This has been an interesting thread.
It would seem that most people have a "favorite" instrument for making
these measurements. I guess I'm the odd man out here because I don't
have a "favorite". I have a number of different instruments and/or
methods of measuring L's and C's and use all of them. Admittedly some
are used more than others. Those that are used more are the ones that
are quickest to grab and use and are not necessarily the most accurate.
In my opinion, an inductor is the most nearly perfect example of the
Schrodinger's Cat concept. In other words, the very act of observing
(measuring) can cause the observed object to change enough to possibly
cause the observation to be meaningless.
The object of the whole issue of measuring inductors is to insure they
will work properly in the end application. An inductor that exists out
in "free space" is pretty much useless to anybody except maybe
theoretical scientists. So are inductors attached to test equipment, no
matter how much that test equipment costs.
Over the last 50 years or so, between on the job and at home, I've had
the opportunity to all of the instruments and test method describe both
in the posts here and direct emails I've had. If not the exact
make/model, then equipment similar enough to be little different. In
general I've found that the more the equipment costs, the less useful it
is to me as a ham! Part of this is due to the realization that
attempting measurement accuracies of 0.1% or even 1.0% is a fools
errand! For home ham radio projects, excepting critical filter
applications, getting an inductor measured to within +/-5% is better
than good enough. Why? Well because of that damn cat mentioned before.
Just moving the inductor from the test set to the final application will
change it's parameter. The difference can range from "insignificant" to
"considerable". The only way you can tell for sure is to carefully
measure the performance of the application network in actual operation
to see if the inductor came out as it was supposed to.
Most amp builders don't care, they just twist the crap out of those two
knobs on the PI network for max power out and move on. If you want to
impress me with your "professional skill" level and measurement
capability, throw away those knobs and use fixed capacitors!
With the inductor design software available these days, you really don't
even need to bother measuring tank coils and taps on them. The last tank
I made, I used KM5KG's Rf network Designer to design the inductor. It
even lets you calculate where the taps should be placed. Granted there
is a steep learning curve associated with these tools and the physical
measurements should be made within +/- 0.001". Making the tank coils
this way, other than a cursory check with the MFJ, no measurements were
made prior to final test of the PI network. Using Jim's criteria, the
caps were set to the values calculated by GM3SEK's wonderful spread
sheet and on all bands, the MFJ showed 50+J0 +/- 5% when the design load
resistance was placed between the anode to ground.
An even more critical evaluation ot the completed tank can made by
measuring the actual loaded Q of the entire tank network. This requires
a signal generator and spectrum analyzer or a VNA like the N2PK-VNA that
can do transmission measurements. If there is interest in this
measurement, I will go into details in another post. This measurement is
always interesting and sometimes surprising as the actual Q is often not
anything like the Q you think you designed the tank for!
If the software designed inductor does not work as advertised, it is
only because the input data to the software was incorrect or (more
likely) the inductor was jammed into the chassis too close to other
objects which significantly altered its parameters.
Now I have only three samples to base my MFJ-259B performance
observations on. One borrowed instrument, and my own, before I zapped
the detector diodes and after MFJ replaced the detector board. In all
three cases, the MFJ produced measurement results that were at least an
order of magnitude better than what VE7RF reported. The caveat here
being that only measurements the instrument was capable of making were
considered and the "fixturing" was properly tended to.
When I zapped the diodes, I didn't realize it for almost two days. The
measurement results were off by about the amount reported by VE7RF and
for what I was doing at the time were believable. It wasn't until I got
around to "final testing" the networks that I discovered they didn't
work properly. After I had MFJ repair the instrument, it went back to
being the "go to" L meter on the workbench. Proof of it's accuracy is
that toroids set with it work as they should in complex filters.
However, the accuracy suffers some on large tank type coils due to the
need for long connections from the instrument to the coil. There is no
way to "zero" out the lead contribution and depending on how they are
positioned with respect to the coil under test and the instrument
itself, the results could be misleading.
If I had either the B&K or the AADE meters, I would use them (a lot!).
But only for air core coils. Toroids would mean back to the MFJ again,
or one of the other half dozen ways of measuring I have.
73, Larry
Larry - W7IUV
DN07dg - central WA
http://w7iuv.com
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|