Hi John,
Back in the seventies I did some research to develop a custom Q
meter at my employer's.
It had to cover from 1 KHz to 5 MHz. And yes the transformer
injection method was used as in the HP4342A.
For amateur applications, you probably want to cover 1 to 30 MHz.
I used a very high mu tape core which worked well down to 1 KHz.
Getting the low impedance drive (less than 5 milliohms) was
difficult at the higher frequencies.
I found that I had to use a large copper strip at the secondary.?
The HP4342A seems to use a copper cup
around the toroidal transformer. The cup is connected to ground at
its perimeter while the LOW terminal
is connected to the center of the cup.? This provides the lowest
residual resistance and inductance for the
excitation source.? HP claims about 1 milliohm.? And the measured
excitation voltage varies only +/- 1 % over
the frequency range (22 KHz to 70 MHz).
The 75 ohm termination at the transformer input establish a a
fixed load over the frequency range.
Looking at the design of the HP4342A shows that they took a lot of
precautions to maintain the highest
impedance across the main variable capacitor.? First they use a 30
to 1 capacitive divider which increases
the amplifier resistance by a factor of 30x30 = 900. The amplifier
itself uses a FET at input with a 28 megohm
gate bias resistor, plus it's bootstrapped to further increase the
input resistance.?
Note that the high input resistance is required at the lowest
frequencies.
At 22 KHz, with C = 25 pF,? 58 Megohms resistance across the L-C
will cause a 0.5 % error.
At 1 MHz, with C = 25 pF,? 1.3 Megohms resistance across the L-C
will cause a 0.5 % error.
I have just calibrated of my HP4342A.?
The injection voltage varies with the Q range as follows:
Q????????? Injection V
30???????? 30 mV???? (as calibrated)
100?????? 9 mV
300?????? 3 mV
1000??? 0.9 mV
The calibration is done at 30 mV with the Q range set at 30.
This gives 900 mV full scale at the voltmeter input, on all Q
ranges.
The voltmeter is adjusted to read 900 mV at 50 KHz first.
Then two high frequency compensations are done at 20 MHz and 70
MHz.
These calibrations assume that the variable resonating capacitor
is 'perfect' since it's the reference.
I have never seen an external capacitor that will give an increase
in the measured Q, so it
must be very good.
Note that a 10:1 scope probe will present an impedance of 10
Megohms at DC, but at RF it will be much lower.
I measured my TEK scope probe P6131 using the delta Q method:
At 13 MHz, the measured parallel resistance was 11845 ohms and the
capacitance 10.2 pF.
See:?
Other articles which might be of interest:
??? (Published in QEX jan. 2012)
Using a VNA or a Spectrum Analyzer to measure attenuation allows Q
measurements with ONLY a resonating capacitor.
It also computes the L and C values as seen by the resonant LC
circuit.
See:?
Hope this will help those that are looking in Q meter designs.
Jacques,? VE2AZX??
Web:? ve2azx.net
