[Elecraft] [QRO] KPA500 variation of PA current band-by-band

[Don Wilhelm]

Tony,

All you say is true. and should be heeded by all hams.

I have dummy loads that are known to be flat (within 5%) to 1 GHz, and I 
have several more that are flat to within 500 MHz.  Others will display 
the same characteristics to 60 MHz.  I know which ones are which, and I 
rely on them to give me precision power measurement capability.

I do not depend on anything other than the RF voltage measured across 
that known good dummy load for true calculation of RF power (although I 
do have one wattmeter calibrated to NIST standards that I use as a 
secondary standard to calibrate my other meters).  In other words, I 
have become more and more distrustful of in-line wattmeters which may be 
in error by a substantial amount (20% of the full scale reading) - just 
try to measure 5 watts with a wattmeter having a 200 watt scale - the 
error may be as much as 40 watts anywhere on the scale - so a reading of 
5 watts might be anywhere between zero and 45 watts actual power - does 
that seem ridiculous - certainly, but it is true - some wattmeters can 
be in error by that much when used at levels considerably less than 
their full scale reading.

OK, this is my wattmeter inaccuracy rant for Feb 2012 - thanks for 
listening.

73,
Don W3FPR

On 2/7/2012 10:09 PM, Tony Estep wrote:
>> On Tue, Feb 7, 2012 at 3:23 PM,<n5ge at n5ge.com>  wrote:
>>
>>> Is your dummy load absolutely a flat 1.0:1 from 1.0MHz to 30MHz?
> ================
> Suppose that the SWR is 1.2. This means that 0.8% of the power is
> reflected (80 milliwatts reflected for 10 forward). If the SWR is 1.1,
> the reflected power is 0.2%. A simple coax jumper can sometimes cause
> an impedance bump large enough to reflect more power than that. A few
> minutes spent on a test bench with even the fanciest precision
> equipment will quickly disabuse one of the notion that a dummy load
> can be trusted to be "absolutely flat 1.0:1 from 1.0MHz to 30MHz."
>
> Tony KT0NY
>