I've been experimenting with the DSP filters on my K2 and I have found several combinations of DSP and crystal filtering that work very well on CW (at least for me). However, I have run into a strange phenomenon. Here is how I reproduce it: I set my crystal filter at its widest setting (1.8 KHz). The DSP filter is set for "low pass" with noise reduction on. With this combination, I am able to find two identical CW signals on either side of a center frequency. It operates just like a direct-conversion receiver. So here are a couple of questions for the group: First, has anyone else seen this happen? If not, do you think it may be a construction error on the DSP board? How about a screw-up in the filter alignment or programming? Thanks and 73, John, WA6L -- The world is best viewed through the ears of a horse. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
John...
Perhaps you have your BFO set in the middle of the wide band. That way you'd get a tone on either side. That said, I sure would like to hear about the CW filter settings that you found useful. Thanks in advance. ....robert #5957 At 06/04/2007 12:42, John Graf wrote: >I've been experimenting with the DSP filters on my K2 and I have >found several combinations of DSP and crystal filtering that work >very well on CW (at least for me). However, I have run into a >strange phenomenon. > >Here is how I reproduce it: I set my crystal filter at its widest >setting (1.8 KHz). The DSP filter is set for "low pass" with noise >reduction on. With this combination, I am able to find two identical >CW signals on either side of a center frequency. It operates just >like a direct-conversion receiver. > >So here are a couple of questions for the group: First, has anyone >else seen this happen? If not, do you think it may be a construction >error on the DSP board? How about a screw-up in the filter alignment >or programming? > >Thanks and 73, > >John, WA6L > > >-- >The world is best viewed through the ears of a horse. > >_______________________________________________ >Elecraft mailing list >Post to: [hidden email] >You must be a subscriber to post to the list. >Subscriber Info (Addr. Change, sub, unsub etc.): >http://mailman.qth.net/mailman/listinfo/elecraft > >Help: http://mailman.qth.net/subscribers.htm >Elecraft web page: http://www.elecraft.com Robert G. Strickland PhD ABPH - KE2WY [hidden email] Syracuse, New York USA _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by John Graf
John,
That is not a consequence of any construction error, but a consequence of your filter alignment. When you align a wide filter for either CW or SSB, do NOT CENTER it. Use Spectrogram and align it so that the low frequency edge of the passband cuts off above zero frequency (I prefer 200 Hz at the lowest corner). In other words, an 1800 Hz wide CW filter should have an audio response from 200 to 2000 Hz. Aligned that way, you will not be able to hear a signal go through zero frequency and come out on the opposite sideband (like a direct conversion receiver). I would wager that you have that wide filter aligned so it receives from perhaps 900 Hz down to -900 Hz (the -900 Hz is on the opposite sideband, so you hear both sidebands for pitches less than 900 Hz). If you like a really wide CW filter, try using the OP1 filter instead of the variable filter (assuming you have the KSB2 installed). It is flatter and more 'listenable' than using the varaible filter at a wide setting. I usually recommend setting the variable filter to 1200 Hz wt the widest although some filters are quite usable out to 1500 or even 1800 Hz, but on some, the passband ripple becomes quiter pronounced at wider bandwidths. Spectrogram will show all that to you visually. If you have not used it to set up your filters, I recommend that you download it (version 5.17 is available free from www.n0ss.net) and try it out. Once you have the CW filters set the way you want, then you can use the DSP to augment those filters and improve the ultimate rejection and provide a sharper filter slope, but operate with the most narrow CW filter as your first course of action to combat QRM - if you do not, strong undesired signals in the CW filter passband will capture the receiver AGC and de-sense the receiver making all signals in the passband (including the weak one you are trying to copy) weaker - operating with a narrow CW filter is the only way to combat that (other than turning off the AGC). For additional information on filter alignment, check the info on my website http://w3fpr.qrpradio.com or Tom Hammond's site www.n0ss.net or on the builder's Resource page of the Elecraft website. 73, Don W3FPR John Graf wrote: > > I've been experimenting with the DSP filters on my K2 and I have found > several combinations of DSP and crystal filtering that work very well on > CW (at least for me). However, I have run into a strange phenomenon. > > Here is how I reproduce it: I set my crystal filter at its widest > setting (1.8 KHz). The DSP filter is set for "low pass" with noise > reduction on. With this combination, I am able to find two identical CW > signals on either side of a center frequency. It operates just like a > direct-conversion receiver. > > So here are a couple of questions for the group: First, has anyone else > seen this happen? If not, do you think it may be a construction error > on the DSP board? How about a screw-up in the filter alignment or > programming? > > Thanks and 73, > > John, WA6L Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by RobertG
Hi, Robert, Yep -- I am sure that is what it is. You have confirmed what other people have told me and I will change the BFO this weekend. It is sure nice to have the resources and experience that I have found on the Elecraft reflector. Well, I think filter settings are a lot a chile recipe. Different people have different ideas of what is good and bad. But having said that, I am happy to share what works best for me. I haven't owned the K2 for that long. I also have a K1 and an IC-746pro. Both of these have different filtering setups, and the IC-746pro has DSP, but implemented much differently than the K2. So my settings are the result of experimentation. I have my 2nd crystal filter set for 800 Hz. I have found that this is a great compromise between the 1st filter (wide open) and the narrower filters. This then feeds 800 Hz bandpass to the DSP. So I just keep the crystal filters at that setting. The DSP I have set up with 800, 400, and 200 Hz. Noise reduction is on and set to 2. When I am searching for a signal I keep it at 800. Once I find the signal I have the option of narrowing the DSP to 400 or 200 if the band is busy. I have tried both the "soft" and "regular" CW filter settings on DSP. Honestly, I don't see much difference. However, after operating on a couple of contests I decided I liked the way things worked with the regular setting (big 'C' on the filter window) a little better. Anyhow, this works for me. Thanks again for your help on the BFO issue! 73, John, WA6L On Jun 5, 2007, at 8:03 PM, Robert G. Strickland wrote: > John... > > Perhaps you have your BFO set in the middle of the wide band. That > way you'd get a tone on either side. That said, I sure would like > to hear about the CW filter settings that you found useful. > > Thanks in advance. > > ....robert #5957 > > At 06/04/2007 12:42, John Graf wrote: > -- The world is best viewed through the ears of a horse. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
In reply to this post by Don Wilhelm-3
Excellent!! Thanks for taking the time to write and for the in-depth explanation. I am constantly amazed (and grateful) for the expertise available here on the reflector. I will be working to correct the problem this weekend. I have used Spectrogram (though obviously not well) and have a copy. I like your idea of using the OPT1 filter and will give that a try. Thanks again and 73, John, WA6L On Jun 5, 2007, at 8:54 PM, Don Wilhelm wrote: > John, > > That is not a consequence of any construction error, but a > consequence of your filter alignment. > > When you align a wide filter for either CW or SSB, do NOT CENTER > it. Use Spectrogram and align it so that the low frequency edge of > the passband cuts off above zero frequency (I prefer 200 Hz at the > lowest corner). In other words, an 1800 Hz wide CW filter should > have an audio response from 200 to 2000 Hz. > -- The world is best viewed through the ears of a horse. _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
Free forum by Nabble | Edit this page |