[Elecraft] Tuners and spurious responses

[Wes Stewart]

Generally speaking, with components of the same Q, an L network will be the 
lowest loss solution to any matching problem. The network Q is set only by the 
two impedances to be matched. (I did once contrived a situation where matching 
the very low R, high C gate impedance of an FET could be done with lower loss 
using a series pair of L networks, but this is unusual.)

In correspondence with Dean Straw 20+ years ago when we were prepping my article 
on ladderline, I pointed out to him that the problem with T networks with three 
variables was that there are an infinite number of solutions, with only one 
giving the lowest loss and some giving huge losses, with the operator being 
clueless. I believe the same would be true with the Match Box.   (By what I'm 
sure must be coincidence, it was shortly after that an ARRL favored author wrote 
article on tuner losses.)

Any rejection due to an antenna matching system is just serendipity and 
depending on it for that purpose is foolhardy, IMHO, of course.

To pick one nit with Scott, a pi-network can be high pass and a tee-network can 
be low pass.

Wes  N7WS



On 12/22/2018 7:21 PM, K9MA wrote:
> While I seriously doubt any rejection of nearby frequencies by a tuner is 
> likely to have any effect on the K3 receiver,
> I'd point out that any tuner configuration other than an L network can provide 
> a match over a wide range of Q. One generally tries to tune them for minimum Q 
> to minimize losses. If tuned to a high Q, however, both the T and pi networks 
> generally will provide some rejection of adjacent frequencies. Just how much 
> is impossible to predict, unless you know just how the antenna impedance 
> varies with frequency. At far removed frequencies, of course, a T does act 
> like a high pass, and a pi like a low pass, but in neither case do they match 
> the antenna to 50 Ohms, unless it happens to actually be 50 Ohms at some 
> frequency.
>
> That said, is suppose it IS possible that a tuner/antenna combination just 
> happened to have a deep null right on the frequency of a nearby broadcast 
> station which was causing intermodulation. If that were the case, a more 
> reliable solution would be a trap or stub.
>
> 73,
> Scott K9MA
>
>
> On 12/22/2018 08:52, Bob McGraw K4TAX wrote:
>> The Johnson Matchbox configuration is indeed a banpass filter meaning it 
>> attenuates both above and below the frequency to which it is tuned.    I use 
>> mine at Field Day to provide attenuation to stations operating both above and 
>> below the band being used.  The amount of attenuation does vary as it is not 
>> symmetrical in nature.
>>
>> Probably you were using a less than optimum balun which had little common 
>> mode rejection or poor balance.   The best way to check the two 
>> configurations is to measure the current in each leg of the balanced feed 
>> line.  Many baluns do not do a good job or making a "balanced" feed.    The 
>> work of DJ0IP  {see his website} has a lot of information from real field 
>> measurements on baluns, good ones and bad ones.
>>
>> 73
>>
>> Bob, K4TAX
>>
>> On 12/21/2018 11:03 PM, Victor Rosenthal 4X6GP wrote:
>>> I just replaced my single-ended T-network tuner plus balun with a massive 
>>> old Johnson Matchbox. It is very selective, unlike the T network. It is as 
>>> if there is a sharp bandpass filter between the antenna and the rig.
>>>
>>> This may be totally imaginary, and there's no easy way to A/B test it, but 
>>> it seems as though the K3 sounds "cleaner" in some sense.
>>> Could it be that since the mixer sees a much narrower spectrum, there are 
>>> fewer spurious responses?
>>>
>>> Do those of you who use bandpass filters for SO2R or multi-transmitter 
>>> contesting notice such an effect?
>>>
>>> It also seems that the better balance (my antenna system is a dipole fed 
>>> with balanced line) has reduced RF in the shack and possibly local noise 
>>> pickup.
>>>