LCR Meters - Inductance Mesurement Frequency

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rodabod

Well-known member
Joined
May 12, 2005
Messages
2,896
Location
London
Hi,

I'm just wondering why LCR meters often offer different frequencies for measuring inductance? I've used modelswith different frequencies (eg. 100Hz + 1000Hz) but have just bought a cheap model with a fixed frequency of 175Hz. It does have a wide range though.

I realise using different test frequencies gives different results, but why? I assume the meter works using the principle of Z=jwL, or is there something more complex going on?

Either way, when do you knwo which value to take when you have tested with more than one frequency?

Roddy
 
rodabod said:
Hi,

I'm just wondering why LCR meters often offer different frequencies for measuring inductance? I've used modelswith different frequencies (eg. 100Hz + 1000Hz) but have just bought a cheap model with a fixed frequency of 175Hz. It does have a wide range though.

I realise using different test frequencies gives different results, but why? I assume the meter works using the principle of Z=jwL, or is there something more complex going on?

Either way, when do you knwo which value to take when you have tested with more than one frequency?

Roddy
Ther are two main reasons: One is that, if the inductor has a core, the permeability of the core varies with frequency. The other reason is that parasitic capacitance is seen by the inductancemeter as a reduction of inductance, and the fhigher the requency, more its effect is measured.
I have worked for a long time with RM cores for graphic EQ's. The procedure for inductance measurement used several different frequencies, chosen to be close enough to the band centers.
So, to answer your question, the right value is the one measured at the frequency where it will be used!
 
I measure inductors at the frequency of interest using the voltage divider method (with a calibrated decade resistor box). Then I just convert that reactance to an inductance figure. It's really less hassle than it sounds, especially if you use a handy Java calculator like this:

LINK

Digital LCR meters are cool for quick sorting, but I use the method above for an exact figure.
 
And let's not forget electrolytics... most of your uF's are gone at higher frequencies
(well, not at 1kHz yet, but more expensive meters do test-freqs of 100kHz...10Mhz as well).

It's a potentially confusing subject indeed, good to have it 'documented' here in this thread.

Bye,

  Peter
 
Thanks for the suggestions. I've now also remembered that signal level comes into play with transformers due to the core permeability changing.

Now to start testing these low-inductance Lundahl LL1517s.....
 
looking for bits and pieces to this simple but complicated question.

1000 hz has been a standard from way back, probably because mid range and voice are around there, and telephone inductors were really all you cared about, nobody was checking Henries of power transformers once the lam company told you that if you take core X and wind n turns, you will get 10 Henries for your choke.
Since you only need a minimum value to pass QC, not an exact value, a precise measurement was not really needed.
More important might be he hi-Pot test.

With RF inductors, that is a different story.
They are probably very hard to measure back then because of stray C.
Your leads and your Standard Inductor will have bad stuff that makes them in accurate.
But, usually RF coils are air inductors, which are failrly easy to calculate since they are usually wound in a precise fashion, on a form whose diameter can be used with the turns to get a pretty good ballpark figure.
Then it gets tweaked on the chassis so all the other leakages can be compensated, so not a real big need to check the inductance of a 10 turn coil on a 1/2 inch tube at 100 KC, it will always be 4.7 uH or some number.


So that just leaves Voice EQ, so that is where those instruments are probably aimed at.

Here is part of the problem
Since the frequency subtraction is quadratic, you do not want to square a million hertz and subtract 60 hz^2, much better to square 1000 hz and hope for the best.


11s2wb8.jpg



 
Back
Top