My Kiwi and the ground
Hi,
My Kiwi antenna is a EFRW with a conterpoise, a UNUN 9:1 close to a CMC then 6 meters of coaxial cable.
There is no ground connection.
When I connect the Kiwi to the ground in the shack my SNR mesurement goes from 26/17 down to 21/17.
What should I do to keep a good SNR together with a ground connection ?
73 to all of you
Comments
Remember that the SNR1/SNR2 numbers cover different frequency ranges:
SNR1: 0-30 MHz (full range)
SNR2: 1.8 - 30 MHz (HF only)
So your HF SNR is unchanged at 17 dB. But since the full range is reduced by 5 dB I would guess this means strong signals in the AM BCB are being reduced. Or perhaps some RFI/EMI in the LF/VLF range which I imagine could be easily influenced by grounding configuration.
Use the spectrum display, set the zoom to level 4, and shift the waterfall to zero Hz on the left so that 0 - 1.9 MHz is shown. Use the peak hold function on the WF tab to compare what happens when the ground is changed.
Tnx for the tip.
When I connect the ground I loose around 20 dB between 500KHz and 1300 KHz.
It's important to know what kind of signals you're losing. All of them? Just strong RFI/EMI that might be influencing the outcome of the SNR measurement algorithm?
The SNR algorithm is rather simplistic and is the difference between the average of all signals levels (the assumed "noise" level) and the 95th percentile signal (not the 96-100th percentile to keep the strongest signals from being an anomaly). But this means that broadband, high-level RFI/EMI signals can establish the SNR high value.
If your HF SNR is unchanged that's all that matters, right? That's why there are two different frequency measurement ranges.
any idea to fix this ?
From a SWL point of view you are right the HF signals are the most important.
But from a ranking point of view when you filter the Kiwi list to view only the french ones I loose quite a lot of places.
So if I could fix this I would be happy. If not you are right it is not a big deal.
There are better people on here to suggest solutions for this than me. Looks like noise coupled on the ground connection (or a ground loop). Maybe choke it off somehow or shrink the loop.
If you are using an unbalanced antenna, it needs something to provide the "missing" half, that would otherwise be present in a balanced antenna. All simple antennas are basically dipoles. It's just that sometimes you can't figure out where the missing bit is, but it's always there !
The trick with unbalanced antennas is to provide the "missing" half in a controlled manner, and not to let random conductors, which may be polluted with noise from other sources, provide it.
Use a separate "clean" earth at the base of the antenna, and choke off unwanted noise that may be present on the feedline, by means of strategically placed common mode chokes and maybe additional "clean" earth connections along the cable route, as this improves the rejection by working a bit like a potential divider network.
Burying the coax can also help, as this provides both additional shielding and a low impedance path to ground.
Balanced antennas with good common mode rejection are nearly always better performers, and low impedance balanced antennas, such as broadband loops, tend to have intrinsically good common mode rejection characteristics.
Regards,
Martin
Oh and also, the SNR measurement method is seriously flawed, and can easily be fooled.
The human eye and brain can very quickly determine if the spectrum and waterfall looks "clean", but it's a lot harder to find an automated method that works just as well. There have been endless discussions about this, with no obvious conclusion or solution, but maybe AI will eventually figure out a better method.
In the meantime, just use the existing stats for guidance purposes, as it still helps to pick out the top 10% best performers, it's just the other 90% that are disappointing, but again that topic has already been done to death.
Regards,
Martin
Tnx for your comments I will try something in this way.
I found a temporarily solution which I would assess as a dirty workaround but leaves me the possibility to stay in the top ranking.
I grounded the kiwi through my kmtronic realy board and select the relay as the default antenna. The ground is connected when the relay is normaly closed.
When a user logs in or when the SNR mesurement works my relay goes on and the ground is disconnected. By this way I have still a good SNR ;o))
Today I tried something else.
Since the power plug of the kiwi is a 2 contacts (phase and neutral) I reversed the way it is plugged into the socket.
The effect of the groung connection on the SNR mesurement has been changed in the other way around.