开云体育

Recently I acquired a type 130 L-C meter, date codes 1963. Accuracy seemed still quite good, judged with some length of coax as capacitors. Next I calibrated the meter using only a 0.2% 300pF capacitor and a frequency counter (DC510/7D15). Only one capacitor is needed under idealized assumptions: fixed frequency is constant (at prescribed 140 kHz), internal L and C of the variable oscillator are independent of frequency. So the variable oscillator frequency is constant/sqrt(L*(C + Cext)) in the pF ranges and constant/sqrt((L + Lext)*C) in the uH ranges. An almost negligible complication is that the capacitor to be tested is connected in series with an internal 0.1 uF. A 300 pF test cap gives Cext = 299.1 pF. Calculations show:
C = 1135.1 pF, L = 1138.5 uH.
Theoretically the detector frequencies (test point #4) for full scale are as follows.
300 pF: 15.451 kHz
100 pF: 5.782 kHz
30 pF: 1.8136 kHz
10 pF: 0.6126 kHz
3 pF: 0.1846 kHz.
The procedure is
1) set T30 for 140 kHz.
2) select 300pF range. Zero the detector frequency. Attach 300pF cap. Adjust T1 for above given detector frequency. Repeat a few times until no further improvement is obtained.
What rests now is setting the display meter sensitivities. Symmetry adjustment should have been done first (same test point #4).
3) select 300pF range, connect (about) 300 pF, adjust course/fine knobs for detector frequency 15.451 kHz and set Adj. 2 for full scale deflection.
4) similar for 100pF range and 5.782 kHz, adjust Adj. 6 for full scale deflection. And so on for the other ranges. In my case I could set for full scale frequency using only the coarse/fine knobs, no capacitor attached.

Tonight I allowed for 30 min warm-up again and then did repeated measurements using General Radio GR900 30 cm airlines. These are specified with capacitance 20 pF +/- 0.07%. I used one airline as "cable" and after zeroing recorded the effect of adding the second airline. This has the advantage that stray capacitances at the open end stay the same and can be ignored. The capacitance measurements (calculated from measured detector frequencies) were within 0.4% of 20 pF.

For the low ranges the manual advices to set the meter at a small bias value instead of zero. I did so with that 20pF measurement. It seems that variable oscillator locks to the fixed oscillator frequency when the variable frequency approaches the fixed oscillator frequency. As it happens, my daily journal mentioned that for the first time a rigorous derivation has been given for the synchronizing behavior of pendulum clocks. I think this is the article:


Albert

Yes, the 130 is quite a nice LC meter. I got one a few years ago, and found it to be in very close (<1-2% about as good as I could resolve on the meter) agreement with some pretty good reference caps, as was. Not bad for a 50-60 year old instrument. I've gathered up some spare tubes and other maintenance stuff for it, and plan to make some special probes to go with it. I found that you can extend the measurement point out from the DUT port connector by up to 2-3 feet if you use 75 ohm cable, and still have enough offset range to zero it. I did some experiments measuring scope input C with this setup - it works great as long as the R is high like the 1 megohm standard inputs.