PCB coil Inductance measurement

[Shotky]

Hello everyone,

I have a PCB coil ( 32 layer, each layer has 8 turns) and the outermost dimensions of the coil are 600 mm x 4.5 mm, resistance is around 11k Ohm.

I would like to measure the total inductance of this coil, I tried with an LCR meter in parallel configuration but it is showing a negative value of L at any frequency until 100k Hz.
Can someone please help me understanding what I'm doing wrong?

Thanks

 

[D.A.(Tony)Stewart]

Hello everyone,

I have a PCB coil ( 32 layer, each layer has 8 turns) and the outermost dimensions of the coil are 600 mm x 4.5 mm, resistance is around 11k Ohm.

I would like to measure the total inductance of this coil, I tried with an LCR meter in parallel configuration but it is showing a negative value of L at any frequency until 100k Hz.
Can someone please help me understanding what I'm doing wrong?

Thanks

What is the self resonant frequency or pulse resonant f.

Also, important are trace width and space as well as trace thickness to some extent.

 

[FvM]

Planar coils have relative large interwinding capacitance, well possible that SRF is below 100 kHz. I would try with 1 kHz measurement frequency.

 

[G4BCH]

Assuming the resistance is measured at DC 11k seems very high resistance for an inductor. For 256 turns that's 43 ohms per turn, something is wrong here, are the connections between layers OK?
If the tracks are thick you may be measuring capacitance between layers.

 

[D.A.(Tony)Stewart]

I presume your layers in series y/n?
What were your expectations ? Design Specs? 50uH per layer?

Check for via faults with continuity , hopefully not BBV's

cite your design reference, e.g. http://smirc.stanford.edu/papers/JSSC99OCT-mohan.pdf

If this was an ESR series measurement with high R for a design of 2.4GHz antenna, interlayer capacitance was your big oversight.

 

[Shotky]

I presume your layers in series y/n?
What were your expectations ? Design Specs? 50uH per layer?

Check for via faults with continuity , hopefully not BBV's

cite your design reference, e.g. http://smirc.stanford.edu/papers/JSSC99OCT-mohan.pdf

yes they are in series
copper tracks are 50um width and around 12 um thick, distance between tracks is 100 um

I expected R around 10 kOhm, I have performed also a magnetic calibration of the Area, it seems there are no shorts between layers

 

[Shotky]

Assuming the resistance is measured at DC 11k seems very high resistance for an inductor. For 256 turns that's 43 ohms per turn, something is wrong here, are the connections between layers OK?
If the tracks are thick you may be measuring capacitance between layers.

yes tracks are very thick, around 12um

--- Updated Sep 15, 2021 ---

Assuming the resistance is measured at DC 11k seems very high resistance for an inductor. For 256 turns that's 43 ohms per turn, something is wrong here, are the connections between layers OK?
If the tracks are thick you may be measuring capacitance between layers.

around 40 Ohm per layer is ok, connections between layers should be ok

maybe I can try with a VNA to obtain L

 

[FvM]

Right, resistance value around 11k is expectable for 50 µ traces (about 38 µ effective width with usual etch factor). I wouldn't have thought that you are using so high density.

Did you estimate interwinding capacitance and self resonance frequency?

Is your VNA lower frequency limit low enough to show self resonance frequency?

 

[D.A.(Tony)Stewart]

a negative value of L means a positive value of C at same |impedance| but opposite phase

yes tracks are very thick, around 12um

--- Updated Sep 15, 2021 ---

around 40 Ohm per layer is ok, connections between layers should be ok

maybe I can try with a VNA to obtain L

You can measure L easily with any CMOS inverter with known shunt caps as an oscillator.

 

[FvM]

If winding capacitance has a significant effect on total coil impedance in frequency range of interest, it's not possible to determine the coil inductance by a single frequency measurement. You can either choose a significant lower frequency as suggested in post #3, or derive L and C components from a multi frequency measurement.

A VNA measurement is a possible way, however the large ratio of DC series resistance to 50 ohm doesn't promise much accuracy.

By the way, what's the intended operation frequency of your coil?

 

[D.A.(Tony)Stewart]

Unfortunately, your coil must have a lot of work and expense but can serve no purpose as it is too lossy (resistance) with tracks so narrow and undefined L/R ratio and L/C ratio. Perhaps if you share your intent, we can help.

 

[FvM]

Unfortunately, your coil must have a lot of work and expense but can serve no purpose as it is too lossy (resistance) with tracks so narrow and undefined L/R ratio and L/C ratio.

Strictly speaking, you can't know without an operation frequency specification. But the 100 kHz LCR measurement frequency mentionend in post #1 gives rise to doubts.