I guess I do not understand the rationale for a 9:1 unun. The implication is that the random wire has an impedance of 450 ohms. There probably IS a frequency at which the wire has an impedance of 450 ohms but to hit that frequency by chance would be highly improbable and it certainly would not present that impedance over a range of frequencies.
A random wire has two very predictable impedances: nominally 50 ohms at frequencies where the length of the wire is an odd multiple of a quarter wave and nominally 2600 ohms where the wire is an odd multiple of a half wave. At all frequencies in between the wire represents a random impedance between a few ohms and 2600 ohms.
The fairy tale that a 9:1 or a 49:1 balun that makes a random wire into an efficient broadband antenna is simply wrong.
I know, I know...... your SWR bridge says you have a 2:1 or better VSWR over a broad range of frequencies. The reason for that is the 9:1 balun is relatively efficient at the frequency, if there is one, where the wire represents a 450 ohm load. But the balun is wildly inefficient at all other frequencies where it is badly mismatch and the majority of your transmit power on those frequencies is turned into heat in the balun. The reason your SWR bridge shows a low number is it is measuring reflected power and any loss between the transmitter and the load represents a double loss, once for the outbound power and a second time for the reflected or return power. Your SWR bridge sees low reflected power because it is mostly lost in the balun. In most cases less than 50 percent of your power actually gets to the wire and in many cases it is less than 10%.
The principle is exactly the same for the famous EFHW antenna fed by a 49:1 transformer. This antenna is an excellent, efficient radiator at the frequency where the wire represents EXACTLY an odd multiple of a half wave. At that frequency the wire presents a nominal 2600-3000 ohm load to the transformer. The transformer efficiently converts that impedance to something close to 50 ohms (2600/50=52 ohms) and you will get a low VSWR and an efficient antenna.
At all other frequencies the transformer sees a horrible mismatch because the wire presents nothing close to the necessary 2600 ohms. So the transformer delivers whatever power the mismatched wire will take and converts the rest to heat. Reflected power is low because it suffers a double loss in the transformer. So your VSWR bridge tells you your magic transformer has violated the laws of physics and turned a random wire into a constant impedance load across a broad range of frequencies. It is lying and there is no such magic transformer. And you are getting very little power to the wire on the vast majority of frequencies.
I realize this is heresy to the "EFHW" advocates. "I make contacts on different bands all over the world with my EFHW". Well, if you start out with 100 watts and you end up with 10 watts or less at the antenna you have suffered a 10 dB loss, and you will on most frequencies where you use such a contraption, then 10 watts can make contacts all over the world. I regularly make contacts into Australia and New Zealand on 100 milliwatts. But those who think it is a magically efficient broadband antenna are just fooling themselves. Or, more accurately, are being misled by their SWR meter into thinking that is what they have.
WA8TOD
Warren Allgyer