Find LC resonance hidden by inductance

Now that we know the 'LC' resonance is actually mechanical, that's a bit puzzling. The 50mH-range electromechanical systems I can think of are lightly mechanically damped.

(There's a black box surrounding your black box, and you have the advantage on us of knowing what's inside it.)

It might be easier to excite the resonance electrically, and monitor for the mechanical resonance mechanically, with a microphone or transducer.

Cheers, James Arthur

Large variable capacitors get difficult, I've always wondered about putting the largest air variable I could find into an oil bath. I see Glycerol (wiki says it glycerine, but maybe not) has a dielectric constant of 42 vs. 1 for air. I haven't scoured the list, may be something better. Then you may also need to parallel it with lots of with polystyrene caps. Mikek

We don't know if this is a product, a test-jig, a one-off, etc. If this were a product, a variable cap could be cobbled with an array of capacitors, switched in D/A fashion.

At 360Hz, he needs ~3.9uF to series-resonate 50mH.

Cheers, James Arthur

Slap it hard and watch it ring,

feed it with a current source then once it's stable, open the feed and measure the voltage, first there'll be a spike as the bare inductor dumps its energy and rings down against parasitic capacitance, after that subsides there should be a lower frequency signal from the buried tank.

OK no more ideas until you tell us what the mechanical resonance is. (BTW the "size" of the dip depends on both Q and coupling, you can measure Q by the width of the resonance and then guess the coupling.... I guess I shouldn't have said that :^)

George H.

It's nice to swing around that last 0.1uf with a knob while watching for a peak. :-) Mikek

I've got a flexure mount holding an optical grating, driven by a piezo that resonates at ~3k Hz.

Maybe some magnetic/ solenoid/ mechanical thing.

I've got some silly ideas, but I can't share till I know what it is . :^)

George H.

I think I know (p=.75), but it's Klaus' secret, not mine to spill.

You sure do get to play with some fun toys...

Cheers, James Arthur

Yep. A 5uF variable looks good in your collection too, sitting next to your

Cheers, James Arthur

I think swinging around large sacks of $100 bills is nice too... but I don't seem to have the time to do it.

Let me try it as follows. You have a large inductance in series with the LC that you want to investigate. So ideally you just don't want to 'see' that large inductance.

If you have a good quality programmable current source then you can inject any current you like into the large inductor, which then also goes into the LC that you want to investigate.

Disconnect the current source, then current-less measure the voltage with a high impedance device. This way you get the voltage response of the LC.

In fact it's a step response measurement which doesn't involve the large inductance. In fact I think someone here already proposed the step response, so this post of mine means that I go with that poster.

joe

I nearly missed reading this part; yes, that's a GREAT idea, if the inductor isn't self-shielding, because resonance will sharply peak the flux (and flux leakage) One can get fancy, and note that the series inductor will have B in phase with the drive current, and the resonant inductor will be 90 degrees out-of-phase with that...

Oh; it's ACOUSTIC leakage. Contact microphone, then.

I have looked into this. The resonance coupling is partly through L1 as well as the original schematics I linked to, so L1 and L2 are partly coupled

Cheers

Klaus

ote:

t in series

records something like 15% change of the current at resonance

pdf?dl=0

arge inductance in series. The resonance I need to detect is at 360Hz

robust way)

e small

ll stand better out

nknown inductance?

l...hang a

onnect a

n impulse.

econd capacitor.

whole

see how the

onents in

vy :(

sible to

le if the

llowed by

be wrong

E1.pdf

and

hows the impedance did not go to infinite, but there is some extra parasiti cs in the model which results in only a small dip on the curve

ohms

ou've

ks

small

idea

ble node.

think.

e feasible.

nical resonance I have not access to.

htly

ge

I would like to explain exactly what it is, but I cannot, since our competi tors might be lurking around here also. I can say it's a big mechanical sys tem, weighing more than 100kg

the

Yes. We can easily measure it that way, but if the system is in an environm ent with other systems that has vibrations we might just end up measuring o n that system instead (and precise vibration sensors are expensive, I am tr ying do do it in cheap electronics)

Cheers

Klaus

This is a product

Cheers

Klaus

Yes, that was my first idea

In the schematics I posted I had a switch which turned on a current into the load and released it, but I didn't get good results right away. I think I need to go back and dig deeper into that one

Cheers

Klaus

I am confused by this. You said L2-C1 is a mechanical resonance, but you are also saying that somehow L1 is coupled to L2 by mutual inductance? I'm not following this at all. What is the "thing" that constitutes L2-C1?

I guess there are other questions I have. Why do you need to measure this resonance? Is it an important part of the design? To make sure it is not in a bad frequency that would make something fail? Can this resonance be damped to get rid of it?

What part is the product?

Is it the measuring system you are designing, or the complete black box plus a system to measure the internals?

Mikek