Help with Maxwell-Wien Bridge

Well, they tested it at their specified impedance, 1Meg to 15K. But I understand, the bandwidth would get wider is you tested it with 100K and 1.5k impedances. The high impedance response fulfills the Bell Labs 300Hz to 3kHz for intelligibility frequency span. Mikek

No problem. 5000 ohms DCR implies a lot of turns of very fine wire.

If you apply a DC voltage to it and measure the current, the current will reach 63.2% of the (DC voltage/winding resistance) in a period of L/R. If your numbers are correct, it should take 200ms.

Why won't that work for you?

Half voltage is 6dB down, not 3dB.

Don't I want 1/2 the voltage, for a series circuit? !/2 V across the resistor, 1/2 voltage across the inductor. Then resistance equals the reactance of the inductor. ?? Mikek

I'm very sceptical of calculations that give transformer reactance in the megohms at audio frequencies. That implies a very small amount of stray capacitance would resonate with the inductance and kill the performance.

Any audio transformer is going to have winding capacities that are larger.

Have you mentioned the manufacturer and model number of the transformer and I missed it? Or is the transformer homebuilt?

I sorted your list to show the strong relation of inductance with frequency.

There have been several discussions on the difficulty of measuring inductance of ferrite or iron core transformers due to nonlnearity with changes in drive level.

Are you perhaps having a similar problem with changes in drive level as a function of frequency?

The probe capacitance should not be a problem at these frequencies.

Also, your inductance values seem impossibly large. For example, 893H at

Something seems very strange with the transformer, the measurements, the calculations, or the fixture.

I don't know how you can get an LCR circuit to ring at two different frequencies.

You are forgetting the reactance of the stray winding capacitance in parallel with the inductance.

I'm trying to avoid DC on the coil.

Mikek

I'm not forgetting it, I have mentioned it several times, I've also ask how to tease out the capacitance number.

Thanks, Mikek

dB = 20*LOG(V1/V2) if V1/V2 is .5, then dB = -6.021

Yes, not something I kept track of.

Hmm. I'm using almost 6meg to get 1/2 voltage so just the 10meg probe resistance affects the measurement, And the capacitance, maybe your right, 15pf is probably swamped by the interwinding capacitance of the inductor. I don't know yet.

Yep

A capacitor of

That's what makes this interesting, a need to know!

Oh, you ring it with added capacitance, then you add more capacitance and ring it again. With that data and pretty basic algebra skills (which I don't have) you can calculate the self capacitance.

Thanks for your attention, Mikek

I'm not sure what the disagreement is here.

2) I have setup a series RL circuit, driving it at 1000Hz. I'm adjusting the R until I have 1/2 V across the L. My assumption and maybe I'm wrong, (please correct me) the value of R is the impedance of the inductor at 1000Hz.

From there I calculate the inductance. Mikek

Surely you mean sqrt(2)/2 on each?

There is no "half and half", the closest you can get is sqrt(3)/2 and 0.5 (at -6dB).

Tim

It messes up your calculation of "1/2 voltage across the inductor. Then resistance equals the reactance of the inductor." It means your inductance calculations are wrong.

The Owen bridge looks like it ignores the winding capacitance. In fact, all the conventional AC bridges ignore the winding capacitance. There is no term in any of the calculations for different bridges that refers to the winding capacitance. Please see

Even the HP Impedance Measurement Handbook begs off on inductor stray capacitance. The "Paracitics of an inductor" section 5.2.1 starting on page 94 assumes a high Q circuit at RF, and is useless for low Q audio:

It also discusses the errors due to variation in permeability with drive level on page 96.

The paper "MEASURING TRANSFORMER DISTRIBUTED CAPACITANCE" offers some guidance, but requires measuring the SRF of the transormer. This is difficult when the Q is low as in your case. However, it also discusses sources of error, including core permeability variation with frequency and drive level:

The National Instruments paper on Capacitance/Inductance Measurements briefly discusses inductance measurement variation with frequency and drive level on page 4. They helpfully observe "All of these factors can combine and cause inductors to have different values under varying conditions of temperature, frequency, and signal level.". See

Since these papers illustrate the variation in inductance with frequency and drive level, we conclude it is difficult to know the true inductance. Consequently, it may be difficult or impossible to determine the stray capacity of the winding. It could mean a futile waste of time and effort.

So we go back to the beginning. Why do you need to know the inductance of the winding?

Who is the manufacturer and what is the model number?

You are ignoring the stray winding capacitance. This invalidates your inductance calculation. Also, the core permeability changes with frequency and drive level. So any result you get is invalid at different frequencies and drive levels.

Explain it to me, is the math changed because it is an inductor and has phase shift?

Not sure what you mean by the "closest you can get" I can get any voltage ratio I want by varying R or frequency. When I get 1/2V, does the R equal the reactance of the L?

I'm here because I don't have the info I need.

Thanks, Mikek

NO - the resistor and inductor are 90 degrees out of phase, so it will be 0.707 times the voltage over both. 0.707 is the sine and cosine of 45 degrees.

Catalog page 16, (Ouncer Audio Units) Model 0-15 The Graph for the 0-15 is at the bottom of page 17.

I'm not ignoring it, I just don't have a number for it. I understand that any inductance I measure is smaller than the real inductance.

Ok, so if I set up the series RL circuit, input, resistor, coil, grd. Measure from the resistor/coil connection to grd, I want to see

What if I reverse the order to input, coil, resistor, grd. Now I'm measuring across the resistor. Do I still want 0.707 of the input voltage?

So, when John L said 3db point, that's what he meant :-)

Thank you, Mikek