How covers inpact the inductance, and then some...

[Joe Gwinn]

I remember building the 'ol Wein Bridge back in the 80's, something like using a lamp as the element, will check your links but also I'll just have to dig out my old papers on it, ...

The lamp approach still works, but actually limits how low distortion one can achieve. The chapter in the Williams book lays it all out.

... dang you're high maintenance...

All for your better education.

BTW, it's <3% at maximum amplitude (which is like 10 volts P-P) but very small sine waves (under 500mV) it is quite good, don't know the exact dist spec though.

Even if it drops to 1% at lower output levels, it will still be too high. I had a function generator with 3% distortion (BK 4040A), and it proved unusable.


This is probably a good time to explain why low distortion is needed.

Unlike most kinds of AC bridge, the Maxwell-Wein bridge's balance equations are independent of the test frequency, a great simplification. One could conclude that this means that the spectral purity of the test signal is therefore unimportant, as all frequencies (and thus harmonics of a distorted test signal) will come to balance at the same time. And in a perfect world this would be true. However, the inductance and capacitance of practical components varies with frequency, and in particular iron-cored inductors vary a lot even if there are no eddy currents. And in pickups there are almost always eddy currents.

The net effect of all this is that the fundamental and harmonics of an impure test signal will come to balance at different places, and this will smear the "null" out sufficiently to make finding balance difficult.

Withe a pure sinewave (low harmonic distortion), there is a single null, and no smearing.

[David Schwab]

I see on the Duncan website they talked about the test equipment they use:

Rigorous bench testing

Resonance was calculated using an Amber 3501 distortion and noise analyzer and an HP 3314A function generator in conjunction with an impedance measurement fixture designed and built in SD Engineering.

[Joe Gwinn]

I see on the Duncan website they talked about the test equipment they use:

Rigorous bench testing

And "Inductance measurements were made directly using a GenRad 1658 RLC DigiBridge". The 1658 was replaced by the 1659. Datasheet available here: http://www.ietlabs.com/Genrad/1659.html. The 1658 and 1659 have built in test-signal generators, most likely a buffered Wien Bridge oscillator. Only $5,789 new, and probably also available on the used market.

IETLabs has the LCR-500 ($370 new), which sounds like it works by the same principles as the Extech LCR meter (which may be a clone): http://www.ietlabs.com/IET/LCR-500.html.

IETLabs bought the rights to and wreckage of GenRad (nee, General Radio).