On Feb 2, 9:10 am, "Ralph Mowery" wrote:
"Richard Clark" wrote in message

...

On Sat, 2 Feb 2008 08:45:27 +0000, Ian White GM3SEK
wrote:

As with the AN-URM120, the Bird design adjusts the coupling so that all
detector diodes in all the inserts are operating at the same RF levels,
so they can all share the same meter scales. The difference is that
instead of physically moving the same insert inward or outward to adjust
the coupling, Bird do it by selling us more slugs :-)

Hi All (Congratulations Ralph),

There are more differences than that. The Bird suffers more from
parts erosion than the URM120 as the URM has bigger elements. With
the bigger elements, the geometries are held to a tighter precision.

I was luckey enough to find one that was in the origional wrapping paper. I
don't remember what the date was on the wrapper. At a hamfest several years
ago soneone had several new ones. They were in the heavy duty aluminimum
foil , cardboard box, and the hard plastic foam filled case. Never opened
from the day they were calibrated about 10 or 20 years before. It came
with 3 elements that would go from about 3 mhz to 1000 mhz. Think it topped
out at 1 kw up to 30 mhz and 500 watts after that. I bought it for what
some used Birds without elements seem to go for on e-bay now.
One minor thing is the meter movement on those meters are sluggish so it
takes a slow adjustment of amplifiers to get them to the max output.

I have one of those, or something very similar; a Struthers 1219-D.
The HF slug that came with mine goes up to 5kW, though. The meter
movement measures 24uA full scale, and the coil resistance is 2k ohms,
so the meter is pretty sensitive: about 1.1 microwatts for full scale
deflection. It takes that sort of sensitivity to get down to the
linear detector out versus RF power in region.

Interestingly, if I simulate a single-diode detector using an
HSMS-2850 Schottky diode into a 2000 ohm load, and find the RF voltage
that gives 24uA output, and then re-simulate with voltages in the
right ratios to give 90% of that power, 80%, 70%, etc., down to 10%,
the outputs are not exactly linear with input power. The error is
small, though, no more than about 1.5% of full scale. So then I
checked the calibration on my meter, and put in 50% of full scale
current, and noted that the meter reading is not exactly 50, but just
enough off to be correctly reading the RF power. So though the scale
on my meter appears at first glance to be linear, it's not exactly so,
and I believe that must be intentional. The worst-case "error" to
linear output in the simulation was at 70% of full power, where the
current is predicted to be 71.43 percent of full scale. On my meter,
when I put in 71.43% of full scale, indeed it reads almost dead-on 70%
power. I doubt that's a fluke.

Cheers,
Tom