[CQ-Contest] Bandpass filter

Tom W8JI w8ji at w8ji.com
Wed Jul 4 11:21:12 PDT 2012

> Sorry Keith, the measured isolation between 15m TX and 20m RX on the C31XR
> with physical separation of only 18" is legitimate.

It could be much more than that, actually.

Maximum power transfer would occur when the optimum impedance on 15 meters 
for the 20 meter element is presented. The simple cheap test would be 
installing a T network tuner in line between the load and the 20M feedline, 
and adjust for maximum power into the load. I'd bet you have a whole lot 
more than 135 watts available.

The dummy load test presents 50 ohms to the 20 meter element. Most likely it 
wants a capacitance and higher resistance.
Your radio will look nothing like 50 ohms to the transmission line at 15 
meters while receiving on 20 meters, so the power will be not be the same as 
the dummy load. It can be anywhere from a really low power (but that's 
unlikely) to a much higher power than the dummy load indicates, because it 
is probably not matched to the antenna on 15 meters while on 20.

Not that any of this matters, because 135 watts indicates a real coupling 

I can alter the F/B ratio and gain of my interlaced Force 12 antenna for 15 
and 10 by changing termination on the unused feedline, or the unused 
feedline length!!! It not only has a good bit of power available on the 
unused feeder, antenna perfornance can be changed.

I wonder how many interlaced antennas are tested for this type of 
interaction, by varying termination impedance through a realistic range?  Or 
do designers just assume the unused feeder is opened at the antenna 

> combination. I found that Ice 419As with no stubs would not burn out under
> those conditions while running SO2R as a very novice contestor (A better
> SO2R operator would probably burn them up because they would transmit on
> both rigs more often than I was able to). Now that I am wiser I have
> invested in both stubs and higher overhead band pass filters.

I would not expect a filter of reasonable headroom to be damaged. A filter 
misterminates the line, it does not just dissipate all the energy it does 
not pass. I'm sure there are sweet spots in TL length between the antenna 
and filter that will maximize heating (which should be around minimum 
attenuation lengths) or minimize filter heating or filter input (antenna 
side) voltage at the 15M frequency.

I notice all sorts of things with stub articles, but I rarely see any that 
point out how stub location changes effectiveness. It appears they all 
assume the TL is matched at the fundamental, and also matched at all 
frequencies above and below the band being used. They also assume the 
amplifier output, radio, and the antenna all appear as a 50 ohm j0 sources 
or loads on all frequencies.

Another problem with stubs is bandwidth decreases as they are longer in 
fractions of a wavelength. A 3/4 wave stub is narrower than a 1/4 wave stub, 
and lossier. If you have one 1/4 section with a 10% error from frequency 
change, a stub 3/4 wave long has three times the error for the same 
frequency change. Each 1/4 wave section changes 10%, and there are three in 
series. :-)

I'm not saying they don't work, but simply pointing out I have not seen an 
accurate description of how they work or where to place them in the stuff I 
have read. I can imagine many cases where moving them a few feet along the 
line could significantly change their effectiveness.

What you see is reasonable....except I'm a bit surprised it is **only** 135 
watts. :-)

73 Tom 

More information about the CQ-Contest mailing list