+++++++++++++++++++ See also 17M SW Interference Mod ++++++++++++++++++++ Date: Mon, 05 Nov 2001 11:34:34 -0800 From: Wayne Burdick Subject: [Elecraft] Re: K2 receiver images Dave Johnson wrote: > >From time to time, I suffer from 15 MHz broadcast band breakthrough when > listening on 24.9 MHz. On checking the image rejection with my signal > generator, I get an image rejection of approximately 50 dB on 24.9... > > Has anyone modified their K2 (without losing too much sensitivity in the > process) to reduce this image level and what figures does the mod improve > the image rejection to? > > Dave, G4AON Hi Dave, The VCO is below the band on 10, 12, and 15 meters. So all image rejection has to come from the band-pass filter on these bands. We have tested a bit "stiffer" 10/12 meter band-pass filter that should provide some 10 dB better image rejection. Just change both 220 pF caps to 330 pF (approximately), and change the 2 pF cap to 1 pF. This new filter shows around 0.5 to 1 dB more loss than the original at midband, but the rig is still able to reach full power output. In addition to the better image rejection, the -3dB bandwidth decreases by about 50% on both bands (from 1.8 to 0.8 MHz on 12 m, and from 2.5 to 1.2 MHz on 10 m). Similar improvements are possible on other bands, if required. In the case of both the 15/17 and 20/30 meter band-pass filters, change the 330 pF caps to 470 pF (approx.) and the 3.3 pF to 2 or 2.2 pF. Again the compromise is similar; about 0.5-1 dB additional mid-band loss, but better rejection and a narrower passband. 73, Wayne N6KR ++++++++++++++++++++++ Date: Tue, 6 Nov 2001 10:01:00 +0100 From: "Sverre Holm \(LA3ZA\)" Subject: RE: [Elecraft] Re: K2 receiver images An alternative way to get better image rejection could perhaps be to use a trap for the 15 MHz band? A series tuned L-C network from the junction of C43 and C42 to ground and/or from the junction of C47/C48 to ground could do the trick in the same way that C104 and RFC7 today traps the IF frequency. The trap(s) would be inductive (I think) at the higher signal frequency, so some thought would have to go into finding the right L,C ratio in order not to influence the present BP-filter too much. - - Sverre, LA3ZA - ----------------- www.qsl.net/la3za +++++++++++++++++++ Date: Sun, 11 Nov 2001 20:14:04 -0800 From: Wayne Burdick Subject: [Elecraft] Re: Broadcast AM Harmonic on 15 Mtrs (K2)? Hi Lynn, What you're hearing is probably an image reponse from an extremely strong HF AM broadcast station, not a broadcast band AM station harmonic. I'll explain how such images are created and what you might be able to do about them. On 15 meters and above, all image rejection in the K2 is provided by the band-pass filters, since the VCO frequency is below the band itself. The TS440's VCO is probably above the band--a more typical situation for a multi-conversion rig, but not possible with the K2's present conversion scheme. In the case you mentioned, the K2's VCO is operating at about 21365 - 4915 = 16450 kHz. The 15-meter band is reached via the sum of the VCO and BFO (16450 + 4915 = 21365). But a high-power broadcast station at 11535 kHz could create an image response due to the difference between the VCO and BFO (i.e., 16450 - 4915 = 11535). It's then left to the 15/17 m band-pass filter to reject 11535 while passing 21365. The rejection of this frequency is probably in excess of 50 dB, but this might not be enough to eliminate an extremely strong image. We recently tested more aggressive band-pass filters on 15 through 10 meters, with good success. You can easily modify your K2 if you'd like improved image rejection on 15 m. Just change the 330-pF caps in the 15/17 m band-pass filter to 470 pF, and the 3.3 pF to 2 or 2.2 pF. Then re-align the rig on 15 and 17 meters (you can probably just peak both bands on receive--refer to the manual.) The result is an insignificant additional mid-band loss (about 0.5 dB), but better image rejection and a narrower passband. Let me know if this helps. You may need to monitor the image signal for a few days before doing the mod to see if it varies in strength from one day to the next. If the image really is at 11.5 MHz, it will probably be reasonably strong day and night, but may peak at some particular time in your location. 73, Wayne N6KR lbailey at aep.com wrote: > > Here at the office, I made a contact today on 15 mtrs w/my TS440S & when I > switched to my K2, I couldn't find the other station because there was a > strong AM station blanketing the frequency. I believe I was operating on > 21,365 + or -. I was able to go to other freqs on 15 mtrs. and copy on > either rig, but at this frequency, I encounter a strong carrier & AM sig. > I didn't even try to listen to the AM station since the K2 is not > optimized for AM. > > I'm thinking that it is a harmonic of a local AM broadcast station but why > would I hear it on my K2 and not on my TS440s? Has anyone else > encountered this? Is it a characteristic of a single-conversion rcvr? Thet +++++++++++++++++ Date: Mon, 12 Nov 2001 22:31:49 +0100 From: "Sverre Holm \(LA3ZA\)" Subject: RE: [Elecraft] Re: K2 receiver images Dave, The two different ways of improving image rejection: more narrow bandpass filters vs. traps have different pro's and con's. One issue is that the price you pay for getting 10 dB more image rejection at 15 MHz in the 24 MHz band is a reduction in bandwidth in the 10 MHz band. The bandwith goes from 2.5 to 1.2 MHz according to Wayne's simulation. Therefore this forces you to select which part of the 10 m band you are interested in when you tune the filter. Probably this could be avoided with a trap or two. - - Sverre, LA3ZA - ----------------- www.qsl.net/la3za ++++++++++++++++++ Date: Mon, 12 Nov 2001 15:07:03 -0800 From: Wayne Burdick Subject: [Elecraft] K2, 10 meters: 1.2 MHz too narrow? "Sverre Holm (LA3ZA)" wrote: > The two different ways of improving image rejection: more narrow > bandpass filters vs. traps have different pro's and con's. > > One issue is that the price you pay for getting 10 dB more image > rejection at 15 MHz in the 24 MHz band is a reduction in bandwidth in > the 10 MHz band. The bandwith goes from 2.5 to 1.2 MHz according to > Wayne's simulation. Therefore this forces you to select which part of > the 10 m band you are interested in when you tune the filter. Probably > this could be avoided with a trap or two. Hi Sverre, Is there much SSB activity above 29 MHz? If not, the reduction in 10 m bandwidth due to a tighther band-pass filter would not be a problem. (In fact it would also help prevent problems due to overpowered [illegal] 11-meter transmitters operating near by.) Of course if the K2 had FM capability, higher-band coverage on 10 m would be required. A trap in the image band is another interesting alternative. 73, Wayne N6KR ++++++++++++++++++ Date: Wed, 14 Nov 2001 21:47:28 +0100 From: "Sverre Holm \(LA3ZA\)" Subject: [Elecraft] Trap vs tighter BP for mirror rejection Wayne, You're right, your tighter bandpass filter with only 1.2 MHz bandwidth in the 10 m band is probably no problem with a CW/SSB rig - in fact for me running CW only, I have never been above the beacon band from 28.2-28.3 MHz. But, this discussion awakened the dormant engineer in me, so I made some traps and did some measurements. My test frequencies were 24940 kHz (middle of 12 m band) with mirror image at 15110 kHz (24940-2*4915). I made a trap from a 1 microH inductor in series with 100 pF in parallel with a small variable capacitor of some 10-20 pF. I tried several locations to connect it, but the maximum rejection with no influence on the desired signal was at the junction of C47 and C48. The output of the BP filter, junction of C42 and C43, did not give as much suppresion, and connection inside the filter (either end of C45) would attenuate the desired signal. The initial mirror rejection was 48 dB (very close to Dave Johnson's 50 dB), and it was improved by 7 dB when the inductor was a standard 1 microH inductor (pulled from an old video-recorder). The improvement increased to 11 dB when I made the inductor from 18 turns on a T37-6 toroid. The capacitance has to be tuned exactly for maximum rejection, so this schme works best for only a single mirror to be suppressed. The up-side is that it did not change sensitivity at all in the 12 or 10 m bands, as checked at the ends of the bands. I measured sensitivity by turning up the output from my Leader signal generator dB by dB until the second green LED in the S-meter would light up, and noted down the setting. Conclusion: It works as well as Wayne's suggested tightening of the 10/12 m bandpass filter. The trap does not tighten the passband of the 10 m band, but on the other hand it has to be tuned exactly and it contains one of the dreaded toroids. It also adds components while Wayne's modification consists of a replacement of existing components. So, if I had a problem with mirrors, I would ... not use the trap, but the tighter BP-filter, unless the 1.2 MHz 10 m bandwidth was a real concern. The only situation I can think of is if the K2 was to be used with a transverter for 6m or 2m > 10 m, but since I am not active on these bands, I can't really tell if >2 MHz of bandwidth is required then. - - Sverre - ----------------- www.qsl.net/la3za +++++++++++++++ Date: Wed, 21 Nov 2001 20:12:02 +0100 From: "Sverre Holm \(LA3ZA\)" Subject: [Elecraft] Trap for12 m image suppression I was encouraged by several responses to pursue the trap idea as a way to reduce the image problem in the 12 m band. The trap is now permantently installed in my K2, and the circuit and the experience with it has been written up on http://www.qsl.net/la3za/K2/K2_trap.htm Dave, G4AON, and myself are both hearing images at 24911 and 24932 kHz (from broadcast stations near 15031 and 15052 kHz), especially when 19 m band conditions are good and 12 m is not. I am curious to know if this is because of our Northern/Western European location or if it is also the case other places in the world, in particular in the US? - - Sverre - ----------------- www.qsl.net/la3za ++++++++++++++++++++