On Jun 4, 10:30*pm, Owen Duffy wrote:

I doubt it. Again it relates to the loss characterisation of the
components. If you could do that, the device could also calculate its own
efficiency... now that would be a feature that would kill the market!!!


Actually, you could probably do this with a series of lookup table. A
tuner mfr has a fairly good idea (or can get a fairly good idea, with
some time to do appropriate measurements) of the loss properties of
the tuner components. Based on measuring some AT200PCs with a TenTec
VNA, there is *some* interaction between the components, so it's not a
perfect linear scaling as you step through the values. The tuner knows
the frequency, it knows the L and C, and it knows the parasitic Rs,
and it knows that it's "matched", so it should be able to calculate
currents and voltages, and figure out loss.

Whether you could do it in a tiny PIC... I don't know.



I'm going to suggest that MFJ make that a feature on an auto tuner. I

See above.

BTW, for referring measurements at one place to another, TLLC athttp://www.vk1od.net/calc/tl/tllc.phpmay be of interest. It would be a
challenge to incorporate those calcs in a small 8 bit microcontroller,
their capacity and performance on logs, hyperbolic cosines, etc is the
issue. Probably why most of the tools that do this, use a client on a PC
to do the calcs and presentation.


You'd be surprised.. the question is whether someone is motivated
enough to try and do it. There's not much commercial market, so it
would be a labor of love, and that requires a somewhat bizarre
intersection of someone who takes pride in putting complex math (in
both senses of complex) into a limited processor AND someone who is
familiar with the relevant equations and their use.

In any case, given that folks have done this sort of thing on
Z80/8085/6502 class processors... It doesn't have to be fast.. 10
seconds is a LOT of CPU clock cycles.
jim