++++++++++++++++++++ See also Spectrogram Notes, Usage ++++++++++++++++++++ To: elecraft at mailman.qth.net From: CHARLES K BROWN Date: Tue, 04 Jun 2002 01:15:59 EDT Subject: [Elecraft] N0SS CW Filter Alignment Available now as text files: N0SS CW Filter Alignment Using Spectrogram/ for the K2 and another file for the K1-- no graphics-- just the text files from the .pdf data files. Recommend you use the graphics files and Acrobat Reader if you receive attachments and .pdf files on your E-mail program. Otherwise, the text files will guide you without any pictures. W3SMK WWV Calibration Using Spectrogram- this provides a guide in setting the Calibration to WWV signals with Spectrogram and explains the basic settings in Spectrogram. Allows a precise VFO Calibration to go along with the Filter alignment. Explains about WWV signals and what to watch for on the Spectrogram display. -files are on this computer-- N4SO Ken Brown Mobile, AL/EM50tk NorCal-20/5 watts/4 ele. beam +++++++++++++++++++++ Date: Tue, 4 Jun 2002 09:27:48 +0100 To: elecraft at mailman.qth.net From: Trevor Day Subject: Re: [Elecraft] N0SS CW Filter Alignment Just completed the CW and SSB filter alignment of K2 #2637 using Spectrogram ($25 was well worth it if only to be able to take it all at a leisurely pace). I had almost as much fun setting up the filters as I have operating the kit :)) It also put new life into an old antenna noise bridge that has been gracing the junk box for far too long. My XYL (G4KYY) thinks I'm crazy as I attempted to explain to her the beauty of watching the smooth change of filter characteristics as the BFO frequency was adjusted. I haven't done the SSB mod yet, so thats going to be a good excuse to do it all again :))) Trev G3ZYY In message <20020604.051224.7655.4.n4so at juno.com>, CHARLES K BROWN writes >Available now as text files: > >N0SS CW Filter Alignment Using Spectrogram/ -- snip -- ++++++++++++++++++++++ Date: Wed, 05 Jun 2002 23:16:58 -0400 From: frastephen at comcast.net Subject: Re: [Elecraft] N0SS CW Filter Alignment To: elecraft at mailman.qth.net, RMariutto at aol.com From: > > > > Subject: N0SS CW Filter Alignment Using Spectrogram > > I think my problem is that I am trying to do two things at the same time: 1) > adjust the filters 2) set the K2 to the proper receive frequency. Right, work on getting the K2 on frequency first, then adjust the filters. I think you want to make sure you are adjusting C22 with the widest filter you have... >I noticed > that my receive frequency was about 200 hz lower than indicated on the > display. In the manual there is a proceedure for adjusting a variable cap > that corrects this, boy is that cap touchy. Yes, it is!!! I found that it was helpful to give it a few twists to loosen it up a bit before I made the adjustment -- took a little of the "stickiness" out of it, and also I understand that sometimes these trimmer cap plates do not seat properly sometimes unless you turn them a bit before you try to adjsut them. >I noticed that the 500 & 600 > hz tone markers don't line up exactly with the respective marks on the > Spectrogram when I shift from usb to lsb or (cw & cw reverse) . Hmmm, I just tried it again on Spectrogram on 10 MHz and both sidebands come out exactly the same.... Curious... Don't forget that the 500 Hz and 600 Hz tones basically alternate .... when you see the big spike at 1000 Hz and 2000 Hz it will shift to the other standard frequency for that minute. >In other > words, if I put the tone on 500 hz on lower sideband then shift the K2 to > upper sideband it is sort of close to 500 hz, if I try I can't get both on > 500 hz . For instance on 10 mhz lower sidband displays 999982 & upper > displays 999977. On 15 mhz lower displays 1499973 & upper displays 1499967. > Both about 50 hz off but different. Does this make sense? No.... >I feel that I > should get the frequency to display correctly before I set the filters , is > this right? Yes! Interesting how the LSB frequency is HIGHER than the USB frequency.... Here are a few ideas. Did you do the rough alignment of the filters per p. 86? I would make sure that is roughly OK before you proceed with the fine adjustments. It sounds like the filters maybe are set to the wrong side of the passband. Perhaps if you had these backwards that would account for the problem you are experiencing. Remember when you followed Tom's instructions you found the center of the passband and marked down the DAC count for that place (probably around 170 or so)? The CW filter should be set to a DAC count LOWER than the center and the CW REV should be set to a DAC count HIGHER than the center. Make sure the rough cut as described on p. 86 is in place before you do these fine tuning adjustments. I would stick with one sideband as you go through this procedure. When you get one sideband correct the other one naturally falls into place. That might be all you need to do. Keep working with USB until you get it dead on and then verify your effort with LSB. Also, you will probably have to run through this procedure several times, getting closer each time, until you are right on! It took me three iterations of the procedure to get within the theoretical limit. When you are following the instructions to adjust C22 to calibrate the VFO, make sure you do the step where you move the counter probe back to TP2 then start up CAL FIL, change the BFO control parameter for the current filter at least one count, then move it back to the original setting (e.g. 126-->127-->126) then tap MENU to exit -- this forces the K2 to re-measure the BFO frequency. Also, don't forget that the specification VFO accuracy is +/- 30 Hz. The display shows down to 10 Hz. I am finding that the K at stays within +/- 10 Hz over time since I made the fine adjsutment. Does that help???? Let us know how you do! Peace & 73 Stephen W3SMK K2 #2644 +++++++++++++++++ From: "Rich Lentz" To: Subject: RE: [Elecraft] N0SS CW Filter Alignment Date: Wed, 5 Jun 2002 22:25:05 -0500 The K2 has 10 hZ steps thus it is almost impossible to get it exactly. After about 10 iterations I was able to achieve the following after getting WWV within 10 hZ. On lower bands (below the reversal of BFO shift) got the following results reading the frequency off of the MIXW 2 display. (Seems to provide a finer resolution than Spectrogram.) FL1 FL2 FL3 FL4 USB REF -4 -6 -10 LSB REF 0.0 0.0 0.0 CW REF -5 -5 -5 CWR REF -7 -7 -7 USB, CW, and CWR all seem to provide absolutely no change in pitch (Using 750 Hz) Listening to a steady carrier on the air with RF gain and ATT set so that it is just audible and at normal listing level. Will probably play some more over the weekend (After I write down all the numbers.) and see if it possible to get closer. Messing around with PSK, It looks to me as there is no reason not to use LSB (on the lower bands), especially if you intend on dropping to a narrow filter. Tom, N0SS provided this in an earlier message I have found that you can get the shift reduced to less than 1 Hz by just repeating the procedure numerous times. Sometimes by just adjusting one filter at a time. Often I just go in to the procedure and just the BFO by one 'notch' and back again and exit, so it will recalculate the offset. I have been able to get it down to the point where there is no audible pitch change (musicians ear) between the filter settings, and just a slight phase twist on the PSK tuning eye. It WILL drive you nuts, but it can be much better than 10 - 20 Hz. I have found that if you change the cap adjusting the frequency, that you need to 1) change the filter (even if back to the same setting) and then 2) re run Call-PLL. I found that when I got close to the proper alignment with WVV, that I needed to use on half of the calculated value for changing the cap. Remember, Only adjust LSB for WWV, the BFO shift will fix the other. Look at the "crystal and BFO settings" section of the manual. Let the rig set for about four hours before starting, and put the cover back on quickly after any adjustment. Don't set BFO/filters until happy with WWV alignment. If changed you start from scratch. Rich -----Original Message----- From: elecraft-admin at mailman.qth.net [mailto:elecraft-admin at mailman.qth.net] On Behalf Of Ron D'Eau Claire Sent: Wednesday, June 05, 2002 9:25 PM To: elecraft at mailman.qth.net Subject: Re: [Elecraft] N0SS CW Filter Alignment > I noticed that the 500 & 600 hz tone markers don't line up exactly with the respective marks on the > Spectrogram when I shift from usb to lsb or (cw & cw reverse) . In other > words, if I put the tone on 500 hz on lower sideband then shift the K2 to > upper sideband it is sort of close to 500 hz, if I try I can't get both on > 500 hz . For instance on 10 mhz lower sidband displays 999982 & upper > displays 999977. > Ron N8SZY As far as I can see, that is entirely "normal" for a K2. The i-f filter is NOT symmetrical. That means that the slope of the lower frequency side of the bandpass is NOT the same when shifting from upper to lower sideband. If I set the filters to produce the same response on a DSB AM signal when switching from upper to lower sideband, the BFO is way off from the right setting for good signal quality. Indeed, I am rather disappointed in the gradual slope on the lower sideband in any case. It produces fairly "muddy" sounding audio compared to the upper sideband. Optimizing the BFO settings to minimize this means that I must retune slightly when shifting between upper and lower sideband. That's not an issue for me, since I don't normally listen to double-sideband signals, Hi! Ron AC7AC K2 # 1289 +++++++++++++++++++++ Reply-To: "Don Wilhelm" From: "Don Wilhelm" To: , , Subject: Re: [Elecraft] N0SS CW Filter Alignment Date: Thu, 6 Jun 2002 00:28:06 -0400 Folks, When you adjust C22, you are changing the 'master oscillator' for the entire K2 and any prior calibration will be in error. After adjusting C22 (each time) remember to run CAL PLL. Until CAL PLL is run the frequency display will still be in error. BTW, that is the reason attempting to adjust C22 to zero beat WWV can take many interations to get it right. If you have a good calibrated frequency counter, you can hit it right on the head by using the external counter method detailed in the K2 manual. 73, Don W3FPR ----- Original Message ----- > > Right, work on getting the K2 on frequency first, then adjust the filters. > I think you want to make sure you are adjusting C22 with the widest filter > you have... > > > >I noticed > > that my receive frequency was about 200 hz lower than indicated on the > > display. In the manual there is a proceedure for adjusting a variable cap > > that corrects this, boy is that cap touchy. > +++++++++++++++++++ To: RMariutto at aol.com, elecraft at mailman.qth.net Date: Wed, 5 Jun 2002 22:18:21 -0700 From: k6se at juno.com Subject: [Elecraft] Filter Alignment Using Spectrogram I've found that the best way to set up the xtal filters in the K2 using Spectrogram is a follows: 1) Allow your K2 to warm up for about an hour to alleviate the BFO frequency drift problem. The internal frequency counter probe must be in TP2 for this procedure. 2) For CW, set up Spectrogram to display from 0 to 1500 Hz and set up a marker at the frequency you have your K2's sidetone pitch set at. 2a) For SSB, set up Spectrogram to display from 0 to 2750 Hz and set up markers at 500 Hz and 2400 Hz. 3) Disconnect the antenna from the K2, turn the preamp on, and use only background noise while adjusting the filter passband displays in Spectrogram. 4) Set up the filter bandwidths that you want for both CW and SSB using CAL FIL. 5) While in CAL FIL, tap BAND-. 6) Turn the K2's VFO knob so that each individual passband for all SSB filters is centered evenly between the 500 and 2400 Hz markers you set up in step 2a above. Do this for both LSB and USB (8 adjustments). This completes the SSB filter setup. 6a) Turn the K2's VFO knob to center the passband display for each filter passband on Spectrogram at the marker you set up in step 2 above. For CW bandwidths greater than 0.70 kHz, set the lower frequency part of the passband at the marker. Do this for both CW normal and CW reverse (total of eight filter bandwidths). 7) Select the narrowest CW filter bandwidth you set up in step 4 above. 8) Temporarily exit the menu. In CW mode, tune in the "birdie" on 7000 kHz so that its pitch is the same as your sidetone pitch. 9) Re-enter CAL FIL and tap BAND-. 10) On Spectrogram, you should see the 7000 kHz signal peaked at the marker you set up in step 2 above. This is your reference point, so DO NOT adjust the BFO at this narrowest filter setting. 11) Tap XFIL to select the 2nd narrowest CW filter. If necessary, adjust the BFO by turning the VFO knob to center the 7000 kHz signal at the marker. 12) Repeat step 11 for the remaining two CW filter bandwidths. 13) Repeat steps 7, 10, 11, and 12 for all four filter bandwidths on the other CW sideband (CW normal or CW reverse). 14) Move the internal frequency couinter probe to TP1 and execute the CAL PLL procedure. (You should already have set C22 on the control board so that the 4 MHz xtal oscillator is exactly at 4000.0000 kHz prior to this procedure). ===== If you use the above procedure, you'll hear little (if any) change in pitch of a CW signal as you step through all CW filter bandwidths for both CW normal and CW reverse. If the received signal's pitch is set to your sidetone pitch, you should hear no difference in the received signal's pitch when you switch between CW normal and CW reverse. Contrary to other procedures posted, my procedure sets up the narrowest CW filter first and uses this as a "reference". You'll have much less trouble getting the pitch of received signals to remain the same for all filter bamndwidths if you do it this way (i.e., narrowest first). If you have the KAF2 option installed, it's very easy to adjust AF1 and AF2 using Spectrogram and the 7000 kHz "birdie". Just tune in the "birdie" so that its pitch matches your sidetone pitch, select AF2 by HOLDing XFIL/AFIL twice, and adjust both R1 and R2 on the KAF2 for max signal at the sidetone pitch marker on the Spectrogram display. 73, de Earl, K6SE 68 countries with K2 #2622 and counting..... +++++++++++++++++ From: "Rich Lentz" To: "'John McClain'" , Subject: RE: [Elecraft] N0SS CW Filter Alignment Date: Thu, 6 Jun 2002 16:58:41 -0500 These readings were taken with MixW 2 using WWV CARRIER (did not use the periodic tone) as a reference. As I recall, the carrier was about 7 Hz high (10,000,007 Hz) when I started collecting the data which is provided as REF. WWV at 5 MHz and the Russian standard at 14.996 MHz was also very close to about 7 Hz high. The table shows the difference from this reference value (all values were taken at 10 MHZ though) and only several were filters were checked at other frequencies and they were comparable. Thus, using USB and FL4 I was reading 9,999,997 instead of 10,000,000. All LSB filters were right on top of each other (no shift) and all were reading 10,000,007 (+/- a decimal point). However, the trace from the waterfall did not move. The table only shows the difference in Hz so that I did not have to write a table with 8/9 digits. The reason I used MixW is so that I can see the shift in the colored trace. With the waterfall expanded to X3 or X4 you can get very accurate readings of the frequency, in the decimal points even though they are useless. I used the snap (arrows) to lock in on the frequency from both (either) sides and kept the signal to the PC low enough that it locked on the same frequency from both sides. Look at K6SE's "[Elecraft] Filter Alignment Using Spectrogram" message and follow his advice. He has calibrated many K2's whereas I have only calibrated mine many times. Haven't used his approach yet as I start at the widest and work down, the same direction as you click through the filters. If the BFO has drifted during the process and you only adjust one filter it will be for naught as the reference frequency stored will be different than those stored for the other filters. Thus you need to change each filter, including the ones you didn't want to adjust, by one click and back so that the frequency stored will all have the same reference. Most important of all is that the reference clock must be right on or you are wasting time. Rich -----Original Message----- From: John McClain [mailto:digi2 at earthlink.net] Sent: Thursday, June 06, 2002 1:26 PM To: Rich Lentz; elecraft at mailman.qth.net Subject: Re: [Elecraft] N0SS CW Filter Alignment Rich, I found your message very interesting but I didn't understand this chart (i.e. what do the values of "ref -4 -6 -10" mean and where do they come from)? I am pleased with my filters but I must admit that there is a change of 20 to 30 hz when I switch between USB and LSB while using the same filter width (i.e USB with FL1 to LSB with FL1). I also see a 20-30 hz change when I switch from CW FL1 (OPT1) and CW FL2 (OPT1) but there is no shift between CW FL2 and CW FL3. Even though I have a registered copy of MixW, I have continued to use Spectrogram because it has a resolution down to 2hz when you use the cursor pointer and then read the pointed to frequency below the graph. Try as I might, even though I don't feel that it is really necessary, I can't get it any better than it currently is. > On lower bands (below the reversal of BFO shift) got the following > results reading the frequency off of the MIXW 2 display. (Seems to > provide a finer resolution than Spectrogram.) > > > FL1 FL2 FL3 FL4 > > USB REF -4 -6 -10 > LSB REF 0.0 0.0 0.0 > CW REF -5 -5 -5 > CWR REF -7 -7 -7 > Was the above information obtained from MixW? John K7SVV K2 #2569 ++++++++++++++++++