+++++++++++++++++++++++++++++++++++++++++++++++++ From: Larry S Cahoon Subject: [Elecraft] Using Spectrogram HI Gang, I have seen the following type of statement several times in regards to using Spectrogram. >I have used Spectrogram several times to set the filters, but never had >been quite satisfied with the results. It's a bit hard to get it just right >when using an on-air signal. So I thought I'd toss into the mix what I do with it. I've had very satisfactory results. It does require the use of a second rig. So here is the process. 1. I have the sidetone on the K2 set at 700 Hz and the sidetone on the TS-570s set at 800 Hz. These are simply my preferences. Pick what you want. 2. Hook up Spectrogram to the TS-570s. 3. Transmit from the K2 into a dummy load and zero beat the signal on the TS-570s. This is real easy with Spectrogram hooked up to the TS-570s as I only have to match the 800 Hz signal. So even if the VFO calibration is off a bit on either rig you get the right setting. Plus, I'm not going to be off by 50 Hz because I can't hear the tone just right. I've always had problems with the OHR instructions to adjust to a nice mellow 800 hz tone. 4. Now switch over so that Spectrogram is hooked up to the K2. Don't change the frequency setting on either rig. 5. Transmit from the TS-570s into a dummy load and proceed to adjust the filters on the K2 so that they are centered on the 700 Hz signal. The only critical feature is that you have to be sure that the offset on the reference rig, in my case the TS 570s is correct. I used the same basic idea last week to readjust the offset and BFO on my OHR 100A and am much happier with the results. You can use the idea to adjust the DSW rigs as well. If fact I suspect there are few rigs that you could not improve the alignment of with Spectrogram. The key in setting the offset/filters or whatever is that when you are done you should be able to fix the VFO on both rigs and get the correct tone in each rig when transmitting from the other rig. You check the sidetone with Spectrogram. 73/72 de Larry................WD3P in MD _____ ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ After aligning the cw filters on K2 s/n 1291, I switched the Spectrogram to "bar" display and learnt a lot in the process. You may like to try these easy experiments, and possibly suggest a few more. (a) Tune to WWV in CW mode, then press the K2 Spot key, then align the steady carrier peak of WWV to superimpose the peak from the internal spot audio signal. The K2 frequency display should show 10000.00, plus or minus any acceptable error. (b) Tune to a strong station that is repeated calling CQ. The attack and decay times of the AGC can be visually observed as the station transmits, waits for reply, transmits again. Change the AGC fast/slow setting and observe again - absolutely fascinating. (c) With the setup as in (b), watch the bar graph peaks of the stations replying to the calling station. You can see precisely how far they are off frequency. (d) Now cycle through the filter settings. The peak of the received signal will not move significantly. If it does move for a certain filter, you can read off exactly the correction needed to the BFO setting for that filter. By adjusting the settings in the Spectrogram "Scan Input" setup screen, one can optimise the frequency range displayed - a couple of hundred Hertz above and below the sidetone frequency is suggested. ............. +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 24 Feb 2001 07:52:43 -0600 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: Re: [Elecraft] K2 Filter Alignment question Ken: >I use the book default BFO settings as per page 85, and when using >spectrogram & noise bridge I see: > >My LSB low roll off is 624 Hz (very defined) >see: http://www.lotts.net/LSB.gif > >My USB Low roll off is about 722 Hz (poorly defined) >see: http://www.lotts.net/USB.gif > >*and* my SSB pass band ripple is awful (Absolutely nothing close to what Tom >Hammond N0SS presented in his paper) I've gotta admit that you seem to have a pretty deep dip in that one part of the ripple. This having been said, what you show is pretty much characteristic(!) of the K2 using its CW filters for SSB reception...! The filters in the 'BASIC' K2 were not really designed for SSB reception, as such (I believe) no significant efforts were made to 'flatten out' the ripple, which pretty much disappears when you narrow them down to more normal CW receiving bandwidths. If I were you, here's where I'd go from here: 1) Start up Spectrogram and connect your K2 to your PC. From your graphics, you're obviously having no problem doing this. 2) Pick USB... and inject some noise so you get a nice display on Spectrogram. 3) Go to CAL FIL, step to the BFO setting mode, and then adjust the BFO frequency to roll your USB response curve a bit further DOWN, so that the UPPER peak is inside the upper band limit marker. This should help to get rid of the high-pitched signals you are hearing AND it will also significantly improve your low-end response as well. You will now hear some low-end signals, but they should be pretty far down on the curve. 4) Save your settings. IF/WHEN you get the SSB OPTION board, perform the 2ND XFIL mod (available on the Elecraft web site) and I think you'll find that the ripple of the SSB filter is more closely representative of the graphics I presented in my instructions. Remember that the BFO settings in the manual are merely BALLPARK setting, designed to at least get your BFO set to the proper side of the filter and receiving something. >When I adjust the USB bfo to 400 Hz roll off, and I use FL1 (220) I end up >hearing beats on either side of zero. You need to sdjust such that the upper peak is INSIDER the band limit markers. Don't br too preoccupied with the lower limit at this time. >The ripple is by far the worst when using the FL1 "220". Not really... it's REVERSED however. Note that the deep dip you see around 1400Hz (USB) is the same one you see around 2kHz on LSB. The response curves are reversed because the BFO's been moved to the other side of the filter for reception of the opposite sideband. >Worth noting is that I have the front panel SSB option parts installed as >per step two on page 27 of the manual. Remove C180 from the K2 RF board. This alone may help to flatten things out a bit. When/if you perform the 2nd Xfil Mod, it'll be removed in that procedure anyway. Good luck, and let me know how things go. 73 - Tom Hammond N0SS - ---+++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 24 Feb 2001 13:20:35 -0600 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: RE: [Elecraft] K2 Filter Alignment question Ken: >The received signal (from a signal generator) frequency shifts when I go >from normal to reverse (the displayed K2 freq does not change though) > >============= START DETAILS ============= > >CW "normal" starts (from a reset "456") with: >bF1t 110 <4913.12 KHz> >bF2 110 <4913.12 KHz> >bF3 110 <4913.12 KHz> >bF4 110 <4913.12 KHz> >and it's pass band looks real close to acceptable (all is well). > >CW "reverse" starts (from a reset "456") with: >bF1t 169 <4914.88 KHz> >bF2 153 <4914.36 KHz> >bF3 151 <4914.30 KHz> >bF4 148 <4914.21 KHz> >The pass band needs some attention (it is shifted too low especially at >bF1t) *and* it shifts the received signal by about 260 Hz (not on the >display). The changing values of the BFO when in "REVERSE" is becaus, when the filter is 'narrowed ' down, it's only narrowed from one side (rather than from both sides). As a result, the REVERSE BFO, being on the the side that moves during narrowing, must be shifted in order to follow the center of the passband as the filter becomes more narrow. This does not occur when in NORMAL filter mode, so the BFO can (usually) be left (relatively) fixed, even when going from wide to narrow bandwidths. >============= END DETAILS ============= > >The pass bands are similar in size but shifted, so if I play around and >shift them so they are nearly the same, I still get a best case freq shift >of about 140 Hz (not on the display). > >More over, I have gotten into situations as I fiddle with the BFO frequency >in CAL FIL where the received signal (from a signal generator) will shift up >to about 100 Hz as I change between bF1, 2, 3 & 4 using CW normal. A "456" >reset always fixes this anomaly but I feel like a dog chasing his tail at >times with the received freq vs. the BFO. Successive attempts to force the individual tones to be identical have produced similar results for me. In general terms, you should be able to get the tonal difference to +/-20Hz (worst case), and often either dead on or +/-10Hz. Hope this helps. 73 - Tom Hammond N0SS +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 24 Feb 2001 17:04:45 -0500 From: "Don Wilhelm" Subject: Re: [Elecraft] K2 Filter Alignment question Ken, I see Tom Hammond is providing you with much sage advice, so I'll not add much more; EXCEPT When I first looked at my wide filters with Spectogram, I was quite surprised to see a big hole in the middle - toward the low frequency end on USB and the high frequency end on LSB. Such an occurance is not unexpected with the variable filter since it is optimized for narrow filters, and the degree of ripple may be different for each K2. I found too that at filter settings above 1.9, I actually developed a spur on the bandpass My solution was to narrow the filter until Spectogram showed a filter width that I thought might be usable without regard for the K2 display of the filter width - I found the width to be much wider than the displayed numbers. After setting a reasonable filter width - then adjust the BFO to place the audio response where you want it - if you prefer to chop off the low end you can do that, or reduce the high end of the audio leaving the lows - all that is your choice - try several settings to see which suits your ears and allows you to easily tune SSB signals. I get along just fine with the low frequency corner of the filter set at 500 to 600 Hz, you might use that as a starting point and vary from there to suit your own sense of 'what sounds right'. 73, Don Wilhelm -Chapel Hill, NC W3FPR home page: http://www.w3fpr.webprovider.com QRP-L # 485 K2 SN 0020 mailto: w3fpr at arrl.net +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 24 Feb 2001 17:59:50 EST From: LHLOUSEK at aol.com Subject: RE: [Elecraft] K2 Filter Alignment question Hi Ken, I had some confusion on the way the filters worked when I first started looking at it and Tom N0SS worked patiently and diligently to get me straightened out on it. Here's my take on it... As you know, when you change from CW to CW-reverse you are operating on the other side of the IF. If the CW note is perfectly beat with SPOT when you switch between CW and CW-rev the tone of the note could shift 10 Hz to 20 Hz or so do to the granularity in the K2's counters and VCO. If you are not perfectly beat with SPOT, say 50 Hz off, when you switch to CW-rev the tone you hear will change by that amount in the opposite direction giving a total change in tone of a 100 Hz in this example. You may notice that when you switch between CW and CW-rev (or LSSB and USSB) the sound of the noise you hear (in the absence of signal) changes even though you've done your best to set up the filters to be the same. This is because the passband of the filter is not symmetric and when you change to the other side of the IF its audio image is reversed. What was the high cut becomes the low cut and vice versa. When the IF frequency changes when switching between filter settings or CW and CW-rev (or USSB and LSSB) the K2's firmware corrects for that so that the VFO still accurately displays the working frequency (and in the case of CW-rev properly accounts for the receive offset). From the operators perspective, the only thing that changes is where the image of filter is in the audio frequencies. Luigi W7DZN +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 24 Feb 2001 19:07:10 EST From: n5ib at juno.com Subject: Re: [Elecraft] K2 Filter Alignment question On Sat, 24 Feb 2001 15:54:09 -0700 "Ken Lotts" writes: >L34 had some influence in the big picture as well.. I could *not* >hear the >"peak" when twiddling L34 but I could *see* it when using spectrogram. > It was not very pronounced. Hi Ken, +That* is a *good* idea! I used a meter, my "good" ear, and a scope. But I'll bet the big-screen spectrogram display would be just the ticket. I think I'll try that next time I lift the lid (ooops, sorry Eric, top cover) to fiddle. 72 Jim N5IB K-2 #1233 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 24 Feb 2001 16:27:11 -0800 From: "Ron D'Eau Claire" Subject: RE: [Elecraft] K2 Filter Alignment question Ken, AA7JC, wrote: "I think I have everything within 20Hz or so....L34 had some influence in the big picture as well.. I could *not* hear the "peak" when twiddling L34 but I could *see* it when using spectrogram. It was not very pronounced. Please advise if L34 should have given me a better peak." 20 or 30 Hz is about the best you will ever see, due to the resolution of the Digital-Analog Converters (DACs) used in the K2. Sounds like you are as 'right-on' as you can be. That's also normal for L34. It is very "subtle" peak. Gary from Elecraft published a procedure for measuring the signal on a DVM to find the peak. I think it's in the "Builder's Resource" area of www.elecraft.com. Ron AC7AC K2 # 1289 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Wed, 21 Mar 2001 00:22:52 -0000 From: "Ed Juge" Subject: [Elecraft] Question about K2 filter tuning... I finally tried Spectogram last night. Wow... what a simple way to tune filters. However... my patterns on SSB didn't look much like what's in the book. There were three peaks in most of the patterns but the valleys in between were 3 or 4 times as deep as shown in the docs. Is this meaningful?? At one time, I had changed the #1 OPT filter to show 2.4 bandwidth. The audio sounded terrible... very tinney, no matter where I set the BFO. When I turned it back to OPT, it sounded good again. Is this normal? Can anyone tell me why this happens? Thanks... Ed, W5EJ +++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Tue, 20 Mar 2001 21:32:33 -0600 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: Re: [Elecraft] Question about K2 filter tuning... Hi Ed: >However... my patterns on SSB didn't look much like what's in the book. >There were three peaks in most of the patterns but the valleys in between >were 3 or 4 times as deep as shown in the docs. Is this meaningful?? Probably not... it depends upon how hard you were driving the sound card. With more drive, the valleys appear to be smaller. And... if you have installed the SSB option, once you perform the 2nd Xfil SSB mod (available at the Elecraft site, a snap to install), much of the valleys will disappear anyway... at least those from the SSB filter anyway. The valleys are 'normal' when receiving thru the wider CW filter settings. >At one time, I had changed the #1 OPT filter to show 2.4 bandwidth. The >audio sounded terrible... very tinney, no matter where I set the BFO. When >I turned it back to OPT, it sounded good again. Is this normal? Can anyone >tell me why this happens? Nope. Sorry. 73 - Tom Hammond N0SS ++++++++++++++++++++++++++++++++++++++++++++++ Date: Mon, 10 Dec 2001 19:41:56 -0600 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: Re: [Elecraft] K1 Filter Alignment using Spectrogram Tom: >... my question is do I have to do anything special that is different with >the K1 vs the K2 as far as the instructions are written. After giving it some considered thought, I suspect that you can do as well with the K1 if you follow the instructions in the K1 manual. It's been a while since I aligned my K1 with Spectrogram, but I think I can still recall the basics. If you decide to try Spectrogram anyway, here's how I think I'd suggest you try doing it: Set-up Spectrogram on your PC as instructed in the doc file. Be sure you have the Spectrogram marker set to the sidetone frequency you wish to use. Connect your PC to the Speaker or Phones output of the K1. If you have a broadband noise source: Connect it to the ANT jack on the K1, set the K1 to 40M and turn the noise source on. If you do not have a broadband noise source: Connect an outdoor antenna to the ANY jack of the K1. Select a band with some background noise, but NO SIGNALS. You want a band where there's light-to-moderate atmoshperic noise, but no significant pops or crashes. 20M may be a decent choice, try 40M if 20M is dead. Adjust the AF GAIN control to produce at least a 20dB (off the baseline) signal deflection on Spectrogram. If you're using the standard 0 to -60dB scale, try to obtain something around -40dB, but even a -50dB display will be usable. Note that excessive AF GAIN can cause unwanted spurious 'blips' (other than a couple which may show up at or very near the 0 Hz indiactor (far left edge of display)to begin to show up on the Spectrogram display. If this occurs, reduce the AF GAIN level until none (or only the blips at the far left edge) are now visible. BFO Alignment The K1 three crystal filter bandwidths are selected by holding the XFIL switch. Hold XFIL until you see FL3 on the LCD. Note: FL1, 2, and 3 are pre-set to bandwidths of about 800, 400, and 250 Hz. They can be set up differently using FLx (page 53). Select the narrowest of the available filter bandwidths. Locate the BFO trimmer, C20, near the front edge of the RF board. Adjust C20 for to bring the peak of the displayed noise dead on (to straddle) the sidetone marker. Tap MENU twice to return to normal operation. Hope this helps. And PLEASE let me know if it works as stated. If so, I'll make a PDF for others to use if they should choose to do so. 73 - Tom N0SS ++++++++++++++++++++ Date: Thu, 13 Dec 2001 16:28:14 -0600 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: [Elecraft] Finally - Docs for aligning K1 CW filters with Spectrogram Hi folks: Well, I managed to hold out longer than I thought I could against preparing a set of K1 CW filter alignment docs, for those who wished to use Spectrogram for the alignment. Tom Mihelich (K8TDM) finally pushed me over the edge when he asked if the steps on the K2 (Spectrogram) alignment docs would work for the K1, and of course, most of them won't, because the K1 filter alignment is a LOT more simple than that of the K2. So, with Tom's able proofing assistance, I now have a set of K1 alignment docs ready for those of you who might be interested. I am submitting them to Elecraft with the hope that they will (eventually) appear on the Elecraft web site. But Eric's pretty busy these days, and I'm sure it'll be a while before he finds time to 1) review the docs, and 2) add them to the web site, should he choose to do so. For the time being, I'm gonna place the docs (a 414kB PDF) on my web site at: http://home.earthlink.net/~n0ss/k1_cw_alignment_using_spectrogram_v1.pdf NOTE: 1) If your e-mail browser wants to line-wrap the URL above, and when you click on it (to move it to your browser) you get only part of it transferred to your browser, FIRST - click on [STOP] on your browser (so it won't beat itself to death, telling you you've got a bad URL), and then return to your e-mail and cut 'n paste the 2nd line of the address to the end of that which is already inn the ADDRESS field of your browner. THEN HIT . 2) Should you decide to hand-enter the URL into your browser, be aware that there are several UNDERLINE characters in the filename which may be obscured by the link underline. If you see a blank space in the filename, ASSUME that it's filled with an underline character. Also, remember that in the URL, my callsign contains a ZERO, not an 'oh', so please enter it as such when you type it into your browser. If you've not visited my web site before, you might want to take a look around http://home.earthlink.net/~n0ss/ on both the main page AND the DOWNLOAD page as well. It's not much, but you never know... I may have something on the DOWNLOAD page you might be able to use. 73 to all, and thanks Tom! Tom Hammond N0SS +++++++++++++++++++ Date: Fri, 31 May 2002 22:45:52 -0400 From: frastephen at comcast.net To: elecraft at mailman.qth.net Subject: [Elecraft] Usnig Spectrogram for Zero-Beating WWV, Frequency Calibration, and Filters K2 #2644 continues to be aligned! There is serious lightning here in the Philly area right now, so #2644 is disconnected from everything! Hence, a good time to send along some insights I had this evening. Maybe this is old hat to everybody, but I am pleased I figured this out and maybe it might help someone else. Well, I wanted to calibrate the frequency of the VFO and align the filters. So I downloaded Spectrogram (and registered it! I used to be a programmer in a previous life and I liked to eat, too!) and used it with N0SS's wonderful instructions from the Elecraft web site for aligning the filters. What a really great program (although, I wish the audio sent into the MIC was also patched out to the speakers so you could listen to the audio as well as watch it!). I then realized I had not done anything to calibrate the frequency of the VFO! Oops! Well, then I got an interesting idea ... why not use Spectrogram to get a really close zero-beat of WWV? You might remember that WWV sends precision audio tones of various frequencies along with a 1000Hz "tick" each second (details on their web site). So I hooked it back up and Spectrogram gave a really great display of these various tones and I was able to get a really great zero-beat of WWV, which I used to ultimately do the adjustment to C22 for the 4 MHz Oscillator calibration (using the instructions from the Elecraft website). It took a few iterations of the procedure (I believe, three), but in the end I believe I have the frequency calibrated to within 10 Hz (theoretically, I see from the Specifications on p. 5 of the manual that the VFO accuracy is +/- 30 Hz). Not bad!!!! Probably as good as you can get. Then I used N0SS's instructions to do the alignment of the filters. Then I got another insight! Why not use Spectrogram to verify and make fine tuning adjustments to the filters with a strong signal on a known frequency.... So back to Spectrogram and I used it to monitor the audio from a certain signal at 7047.5 kHz. W1AW was blasting away, as usual here. With the frequency display calibrated I could tune it in right on the money (although I have seen some posts about W1AW not being exactly WWV when it comes to frequency stability -- never noticed it myself). So I used that steady signal to fine tune the filters exactly where I wanted them in the audio spectrum across all four filters (8 actually, counting the CW RV settings). Nice strong peaks at 600 Hz right across the board without tuning the VFO -- a steady tone across all the filters. Bottom Lines: I got everything aligned quite well! I learned a lot about how the K2 VFO and filters work! I learned a lot about Spectrogram! A worthy evening, if I do say so myself! Fun and educational! Well, the lightning has passed .... Peace and 73, Stephen W3SMK K2 #2644 ++++++++++++++++++ Reply-To: From: "Ron D'Eau Claire" To: , Subject: RE: [Elecraft] Spectrogram Date: Mon, 12 Aug 2002 12:06:06 -0700 Initially I used background noise when I adjusted my K2 filters. Since then I have used a noise generator cobbled up on a bit of perf board which resulted in a much better display. A nice big S9+ noise lets you see further down the slope of the filter, especially when checking out the crystal filter grounding mods. Turn OFF the AGC on the K2. Mike VP8NO #1400 ------------------------------------- Good point! Also, as John KI6WX pointed out to me a while back, if you want to take a critical look at the bandpass, be aware that the capacitors in the headphone output will cause the response to roll off at the higher frequencies if you plug your sound card directly into the phones jack. This produces a very definite downward slope when looking at the 2 kHz bandpass of the SSB filter. I plug my computer into the auxiliary speaker jack on the back of the K2 and use the "line input" on my sound card to avoid this effect. If one wants to use the phones jack, John recommends putting a 10-ohm resistor to ground across the headphone output After aligning the filters on a good wide band noise so you can see where your "sidetone" frequency appears in the bandpass, tune in a carrier (the birdie at 7 MHz is my choice) and see if there is any shift in frequency as you step through the filters or change from Normal to Reverse. If so, simply step through CAL FIL once again, moving the frequency of the BFO at each filter position away from the desired value and then putting it back to make the uP recalculate the BFO frequency. That takes care of any excessive shifting. Keep in mind that the BFO frequencies are not controlled accurately enough to guarantee that the beat note is be EXACTLY the same in each filter position. The DAC's used in the K2 limit the accuracy to 30 Hz, I believe it is. You might get closer, but it's pure luck of you do. I don't have any problem getting mine to within 20 Hz after a couple of tries. That's about all my "tone challenged" ears can detect anyway. Ron AC7AC K2 # 1289 +++++++++++++++++ Date: Mon, 28 Oct 2002 21:09:53 -0800 From: "John, KI6WX" Subject: Re: [Elecraft] Setup for Spectrogram 7.0 To: Elecraft Mailing List I'm still around, but have been away a lot of September and October (for bedtime reading, the adventures on my rafting trip down the Colorado River is posted at http://home.pacbell.net/johngreb/grandcanyon.html , but no ham radio on this trip). I have no further trips planned for the next few months, so I should be able to catch up on many backlogged projects. An updated tutorial on using Spectrogram 7 will have to wait a few weeks; I still need to finish building the field test version of the KAT100 and report any strange results back to Elecraft. The new version of Spectrogram has a lot more flexibility for making filter measurements on the K2 and I want to fully explain those. -John KI6WX > John Grebenkemper, where are you? We need you. Now that I have 7.0 > downloaded, I would like to have some setup parameters and a tutorial on the > use of this updated version. Thanks Roy Morris W4WFB K2/100 #2225 ++++++++++++++++ From: Piet Meindertsma To: elecraft at mailman.qth.net Date: Wed, 30 Oct 2002 19:33:28 +0100 Subject: [Elecraft] SSB xtal replacements DHK2BOB wrote: >1. Using Spectrogram6 I still can't tell when L34 is aligned! I just left it >one and one half turns from the top! This is not a problem but just an >observation which I'm sure relates to my inability to detect the subtle >differences. I had AGC off and fiddled with the RF gain control. I thought >that the only difference I noticed was not in the "height" of the spike but a >slight change in the distance between the top of the spike and the "noise" >down below. That is, tuning L34 seemed to change the noise but not the spike. I installed Spectrogram on my notebook and on my desktop (with CRT monitor). On the CRT monitor it is indeed almost impossible to see differences when fiddling with L34. On my notbook however this is quite easy: the notebook is equipped with a TFT monitor and I can see the peak becoming just one pixel higher or lower when fiddling tuning L34. So with Spectrogram I could find the peak while monitoring the noise level. Works great. Piet, PE1OQX +++++++++++++++ Date: Mon, 25 Nov 2002 08:13:02 -0800 From: Vic Rosenthal Organization: Transparent Software To: TL Cc: Elecraft Reflector Subject: Re: [Elecraft] Using spectrogram - again TL wrote: > > How significant is it to have AGC turned on (fast/slow) or turned off during measurements with spectrogram? If you are using it to adjust the BFO frequencies, then it doesn't matter. If you are trying to compare the attenuation of the filters or check the ultimate rejection, etc., then you should turn the AGC OFF. Vic K2VCO ++++++++++++++ From: "Bob - AG5Q" To: "Bill Marvin KB9IV" , "K2" Subject: Re: [Elecraft] Where To Start?? Date: Fri, 29 Nov 2002 12:22:53 -0600 Hi Bill, The Spectrogram program version 5.1.7 is still available as freeware at this site: http://home.t-online.de/home/on6et/page7.htm (scan down the list of progs) It's a very neat little Windows program that is user friendly. You just need to connect from the headphone jack or the external speaker jack (with a stereo plug) to the Line or Mic input of your PC sound card. I connected the two-conductor cable right across my head phone terminals so the phones still work. Then plugged into the Mic connector and after adjusting the volume, it took right off. Spectrogram initially came up with the sensitivity set to 60db. At first my volume was too low and I didn't see anything until I set Spectrogram to 90 db. Then I got orientated and it was very easy to use after that. You can play around with the sample rate and the number of samples to control the horizontal resolution. I usually use 16 bits and 44K samples/sec. You do need a noise generator; there are schematics on the net. These are small circuits and construction is not critical (it's supposed to be noisy, right? :) I found Spectrogram to be a terrific program for aligning the filters and it provides insight into how the filters actually work. It's also good for checking the adjustment of the VFO frequency. You can see the signal from WWV and tune it right in (in CW mode) using a marker corresponding to your sidetone frequency. 73/ Bob - AG5Q +++++++++++++++++ From: "JAKidz" To: Date: Tue, 10 Dec 2002 21:47:23 -0800 Subject: [Elecraft] Spectrogram - Again Greetings: I was having a lot of trouble aligning my K2's crystal filters even though I was using the N0SS noise generator and Spectrogram. Starting from the K2 manual settings for the filters, lots of peaks of variable heights were appearing everywhere across the spectrum no matter the filter settings even though the radio seemed to be working. After a lot of fiddling, two things became apparent. I was setting both the N0SS noise generator and the K2 AF gain too high. Using the output pot to set the noise generator just above receiver noise on my headphones and then setting the K2 AF gain to minimum and increasing it slowly, the desired shapes began appearing. I'm guessing my computer's soundcard was clipping even at moderate audio levels and the FFT processing of the resulting "squared" waves in Spectrogram resulted in all those interesting peaks. John, K7JG and K2 #1932. ++++++++++++++++ Date: Wed, 11 Dec 2002 09:06:59 -0600 To: "JAKidz" , From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: Re: [Elecraft] Spectrogram - Again John, K7JG, wrote: >I was having a lot of trouble aligning my K2's crystal filters even though >I was using the N0SS noise generator and Spectrogram. Starting from the K2 >manual settings for the filters, lots of peaks of variable heights were >appearing everywhere across the spectrum no matter the filter settings even >though the radio seemed to be working. After a lot of fiddling, two things >became apparent. I was setting both the N0SS noise generator and the K2 AF >gain too high. Using the output pot to set the noise generator just above >receiver noise on my headphones and then setting the K2 AF gain to minimum >and increasing it slowly, the desired shapes began appearing. I'm guessing >my computer's soundcard was clipping even at moderate audio levels and the >FFT processing of the resulting "squared" waves in Spectrogram resulted in >all those interesting peaks. John makes an EXCELLENT point here... especially with regard to the use of the AF GAIN on the K2. While an output level pot on the noise generator may be a worthwhile addition for many (the generator DOES have quite a bit of output), the MOST important control when performing a CAL FIL alignment using noise (from ANY source) and Spectrogram is to NOT overdrive the input to your sound card. Overdriving the sound card input WILL (almost guaranteed) result in a bunch of spurious artifacts appearing on the Spectrogram display. Use of the 60dB scale, and adjusting your AF gain to the point that the peak signal level displayed on the screen doesn't exceed 40dB should work well. Adjusting the AF Gain to produce a 30dB trace may work even better. When using Spectrogram, NOTE that there is a slider bar along the RIGHT edge of the Spectrogram screen. This slider should be run to the top end of its range before you begin establishing an AF Gain level setting. Thanks for the reminder, John. 73, Tom Hammond N0SS ++++++++++++++++ Reply-To: From: "Ron D'Eau Claire" To: Subject: RE: [Elecraft] Spectrogram - Again Date: Wed, 11 Dec 2002 18:20:12 -0800 John makes a good point about doing a Spectrogram alignment AND it is useful when simply operating the K2 on the bands. While the noise generator is handy, there is NO need for it to set up the K2 BFO frequencies. "Band noise" on 80 is more than enough to do a perfect alignment job. The only advantage of the noise generator is to avoid any signals in the passband. You don't want any "real" signals in the passband while doing the filter alignment. But I've never found it difficult to find a 2 kHz chunk of band down around 80 or 160 where I can't hear signals. Also, turn the "RF GAIN" down! It even helps to turn the AGC OFF. Then set the Audio Gain about 1/4 the way up and control the signal output using the RF Gain control, not the audio gain control. The "bandpass" will then rise neatly out of the base line on the Spectrogram display. Be sure to plug the sound card into the AUX SPEAKER jack, NOT the Phones jack, if you want to see what the bandpass really looks like on SSB. The phones jack output will show a slope on the bandpass that is NOT really there. In fact, controlling the volume with the RF Gain is an excellent way to operate the K2 on the air, too. Running the RF Gain full up and using AGC is nice for roaming around the bands, but I much prefer to turn it off and adjust the level with the RF Gain when in a QSO. It makes the signals stand out much clearer above the background noise! That's because when the RF gain is full up and the AGC is on, the AGC keeps turning up the gain between code characters so the noise in the phones or speaker as always the SAME as the signal! Using "Slow" AGC helps, because that keeps the AGC from turning up the gain so quickly between dits and dahs, but even a brief space in sending allows the AGC time to pump the noise up to full volume. Personally, I'd rather listen to the signal and let the noise stay down in the background where it belongs... Ron AC7AC K2 # 1289 ++++++++++++++++ Reply-To: From: "J. Edward (Ed) Muns" To: , Subject: RE: [Elecraft] Spectrogram - Again Date: Wed, 11 Dec 2002 19:46:50 -0800 One other setup detail is to turn off the audio filter (if you have that option) using the slide switch on the PC board. If this switch is on, the low pass hiss filter is on, even if AF1/AF2/OFF is 'OFF' on the audio filter menu. Both this and using the headphone jack (with it's own low pass hiss filter) will really roll off the Spectrogram plots. I spent months troubleshooting my KSB2, trying to get a "flat" response, only to find out I had both low pass filters activated. 73, Ed - W0YK +++++++++++++++ Date: Sun, 18 May 2003 20:42:58 -0700 (PDT) From: Scott T Emery To: elecraft at mailman.qth.net Subject: [Elecraft] Spectrogram-like program debugging vindicated... After a suggestion that I attenuate the RF input and attenuate the AF output I did both and discovered that the filter spectrogram outputs looked *a lot* better. Thanks to those who made suggestions. Scott Emery AD6RY ++++++++++++++++++ Date: Thu, 05 Jun 2003 07:54:24 -0400 From: "David A. Belsley" To: Joseph Wiedeman , elecraft Subject: Re: [Elecraft] Spectrogram Joseph: On the Elecraft web site, under Builder's Resources, there is an article by Tom Hammond listed under "Spectrogram Filter Alignment Articles". It is excellent and will have you going in no time. best wishes, dave belsley, w1euy --On Thursday, June 5, 2003 12:43 AM -0400 Joseph Wiedeman wrote: > How can I get instructions on the step by step procedure to use > spectrogram. I would buy it if I had some way to evaluate it. You can > download it but I haven't a clue on how it works. > Does it come with instructions that only an engineer will understand??? > or none at all. > ++++++++++++++++ From: "Roy Vickers" To: Subject: RE: [Elecraft] Spectrogram Date: Thu, 5 Jun 2003 07:06:00 -0500 At one time they had an eval copy and I used that to align my filters. When I upgraded my mcu I had to realign my filters and the eval copy had expired and I was reluctant to fork over the $$$ for a real copy. But, I finally did and have found that I use it all the time for zero beating cw and experimenting with filter settings. It's well worth the bucks and easy to use. Good Luck Roy, (KF5YU) +++++++++++++++++ Date: Thu, 5 Jun 2003 09:58:12 -0500 From: "Kenneth E. Harker" To: Joseph Wiedeman Cc: elecraft Subject: Re: [Elecraft] Spectrogram On Thu, Jun 05, 2003 at 12:43:14AM -0400, Joseph Wiedeman wrote: > How can I get instructions on the step by step procedure to use spectrogram. > I would buy it if I had some way to evaluate it. You can download it but > I haven't a clue on how it works. Does it come with instructions that only > an engineer will understand??? or none at all. There are several articles on the elecraft.com web site that can help you. The current version of Spectrogram is 7.2. Some of the articles were written using earlier versions of the software, so dialog boxes and such may have changed slightly from the screenshots in the articles: http://www.elecraft.com/Apps/new_fil_docs/k2_cw_gram.pdf http://www.elecraft.com/Apps/new_fil_docs/k2_ssb_gram.pdf http://www.elecraft.com/Apps/new_fil_docs/filter_apps.htm -- ++++++++++++++++++