On Sat, 3 Jan 2009 16:50:39 -0800 (PST), wrote:

A ham seriously interested in 0.1 dB measurements will probably be
able to scrounge up something to use as a transfer standard

Name one.

73's
Richard Clark, KB7QHC

On Sat, 3 Jan 2009 16:50:39 -0800 (PST), wrote:

No.. if you KNOW the mismatch, it's not an uncertainty anymore.

Well Jim,

It is painfully obvious that you haven't done any of this and
dismissing error is easier said than done for you. The reference
material from the vendor of your choice obviously says quite the
opposite - so I can add very little that would lend any more authority
that you are so ready to discard for the sake of argument.

73's
Richard Clark, KB7QHC

On Sat, 3 Jan 2009 16:50:39 -0800 (PST), wrote:

The datasheet for the head gives this.

Can you quote this instead of making vague references? If you could,
then you would - this is simply more off-hand commentary....

If you're looking at the
thermistor/thermocouple mount style head,

This reveals you don't know which - each technology is very different
in operation and characteristic; much less sensitivity and power
limitation, Z shift, temperature characteristic, circuit topology, and
reference requirement to support the accuracy you suggest.

the Z looking into the head
is basically that of the load resistor, which, if held at constant

IF? Basically? You don't really know - and it appears that what
doesn't serve your defense doesn't matter.

temperature (constant = within a few degrees),

a speculative FEW degrees! You really are wandering the wilderness
here.

I doubt it changes more than a fraction of a percent.

Doubt? No metrics, except for the half-hearted guess that isn't even
close because you cannot settle on any single sensor, much less the
actual sensor!

A diode head (like the 8481,8487, etc.
for HP/Agilent meters) is also going to be pretty good.

Where the temperature characteristic of a diode is a power function?
This is growing more extravagant by each added sentence. You now have
wandered the landscape of power sensing through three technologies,
none of which you can pin down to the actual application, and without
any marked experience in the limitations of any of them.

Agillent
claims the increase in uncertainty for ALL causes from an extended
temperature (0-50) over the specified 25 +/- 5 is something like
0.9%.

And the claim is for what, specifically? The box or the head? I've
offered links to authoritative resources and specific entries, can you
recite your own or is this stuff now the domain of monkeys in lab
coats?

73's
Richard Clark, KB7QHC

On Sat, 3 Jan 2009 16:50:39 -0800 (PST), wrote:

I think that reasonable folks could differ on the concept of "required
calibration cycle" and "aging life"...

This is called delusional.

It's not like the thermocouple
or load resistor in a power meter head has an aging process that
causes it to suddenly go out of cal after 6 months.

This proves it. "Suddenly" - is a manufactured argument if ever one
was offered. You keep the "suddenly" as a trophy, and the rest of us
will be content with the natural progression of aging that doesn't
need to be forced into a spike to serve the convolution of argument.

73's
Richard Clark, KB7QHC

On Sat, 3 Jan 2009 16:50:39 -0800 (PST), wrote:

Some may appeal that their Relative Accuracy has the merit of Absolute
Accuracy - until you ask them to measure the actual, absolute value of
their ad-hoc standard. I have built primary standards and measured
them to 7 places. Over several years they migrated in value by 2 of
those least significant digits, but only in comparison to standards
that had "aged" and been calibrated at a national primary standards
laboratory. By themselves, I could have fooled myself (and perhaps
others less sophisticated) that they were absolutely accurate to the
extent of the number of digits my instruments could resolve.

--
So, you measured to 1E-7, and over several years, they changed by
1E-5? That's a whole heck of a lot better than 1E-2 (which is what
0.1dB implies).

I have accurately calibrated instruments beyond the resolution of
1E-12, but not for power. Metrology encompasses many physical and
electrical phenomenon. The standard mentioned above was a resistor,
not power. I said it was a primary standard, not a sensor.

Would the atypical Ham have the capacity to build a standard resistor?
Sure. Could that atypical Ham qualify the standard resistor? Maybe.
Could that atypical Ham calibrate the standard resistor? Not in
decade except by frequent shipment to the nearest primary standards
lab on their terms of calibration frequency - which would be very
frequent given the lack of aging. In that sense, the atypical Ham is
wholly incapable of calibration beyond the sense of it being performed
through proxy. However, as you have already dismissed the frequency
of calibration to the whim of desire - what price performance?

Of course, that same atypical Ham could simply take one certificate of
accuracy for 1ppm and derate it to 0.1% and probably live with that
accuracy for the rest of his natural life. Unfortunately, 0.1%
accuracy at D.C. doesn't lend itself to precision, much less accuracy
at 50MHz. In fact, that standard resistor would probably be off by
several orders of magnitude (10,000% error). I feel safe in saying
"probably" because that error would be overwhelming to the point of
saturating disbelief.

Those 3 orders of magnitude are why I think it's
reasonable and plausible for hams to make measurements to 0.1dB.

Try fishing harder that simply throwing money at this.

You still don't have a METHOD for achieving this vaunted 0.1dB power
determination, do you? Summoning up vague references to foggy
characteristics isn't going to gather up even a 10% accurate
determination.

is something
eminently doable for ham use (See Larsen's paper from 1975) and mostly
depends on a "really good" DVM for its accuracy.

This is quite a joke. "Really good" indeed - you may as well add
praying to an idol on an alter mounted in the corner of the shack to
boost results.

The original post had to do with comparing measured antenna patterns
against NEC models. That IS the case there.

And I would say that you have done a job of mocking the performance
Mac achieved which conformed to the best practices and METHODs offered
in the modern links provided by me. Those best practices and METHODs
have survived intact through the intervening decades because of their
fundamental necessity. You haven't reduced any of his errors except
by an automated sensor system. Given the other irreducible errors
that still inhabited his measurements (something you have yet to
respond to), power determination still resides in the domain of 10s of
percent error.

The case is this error still eclipes your 0.1dB by a vast margin.

73's
Richard Clark, KB7QHC

The case is this error still eclipes your 0.1dB by a vast margin.

73's
Richard Clark, KB7QHC

Look.. you're asking for specifics and details. That's out of scope
for a discussion like this. In the context of a discussion list, to
which all of us contribute as an avocation (at least, I suspect that
nobody is getting paid to participate in rec.anything), I assert, with
some casual backup, that it is possible for an amateur to make power
measurements such as those to validate a NEC model to an accuracy of
0.1dB. You assert that it isn't, and ask that I provide details beyond
my casual assertion.

We can go back and forth about metrology, but, realistically, I'm not
about to go dragging out data sheets and doing an uncertainty analysis
which is of very little value to me, personally, and realistically, of
little value to anyone on this list.

As with many precision measurement problems, the method has to fit the
thing being measured, so to spend a lot of time and effort on any
arbitrary contrived example (such as the one in this thread) isn't
necessarily going to be general or applicable to any other examples.
So why beat ourselves up about it.

The folks who pay me seem to be happy with the measurements I do (and
I assure you, I do measurements or generate signals to better
precision than we've been talking about here, some are even published
in the peer reviewed literature) . The folks who pay you are
presumably happy with the measurements you do. Neither of them pay us
to solve ham radio's problems.

Pace,

Jim