开云体育

On Mar 18, 2021, at 11:05, Manfred Mornhinweg <manfred@...> wrote:

?

Manfred, Have you ever compared the Q readings measured by the old BRC-160 or
the Hp 4342 Q meters against the Q reported by any network analyzer or even
many of the older 4 terminal instruments from the 70's and 80's and maybe into
the 90's?

No, I haven't, because I don't own any of those old meters. At work we did have a big HP VNA that went up to a glorious 13MHz, but I didn't have a NanoVNA to compare to that, in 1990...

The old Q meter uses a parallel resonant circuit technique and
this was considered the best method until technology caught up.

I normally measure the Q of inductors by resonating them with a high quality capacitor and measuring the resulting bandwidth. For many years I did this using a grid dip meter, and nowadays I do it with other devices, including the NanoVNA. But this technique is based on the assumption that the capacitor's Q is so much higher than that of the inductor, that the total Q measured is close to the inductor's Q. This technique cannot be applied to measuring the Q of capacitors, because sufficiently good inductors don't exist. An inductor Q of around 10000 to 50000 would be required, depending on the quality of the capacitor to me measured!

Which brings up the idea of measuring a capacitor's Q by resonating it with a quartz crystal... With some care, that might be workable in some cases. Measuring the impedance curves of the crystal alone, then with the capacitor in series, and then doing the maths.

But I would like a simpler method, that displays a curve of the capacitor's Q over frequency. The crystal method can't do that. An extremely good VNA should be able to do it.

Manfred