Topband: (2wire) Beverage transformers

[HA3LN]

Hi All and HNY for 2021.

Preparing for the CQ160m with new (2 coax) 2-wire beverages to
cover the missing azimuthal gaps based on LBDX. The first 2x
Bevs worked great back in last Jan.

Now I have difficulties with reaching good imped match with the
T2 transformer (responsible to transform the 745 Ohms wire
impedance to 50 Ohm coax). I use n1=3T/n2=12T tapped @6T
transformer (2m high, 20cm wide with 0.8mm wire)

What can be the reason for the impedance transformation is
rather off to the calculated value?

This is the T2 transformer from 2019:
http://ha3ln.hu/VNA_190116_230811.jpg
...and this from yesterday:
http://ha3ln.hu/VNA_210101_153241.jpg

I have
- same wire with the diam (even from the same roll)
- same BN73-202 cores (tried to use several cores from different
   sources to eliminate the possible mix inconsistencies)
- same winding method (including n2 tapping)
- created a low inductance test resistor network for 744 Ohms

...tried to wind
- lousy, and precise (crossing windings vs. side-by-side, bunched
   wires, etc.)
- n1 first and n2, after n2 first and n1, of course no difference.
- without the tapping, same as above.
- difference turning ratios (3/12, 2/12, 1/12, 3/11, etc.) to see
   the change


The best I could reach now on 160m is
- SWR: 1:1.29 (Rs=40.4 Ohms, Xs=-5.4 Ohms)
vs. in 2019:
- SWR: 1:1.16 (Rs=43.2 Ohms, Xs=-1.6 Ohms)

I know, Beverages are really die hard antennas and this increased
mismatch might have zero effect on performance but still, the engineer 
part of me...


Thanks for the responses.


73!
Csaba  HA3LN
http://ha3ln.hu/