On 4/1/2011 5:44 PM, Richards wrote:
Apropos to what Gerry and some others have said, I thought most ham type
watt meters had two factors that made it difficult to know if one's
power output is up to spec:
1) Typical meters have up to 10% tolerance - meaning up to 10% +/-
accuracy. That makes it almost impossible to know what you have, really.
And that's 10% of full scale, not of the reading. Bird slugs lose
precision at low levels where the diode threshold for forward conducting
is a serious fraction of the applied voltage. Part of the reason for the
nonlinear meter scale.
2) The Gordon West Extra Class license course makes a point of
explaining why one wants a PEP power reading meter - and then goes on to
explain why even that does not always show the real total power out -
especially on SSB - but even on CW - and explains that while close,
there are various technical reasons (yeah... like I can relate and
recall them accurately enough now...) that makes it show less than the
actual power out. Similarly, but in contrast, for example, some rigs
have a surge on first transmit, that might over-state actual power over
time, but a PEP reading meter with a slow decay rate (peak-hold) may
overstate the power out. Thus, our meters do not really give us the
actual power out. I can look up the reasons Gordo gives us for this
discrepancy if anyone is interested...
I hadn't worried about that, I've adjusted my FT-857s so they don't
spike at the beginning. But that spike would cause the peak hold to show
that peak, not the average power with a brick on the key.
I use another alternative when the envelope isn't uniform, a broad band
oscilloscope. In my case a 1970s vintage Tektronix 475 that's flat to
200 MHz and has useful (though uncalibrated for amplitude) to about 456
MHz. I stick its probe into one of my dummy loads and read peak to peak
volts and grab a trusty slide rule, taking peak to peak divided by 2.8
to get RMS, squaring that and dividing by 50, the presumed load
resistance. By use of the built in scope calibrator I can set the scope
gain as closely as the screen can be read, and with an adequate set of
signal generators I can check the frequency response to transfer that
calibration to the frequency under test. Then I can measure the dummy
load with RLC bridge to get a resistance to several significant figures
(a lot more work to do, but more trutworthy than many a digital meter
unless made for high precision and I can acquire resistors with 0.1%
tolerance to check the bridge or digital ohmmeter. So I can calibrate my
scope as power meter with fundamental standards of voltage and
resistance. And with the scope I can detect power spikes, droop, and
other envelope anamolies. I can also detect if the peaks of the SSB
envelope are being clipped because of over drive or the ALC isn't
working soon enough.
One problem with my scope is that it cost me enough bought new in about
1977, that I could have bought 5 top end tentecs at the time and a
calibrated new Bird with several slugs with the change. About $2200
then, but I figured it was fast enough for a long time and it has proven
to be fast enough through today for my applications. And with a 2 mv per
division scale I can measure pretty small powers.
Thus, we may be getting all the wattage we paid for after all... even
though we cannot prove it ! And I am totally sympathetic for anyone
who is skeptical or who has reason to doubt his actual power - as it
would be nice to know for sure since this stuff costs so much. One
should not need a dynamometer to know if his car meters are right - and
neither should we as hams with our radios. So, while think things are
better than we might think, it would be nice to be able to verify it
all.
A dynamometer won't necessarily show that your car speedometer is wrong,
that's most easily accomplished on a road with measured markers at 1/4
or 1 mile intervals and a stop watch. In Iowa on interstates there are
surveyed marks in the passing lane at quarter mile intervals. Some years
ago a state patrolman spent a summer or two laying them out and painting
them so their precision would stand up in court. Then if you do a whole
mile, 3600 / time in seconds is speed in mph. Reaction time is a bigger
fraction of the interval if you use a quarter mile which is what the
state patrol did from aircraft (that I don't think they can afford the
gasoline to fly these days). And on other roads away from rivers, cross
roads on section lines are generally but not always at 5280 foot
intervals. You make repeated measurements and when one is off, you
neglect that one. That section isn't the same size. Section lines are
more precise when going north or south than going east and west because
of the squares being smashed onto a spherical surface.
Or you can arrange to be paced by a police car with a presumably
calibrated speedometer. They do that often when they spot you going over
the speed limit and follow you whether you asked for it or not.
73, Jerry, K0CQ
Just my take.
------------------
Happy Trails.
======================= Richards / K8JHR =========================
|