You must be doing it wrong. 500+ activations using centre fed dipoles without that level of grief. There’s a dipole centre to which each dipole leg attaches. The feeder (RG174) attaches to the centre. All cables can be detached and wound separately. It helps if the pole is more rigid. My Decathlon 5m travel is a bit wimpy at the top, a 6m one with the top section removed would be better. But it still supports the various dipoles.
I’ve been using an EFHW in inverted sloping L because it’s something different to try and there should be more omni radiation from the vertical section (for world wide DX coverage!) compared with low inv-v dipoles.
The mythical 100pF across the primary cannot compensate for winding capacity. What it (hypothetically) compensates for is either
a) leakage reactance (L) in the transformer.
b) a lack of magnetising inductance by bringing the transformer into resonance.
Now if you use a VNA and a resistive load, you can see some benefit from this.
But you are not going to. You are going to put a wire on it. A very long wire, with a lot of capacitance.
Now being a transformer, 100pF across the primary is X=N^2 ie 1/49th = 2pF across the antenna end.
Care to guess how much antenna wire 2pF is?
When you tune your antenna this is tuned out. I believe the capacitor is voodoo.
I would also note that you can make a resonant transformer with low u iron powder or even air, where the transformer is parallel resonant. While N=3 is likely better for a broadband transformer matching a pure resistive load, again that is not our case. We are tuning the antenna to resonance across the complex Ztransformer. By doing this you can make a 2 or 1 turn primary work.
BTW, I use Owen’s 3:21 design myself, but his analysis is correct only within the unstated limits of his analysis. He completely overlooks the fact that we are tuning a complex impedance to our transformer, and that is a very different problem - but one that his transformer works just fine for.
Now in practice it is even more complicated as we use one wire, which is not actually resonant on harmonics, and we would like to have a good match on multiple bands. Add this in, and “tuning” a transformer to a resistive load is a fools errand.
How to prove the above is wrong?
a) Make a clamp on current transformer & meter you can thread on the antenna wire. and put it around the anti-node (middle) where current is max.
Now try transformer A and transformer B, making sure to tune the length for each transformer.
If the antenna wire is the same for both (a couple of metres difference at the far end is neither here nor there), then more RF current = more radiated power.
(This would be an interesting experiment to compare CF dipole and EFHW of the same length: measure RF current in the wire)
The old marine and land mobile HF sets used a lamp or led and tuned for maximum RF current in the antenna wire. It’s a really good idea
I don’t want to go into it any further because I don’t provide scientific justification for all of my statements.
But one thing is certain and confirmed by many statements. Only the capacitor produces an acceptable SWR on the shorter bands.
73 Chris