0500, Rich Lentz wrote: >Back to the original subject - > >I also hear and see (with spectrogram) a noticeable drop in signal (in >excess of 10 db) on the narrowest filter. I am positive the filter is >properly aligned, and this happens on normal, reverse, USB, and LSB. My >filter numbers are almost the same as that in the manual, with a slight >difference for a 750 Hz CW tone. The same decrease is also shown with >the noise generator on the narrowest filter. I cant believe it is >"tuning" or the filters are "miss tuned," I might believe there is an >improper component or miss-wiring that I can't find. But I have checked >each component at least 8 times (counting the four times that I checked >every part when the K2 was built. The level matches the other filters >when it is as wired as the other filters which makes me believe there is >no incorrect components or miss-wiring. > >As the first question in this thread "Any suggestions?" I also see significant attenuation with the 150 kHz filter -- but not as much as you are apparently seeing. Here's a copy of measurements that I made on my K2 in May of 2001 (using a 700 CW Hz tone): Filter Rel. Attn. 1000Hz - 0.0dB (reference) 600Hz - -2.5dB 400Hz - -4.5dB 150Hz - -9.0dB (values rounded to the nearest 0.5dB) Each filter was centered on 700 Hz within 10Hz using Spectrogram. According to Gary S., there shouldn't be as much attenuation as I measured and he sent me a new set of crystals to try. However, I've never installed them for one reason or another. So... your problem could well be a poorly matched set of filter crystals. Even with the 9dB of attenuation, I haven't had any problems "losing" weak signals when switching to the 150 kHz filter. And with the narrow audio filter also in use, signals just pop out of the noise! Maybe one of these days I'll try the new set of crystals... 72/73, Larry W1HUE/7 +++++++++++++++ From: James Hammons To: elecraft at mailman.qth.net Subject: [Elecraft] K2 filter attenuation Date: Wed, 16 Oct 2002 21:58:15 -0400 Reply-To: james-e-hammons at worldnet.att.net I followed the instructions for the setup of the filters provided in both the manual and SSB manual. The recommended setting for FL4 is 0.10 dc. or 100 cps. If this setting is not recommended then why specify the setting. I would bet most K2's are set up this way. I recently changed the setting to 0.20 to keep the drift within the passband. I removed one of the PLL crystals and found that one crystal would pull 11.5 kc, not the recommended 12kc but it worked and it reduced the drift of the PLL to .5 ppm/degree C an improvement of better than half. The BFO still drifts more than 1ppm/degree C. How is the recommended mod coming along. Could we have a few more hints of what is involved with the BFO mod. I was thinking along the lines of if an output could be found which only happens on the reverse cw or LSB setting then the signal could switch to a second crystal or insert a fixed cap to pull the crystal and thus reduced the voltage change required or eliminate the two crystals in parallel. perhaps there is a output at the LCD that would be usable for this? +++++++++++++