Highly resistive windings and true fundamental inductance

Thread Starter

BulbChangeExpert

Joined Mar 26, 2016
54
Hi
i have an LC100-A inductance meter, the bare board you buy like off ebay, it's not so bad and it seems consistent enough in precision etc... while trying to measure the inductance of a small transformer primary (like ~5Watts) i noticed that the more the winding has parasitic resistance (like 500-2500ohms), the more is the inductance result, the fact is that doing calculations to math-out the turns-per-volt etc. gives way wrong results taking into account the inductance shown, quite likely the resistive factor of the winding alters the measurement, doing the same with large transformers gives predictable outcomes instead

i also noticed another thing (maybe to warn others) measuring the inductance on a low voltage side of an hi-v transformer (a flyback is the perfect example) you can obtain like 0 or varying low results ... i was puzzled at first but then i touched the secondary side and it started to show some value, apparenlty there's an hi-voltage generation on the secondary and this messes up the "apparatus" some way, honestly trying to short the secondary can be a bad idea to have a correct reading, transformers are a mess of mutually influencing parameters... is it true?

the question is (regarding the resistive thing)... is there a formula or something that having values of like...
- apparent inductance
- real parasitic resistance
- frequency of reading (the meter shows that also)
...can calculate the actual inductance?
 

Janis59

Joined Aug 21, 2017
1,834
There is huge difference in transformer theory, do the measurement happens at shortened secondary or opened. One case You see solely leakage inductance, while other see it in series with winding inductance. And vice versa when measuring from rear toward front - what is all salt in exact metering any SMPS transformer before try it to kill in-vivo.
 

wayneh

Joined Sep 9, 2010
17,496
the question is (regarding the resistive thing)... is there a formula or something that having values of like...
- apparent inductance
- real parasitic resistance
- frequency of reading (the meter shows that also)
...can calculate the actual inductance?
The inductance is the slope of the line plotting impedance against frequency. The intercept of that line, at zero Hz, is the DC resistance. Impedance, the apparent 'resistance', is probably what your meter measures. Your meter may be 'assuming' the intercept is fixed at zero, and so whatever impedance it sees is totally due to the frequency and inductance. That's an OK assumption in many cases (high frequency work, where the impedance is completely dominated by the inductance) but falls apart in other cases (at low frequency, where the DC resistance can dominate).
 

Thread Starter

BulbChangeExpert

Joined Mar 26, 2016
54
The inductance is the slope of the line plotting impedance against frequency. The intercept of that line, at zero Hz, is the DC resistance. Impedance, the apparent 'resistance', is probably what your meter measures. Your meter may be 'assuming' the intercept is fixed at zero, and so whatever impedance it sees is totally due to the frequency and inductance. That's an OK assumption in many cases (high frequency work, where the impedance is completely dominated by the inductance) but falls apart in other cases (at low frequency, where the DC resistance can dominate).
yeah yeah, i see, honestly math is not my EXACT thing, but i tried a pair of calcs, apparently the problem now lies at the meter behaviour (that we don't exactly know), i tried to calculate the inductive side of the winding starting from a low voltage (low resistance) side, on a random transformer with a stray resistance of 450ohms on the primary, the inductance should be like 500mH calculated for the hi voltage side, the meter says 6H... but the calculated\supposed reactance of 500mH-at-200Hz plus 450ohms, then re-made into inductance didn't coincide to 6H, not at all.... mmmhhhh

there's a chunky poly cap on the meter probes posts, it may be there especially for inductance measurement

i think i'll update something more when my function generator arrives, maybe it'll help also on the scope, this meter may not be enough to see the full picture... or maybe i'm getting the whole thing wrong
 

MrAl

Joined Jun 17, 2014
11,389
yeah yeah, i see, honestly math is not my EXACT thing, but i tried a pair of calcs, apparently the problem now lies at the meter behaviour (that we don't exactly know), i tried to calculate the inductive side of the winding starting from a low voltage (low resistance) side, on a random transformer with a stray resistance of 450ohms on the primary, the inductance should be like 500mH calculated for the hi voltage side, the meter says 6H... but the calculated\supposed reactance of 500mH-at-200Hz plus 450ohms, then re-made into inductance didn't coincide to 6H, not at all.... mmmhhhh

there's a chunky poly cap on the meter probes posts, it may be there especially for inductance measurement

i think i'll update something more when my function generator arrives, maybe it'll help also on the scope, this meter may not be enough to see the full picture... or maybe i'm getting the whole thing wrong

Hi,

When using pure math you would calculate various voltages and currents and then solve for the inductance. But using a simulator you could set it up with a rough model of a transformer with an inductor, then vary the inductance of the inductor until you see the same resutls as you measure. You could probably measure the DC resistance so you would know that for sure, and then you'd only have to vary the nductor vaue until you see results similar to actual real life measurements.
 

Thread Starter

BulbChangeExpert

Joined Mar 26, 2016
54
using a simulator
thought about that as well, also tried an attempt, the problem is that (again) the thing is more complex than expected, there still is some factor that misses, and also transformers on LTSpice are very complex to characterize fully, maybe also capacitive coupling between primary windings can play a role(?)

honestly all of this is useful to me to rewind small transformers for like new secondaries for power supplies or small tube amplifiers (but i don't second the convinction that they sound better:D)
 

MrAl

Joined Jun 17, 2014
11,389
thought about that as well, also tried an attempt, the problem is that (again) the thing is more complex than expected, there still is some factor that misses, and also transformers on LTSpice are very complex to characterize fully, maybe also capacitive coupling between primary windings can play a role(?)

honestly all of this is useful to me to rewind small transformers for like new secondaries for power supplies or small tube amplifiers (but i don't second the convinction that they sound better:D)
Hi,

Well ideally you have to model the core material too. The problem with low level measurements is that the core BH curve rises slowly at first until the current rises somewhat and then we start to see the normal permeability that plays a large part in the inductance.
You'd have to use a simulator that can model the cores too and you'd have to know the characteristic of the core you are using.
Probably have to do some scope work or start using higher currents for measuring.
 
Top