Hello David,
No need to apologise, you did reply.
regards?? Ken g8beq
On 14/08/2022 21:43, Dr. David Kirkby,
Kirkby Microwave Ltd wrote:
toggle quoted message
Show quoted text
On Thu, 4 Aug 2022 at 10:13, Kenneth Greenough via
<g8beqglossop=
[email protected]>
wrote:
Re' a Q meter.
This is only a thought idea, not tested hi.
Have a small enclosure with sockets on opposite sides.
each socket has a
single turn loop attached with sufficient distance between
the two loops
to allow the coil under test to be positioned between
them. one turn is fed
from your sig' gen' and the other feeds a scope or RF
meter.? A sufficient
level from the sig gen may require an amplifier. With the
coil to be tested
placed between the loops tune the sig gen for maximum
throughput, ie
resonance. Then tune the sig gen lower until the
throughput is 3dB
down, ie 0.71 reduction in voltage or 1/2 power. Repeat by
tuning the sig gen
higher than resonance for a similar drop in throughput.
the resonant
frequency divided by the bandwidth just ascertained is the
Q of the coil.
The coil being tested can, of course, have a capacitor in
parallel if wished.
73 Ken G8BEQ
?
Apologies for not replying earlier.
Yes, this is a known method, although it seems more common to
use a VNA now and measure S21 as that makes sweeping the
frequencies much easier.
Both inductive and capacitive coupling
can be done. That measures the loaded Q. I think I am
right in saying that the coupling factor at each end of the
DUT needs to be the same in order that it computes the unloaded
Q, which is what is required. See particularly section 3
of
A good practice guide from NPL does
this, but fits the transmission curve, not just a measurement
of 3 places.
Dave
