Electronics > RF, Microwave, Ham Radio

Do I need a Low Pass Filter for High Q Antennas?

(1/2) > >>

Jeff Weinmann:

I've recently built this antenna design and it works great! (I measure 1.8 SWR, 56 OHM Z):


I'm using this antenna with a WSPR project using a transmitter that utilizes an Si5351a clock chip.

The Si5351a produces a square wave and as a result, produces harmonics on the usual odd frequencies.

The antenna itself however, has quite a High Q,  On my antenna analyzer, the only dip in SWR is the target resonant WSPR frequency and drops off significantly +- 100Khz from resonant.

I 'zoomed' out with my analyzer so I could cover 50Mhz span of bandwidth, and I only see the dip at resonant.  I see > 10 SWR for all frequencies within the span (I am targeting the 20 Meter WSPR frequency)

So my question is:  Do I really need a low pass filter on my square wave output?  I'm transmitting on QRPp level power (10 - 75 milliwatts),  and my low pass filter I think drops output 2 DBm or so.

In other words,  Does the High Q antenna naturally filter out the illegal harmonics enough not to worry about it?

Thanks for your opinion!

73 Jeff W8ZLW

yet another magic antenna; aside from all its properties you certainly need LPFs since you cannot estimate how well harmonics are suppressed by the antenna; the SWR says absolutely NOTHING about that.

the matter about suppression of harmonics is not one of the output power but of the relation between the 1st harmonic and its other harmonics - on shortwave it's by the radio regulation a suppression of -40dBc for all spurious. I seriously doubt this can be reach by that antenna

Generally an antenna resonant at f will be resonant at 3*f as well, although with probably different Z.

A LPF should definitely be used.


--- Quote from: Jeff Weinmann on August 04, 2021, 12:31:48 pm ---In other words,  Does the High Q antenna naturally filter out the illegal harmonics enough not to worry about it?

--- End quote ---

Yes, high Q antenna works like very narrow bandpass filter. But low pass filter is still needed for transmitting, just to avoid unwanted RF leakage from the feeder and other components due to a high order harmonics which are present up to 1 GHz for si5351 output. At working frequency such leakage is not a problem, but for a high-order harmonics it will produce unwanted illegal transmitting at forbidden frequencies.

Put another way: sure, SWR might be >10, or the antenna has a filtering effect; but it's only 1st order, i.e. -20dB/dec*.  And spurious emissions have to be -60dB or so.

*The equivalent circuit is an LC tank, which should be -40dB/dec; but the C of it is an electrically short dipole/whip, which is +20dB/dec in the asymptotic range, so only the inductor really matters!

Not to mention, the inductor may have resonant modes at high frequencies, thus coupling clusters of VHF harmonics into a now very characteristic-length element!

Be careful that the LPF/BPF needs to be matched to the antenna.  For exactly the same reason, i.e. SWR being large, out of band -- it's very reactive at most frequencies, so any leakage through the filter is reflected right back to it, affecting its response.  Best to model both together as a full network.  Which should be easy enough, assuming an RLC network for the antenna.  (Out-of-band resonances should be far enough away that absorptive (i.e. R+C or R||L) elements can be used to terminate the filter, or a constant-resistance or diplexing type filter can be used.)



[0] Message Index

[#] Next page

There was an error while thanking
Go to full version