When adding the radials to my 160 T-vertical (72 ft vertical + 56 ft top loading t-wire on each side for a total electrical length of 128 ft) I kept notes on the SWR as I added 3 – 4 radials. As predicted I had a very good direct feed match starting with only 4 radials…e.g. between 1814 – 1824 the VSWR = 1.1: 1. But the reactive part of the load X = 8. According to the books we have not reached anywhere near peak efficiency. As I added groups of radials the numbers appeared to get progressively worse…but I believed the experts that the efficiency was getting better. But with 10 radials X = 0 over much of the band and the SWR was generally 1.2:1. Skipping all the in-between measurements, when I reached my goal of 32 radials (I could only expect 0.3 dB gain going to 120 radials the nirvana)…the VSWR was now a flat 1.7 – 1.8 between 1800 – 1900…but X = 0 over much of that range….so the load impedence was almost all resistive (R). On the MFJ 259B R = 38 – 26 over the same range. But the VSWR being somewhat higher than I would like, I researched putting a 1.56:1 UNUN (unbalanced antenna to unbalanced coax) right at the antenna base. This would match 32 olms impedenace of the antenna (almost all R ) to 52 ohms of the coax. Balun Designs, LLC makes baluns of all types and can handle up to 2 KW. At $69 I thought I would try this approach for the first time. It took a couple weeks to get it. Sent USPS Priority. Looked nice and installed it today. VSWR is now 1:1 on 1820 and > 1.5: 1 from 1800 – 1870. At 1900 it’s 1.8:1. The MFJ will also measure Percentage of Transmitted Power (a backhanded way of stating VSWR) and it peaks at 99% and drops only to 90% at 1900 kHz. I worked ZL8X this morning BEFORE installing the UNUN. I suspect that this mod will make little difference if any in DXing. But it would appear practically all the watts are now going into the antenna. I think this is probably a good thing. — W1XX