Andy,
I'm working on a detailed reply, but not yet done. The main practical
problem with the model will be in actually attaining the
phase/current/voltage relationships which are set "by magic" in the model.
There also are some gotchas in the design. Extensibility to the bottom of
the broadcast band is questionable, to say the least. I have all of Pete's
reservations as well.
More later, 73, Guy.
----- Original Message -----
From: "Pete Smith" <n4zr@contesting.com>
To: "Andy Ikin" <andrew.ikin@btopenworld.com>; <antennaware@contesting.com>
Sent: Tuesday, May 26, 2009 5:59 AM
Subject: Re: [Antennaware] Dual output K9AY
> Thanks for the quick reply, Andy. I am no theoretician, but I recall
> a discussion of the K9AY loop in which one well-qualified engineer
> asserted that it could/should be viewed as a close-spaced array of
> two short near-verticals, phased by the wire between them. If this
> is in fact the mode of operation, then feeding each "vertical"
> separately, while retaining the top connection between them, seems
> like a real muddle from the standpoint of phasing. Modeling two
> side-by-side loops also wouldn't really be faithful to the real-world
> antenna, would it?
>
> 73, Pete
>
> At 06:49 PM 5/25/2009, Andy Ikin wrote:
>>Pete,
>>
>>Thank you for looking into the modelling feature.
>>
>>The reason for trying to model the Dual output was to determine the
>>phase relationship verses freq. required to increase the rear null
>>by subtracting some of the front lobe from the rear. I did some work
>>on this in May 2001, but ran into problems with the phasing. Misek
>>uses this technique on his Steerable Wave antenna in his Beverage
>>Handbook. However his Phasing system is not wide-band.
>>
>>The test set-up was to use 2 x 9:1z xmfr at each ends of the loop at
>>the junction of the ground wire. Both xmfr outputs connect
>>to Phasing box. One channel is a variable 0-165ns delayline
>>(terminated 50R). The other channel has the amplitude control. The
>>phasing box provides a constant 50 Ohm input for any phase or
>>amplitude setting. Also the phase delay is constant for variations in
>>Freq.
>>
>>First I wired the K9AY xmfrs so that the Phasing box would combine
>>in-phase plus the 0-165ns delayline. Result no increase in gain.
>>
>>Next I re-wired K9AY xmfrs so that the Phasing box would combine
>>anti-phase plus the 0-165ns delayline. The gain dropped by approx.
>>6dB, however, the rear null could be simply adjusted to upto -50dB
>>with just tweaking the Phase and amplitude control. Across the MW
>>band the average delay was 65ns, so most of the null adjustment was
>>with the amplitude balance. The max null was far higher than using
>>remote variable termination. The results were much better than I
>>had in 2001 with the improvement in the phasing system.
>>
>>Btw, this technique doesn't work with RC phasers e.g. MFJ and Dxeng
>>except over a narrow bandwidth i.e. the knob fiddling becomes
>>tedious with changes with freq.
>>
>>Getting back to the model, my thoughts are to simulate the Dual
>>output K9AY is to model 2 K9AY close together and take the source
>>from each antenna.
>>
>>73
>>
>>Andrew
>>
>>
>>
>>
>>----- Original Message ----- From: "Pete Smith" <n4zr@contesting.com>
>>To: "Andy Ikin" <andrew.ikin@btopenworld.com>;
>><antennaware@contesting.com>
>>Sent: Monday, May 25, 2009 10:19 PM
>>Subject: Re: [Antennaware] Dual output K9AY
>>
>>
>>>Andrew, I am very suspicious. Running this model in Multinec
>>>(NEC2) and 4NEC2, it appears to be very sensitive to the value of
>>>the assumed ground resistance. At 200 ohms, it gives me a peak
>>>gain -18.4 versus average of -26. At 100 ohms the figures are
>>>-15.4 and -23.2, and at 50 ohms -12.4 and -20.2. At 5 ohms the
>>>values are -3.1 vs -11. The reference single-feed version gives -26
>>>vs -34.
>>>
>>>I ran the same dual feed arrangement with another model of the
>>>loop, as modified by W7EL. He uses a pair of 50-foot radials
>>>perpendicular to the plane of the loop in lieu of a direct ground
>>>connection, Real-Sommerfeld ground and no series resistance to
>>>ground. That model gives quite similar results, -9.8 dB vs
>>>-17.95. The reference single-feed version gives -25.9 vs -33.94.
>>>
>>>I can't explain the results, but I fortunately, it should be fairly
>>>easy to test. an 8 to 16 dB difference in signal strengths between
>>>the reference loop and the dual-feed design should really be
>>>obvious. I'll be interested to see how your tests come out.
>>>
>>>73, Pete N4ZR
>>>
>>>At 04:57 AM 5/25/2009, you wrote:
>>>>Pete, Guy, Gary and Terry;
>>>>
>>>>Please see attached files. The original K9AY model ( K9AY _ALT.EZ
>>>>) is, I believe one from Terry but with the 200 Ohm res. in the
>>>>ground wire as per a model from Gary.
>>>>The Dual opt K9AY.EZ is with the 390 Ohm load replace with a Source.
>>>>
>>>>Btw, I am going to test it this morning using a varible delayline
>>>>Phasing Unit . If the model is right then the correct phasing from
>>>>500kHz to 2MHz is a 54ns delayline!!!!
>>>>
>>>>73
>>>>
>>>>Andrew (G8LUG not active )
>>>>
>>>>
>>>>
>>>>----- Original Message ----- From: "Pete Smith" <n4zr@contesting.com>
>>>>To: "Andy Ikin" <andrew.ikin@btopenworld.com>
>>>>Sent: Sunday, May 24, 2009 10:08 PM
>>>>Subject: Re: [Antennaware] Dual output K9AY
>>>>
>>>>
>>>>>Can you put the model file somewhere we can download it? EZNEC
>>>>>or NEC-2 or? This potentially looks very interesting, either
>>>>>with switchable phasing or maybe with a phaser in the shack, like
>>>>>the DX Engineering or MFJ units.
>>>>>
>>>>>73, Pete N4ZR
>>>>>
>>>>>At 04:53 PM 5/24/2009, you wrote:
>>>>>>Hello guys,
>>>>>>
>>>>>>
>>>>>>I have been trying to model a Dual Output K9AY I.e. taking
>>>>>>another output from where the termination is normally placed.
>>>>>>Then combine the two equal amplitude outputs with a small phased
>>>>>>shift:
>>>>>>
>>>>>>10 deg. at 500kHz; 20 deg.at 1MHz; 30 deg. at 1.5MHz, 37 deg. at
>>>>>>1.85MHz.
>>>>>>
>>>>>>This provides the typical K9AY pattern. However, the gain 8dB
>>>>>>higher at 1.85MHz and 22dB higher at 0.5MHz compared to the K9AY.
>>>>>>
>>>>>>I have yet to verify the model. However, does the model change
>>>>>>stack-up??
>>>>>>
>>>>>>73
>>>>>>
>>>>>>Andrew Ikin
>>>>>>_______________________________________________
>>>>>>Antennaware mailing list
>>>>>>Antennaware@contesting.com
>>>>>>http://lists.contesting.com/mailman/listinfo/antennaware
>>>>
>>>>
>>>
>>
>
> _______________________________________________
> Antennaware mailing list
> Antennaware@contesting.com
> http://lists.contesting.com/mailman/listinfo/antennaware
>
_______________________________________________
Antennaware mailing list
Antennaware@contesting.com
http://lists.contesting.com/mailman/listinfo/antennaware
|