LCR negative inductance

[bowman1710]

Hi guys,

I am measuring an inductor with an LCR meter but applying an external DC bias to the inductor to see the effects this current has on the overall inductance (as pictured below). I have already checked the SRF with another piece of equipment and the inductor had a SRF of around 230KHz. At the moment I have done tests at 100Hz and 1KHz, through currents at 0.1A to 1A with the setup. When i get to the 10Khz test and apply the DC bias the LCR meter reads a negative inductance. Is this because the SRF is being dragged right down and is showing a negative inductance because the inductor is behaving like a capacitor, or is there something else I am missing here. Any help is appreciated!

 

[dick_freebird]

Dom't know what the internal circuitry is, but it could well
be getting jacked by the DC bias. That transistor supply
could be (and likely is) adding mucho shunt C to the setup.
I think you ought to look at a more resistive / neutral bias
scheme, maybe use a series choke at least 10X what you
intend to measure and de-embed the residue?

 

[mtwieg]

It's likely that the parasitic capacitance of your current source is adding capacitance and dropping the SRF. If that's the case then you may need to change the source design. It's also possible that the small DC bias voltage across the inductor is messing with the LCR meter, try adding a blocking capacitor between the meter and the rest of the circuit.

 

[bowman1710]

That transistor supply could be (and likely is) adding mucho shunt C to the setup.
I think you ought to look at a more resistive / neutral bias
scheme, maybe use a series choke at least 10X what you
intend to measure and de-embed the residue?

Is there any online paperwork around this method? How accurate is this normally?

It's likely that the parasitic capacitance of your current source is adding capacitance and dropping the SRF. If that's the case then you may need to change the source design. It's also possible that the small DC bias voltage across the inductor is messing with the LCR meter, try adding a blocking capacitor between the meter and the rest of the circuit.

My first initial thought was that something was happening with the SRF but I did wonder if there was something else going on so I will give this a try and see whether it has any effect on it, if at all, it won't take a second.

The internal circuitry I am using it is a MJL4302AG NPN fet, with a variable resistor on the base, the power supply is a 6V bench supply. Obviously I am just changing the base value to change the current. The inductor I am measuring is a 10mH inductor.

I will try the cap then see where I can go from there is not....

- - - Updated - - -

LCR didn't change with the cap, is there any suggestions for another type of topology i could try and see if it works?

 

[mtwieg]

The internal circuitry I am using it is a MJL4302AG NPN fet, with a variable resistor on the base, the power supply is a 6V bench supply. Obviously I am just changing the base value to change the current. The inductor I am measuring is a 10mH inductor.

That's quite a big transistor. Datasheet says its Cob capacitance is 600pF with a Vce of 10V. In practice it's probably higher since your bias voltage is lower. Much higher than the self capacitance of your 10mH inductor, anyways.

When I want to test inductance vs bias current on power inductors, I typically don't use an LCR meter. Instead I apply a fixed DC voltage to the inductor and measure the current waveform. Plotting di/dt vs current allows you to see the change in inductance clearly. Doing the experiment requires a very low impedance, high current voltage source and a good switch. Previously I used a DIY spot welder, which was basically a large capacitor bank which connects to the load via an SCR. Crude but very simple and effective, I've used it on inductors with Isat exceeding 100A and the results are quite nice.

Note that this may cause significant remnant flux in the inductor core, so you may want to try the experiment multiple times while flipping the current direction in the inductor.

 

[bowman1710]

That's quite a big transistor.

Yes I needed that because of the temp rise, I used something smaller but I just had issues with a serious case of thermal runaway. Is there no way I can work around this Cob issue, using other in parallel/series etc??

 

[D.A.(Tony)Stewart]

Did you use a PNP current source to supply bias?

 

[mtwieg]

Yes I needed that because of the temp rise, I used something smaller but I just had issues with a serious case of thermal runaway. Is there no way I can work around this Cob issue, using other in parallel/series etc??

Find a device in a high power package (like TO247) but with a low capacitance. Heatsink it thoroughly, with forced air.
Pulse the bias current briefly.
Lower the dropout voltage on the transistor.
Use feedback to regulate the current and prevent thermal runaway. An LM317 can be made into a 1A current source easily, for example. No idea what its output capacitance is, though.

 

[bowman1710]

Did you use a PNP current source to supply bias?

Yes i did the MJL4302AG is a PNP version, sorry i looked i put npn on my other post.

Find a device in a high power package (like TO247) but with a low capacitance. Heatsink it thoroughly, with forced air.

I think I will have to go for something like that if I want this to work