Adjustable AC voltage

[unclejed613]

there's also a safety factor here. let's say you use an AC current source capable of 2A (we'll use a test frequency of 1khz). the inductor under test turns out to be 500mH with a DCR of 4 ohms (this would be a rather large coil of course), and the capacitor is a 0.05uf cap with an ESR of 1 ohm. 1khz is pretty close to resonance with these values, and the ESR and DCR values are reasonable. testing with the components in a series resonant circuit yields some interesting (and possibly disastrous) results... at resonance, the voltages across each component cancel, so all that shows up across the circuit is the voltage developed by the resistances (ESR+DCR=5 ohms), or 10V. but the voltage across each component is the result of 2A ac through Xc and Xl, each of which are 3k ohms, resulting in 6kV at the joint in the middle where the cap and inductor are connected.... that has the potential to damage both the cap and the inductor.... that's why it's best to do these tests with a function generator rather than a power source.....

 

[fezder]

woa, nice increase at voltage, well for my own safety, i probadly should buy function generator then? any good suggestions for, well, cheap function generator for a starter, and for college budget :S?

 

[fezder]

hey, i just found this:
https://electronics-diy.com/lc_meter.php

tell me what you think about this, worth a try? seems to me to be just what i need

 

[fezder]

https://electronics-diy.com/Function_Generator_XR2206.php

but this seems better choise, it doesnt show me value of coil, but it has adjustable frequenzy/wave form, what do you think? any opinions/thoughts?

 

[unclejed613]

wow i didn't know those were still in production....

 

[fezder]

which one? and are these reliable and good for calculating/measuring coils incuctanse? any idea?

 

[MrAl]

Hello fezder,

What kind of coils do you intend to measure? If you want to use an AC waveform to test the inductance you can use a function generator, might have to add on an output booster to boost the current level a little though depending on what kind of inductors you test.

The XR2206 chip is made for sine output too, along with square and triangle. One thing i dont get is how they are spec'ing 1Hz to 2MHz, when i think that chip only goes up to 1MHz or a little over. Maybe they have a better version out now.
All the spec sheets i have found however say 0.01Hz to a little over 1MHz.

So what you should do is let us know what kind of inductors you are going to be testing, small ones, big ones, and for what purpose: power supples or RF or whatever.

Also, do you own an oscilloscope?

 

[fezder]

ok, time for little info
my coils are from random power supplies, SMPU's and such. Well my main purpose for these coils is power supplies and crossover filters, and who knows where i need them.....but i thought if i can measure them somehow, then i could use them better. And yes, i have oscilloscope, dual trace 10mhz analog type.

i hope i dont get scavengers reputation when i post this photo, just showing what cind coils i have

https://i50.tinypic.com/2v33lup.jpg

 

[MrAl]

Hi there,

Well, some of those coils can be checked with the output of the function generator such as the XR2206 chip, but others will require some higher current levels to really test the inductance. The metal core ones most likely will need some higher current levels to get the flux up high enough on the BH curve to get to the part with the normal operating permeability. Not doing that could lead to a quite incorrect measurement of the inductance.
Also, a coil used for a switching power supply is probably best tested in a circuit that actually acts as a power supply. That way you can tell what the inductance is by looking at the waveforms and also how it reacts to higher level DC currents, which is something you'd want to know as well as the inductance. You'd want to know at what level the DC current starts to degrade the inductance. You can get some idea by looking at the lead wire gauge so that might help too.
To test it outside of an actual power supply you have to get some DC current flowing and see how it reacts. You might build yourself a little test jig which is actually a buck converter, where you have some decent level of power coming in and going out, and can look at the current through and voltage across the coil. To find out what a buck circuit looks like you can look on the web im sure there are plenty of examples, but if not i'll post a schematic.

To test the little coils you can use the frequency generator and maybe a series resistor where you can check for the response of the coil with frequency. That will tell you the inductance. To use it with the bigger coils, you'd have to use an amplifier or maybe just a transistor and power supply to pump the coil with a square wave. Measuring the response of the coil to a square wave can tell you the inductance, and you can set the DC current if you use an external resistance and vary the level of the wave (using a square wave that is positive only, not plus and minus, and this kind is quite easy to generate).

So to start maybe you can build a little buck converter. It doesnt have to be regulated so it's going to be quite a simple as all heck circuit. Just a transistor or two or three for the drive, a back 'catch' high speed diode, maybe a 555 ic to provide the frequency wave, some load resistance, and a power supply that can handle the current you need to test at. This is really the simplest way to go about this and you'll see how easy it is once you do a couple chokes this way.

 

[fezder]

https://www.electro-tech-online.com/custompdfs/2012/06/f5.pdf

i found this arcticle regarding SMPS, is that buck converter the circuit you meant? and can i control the pwm via 555.....pretty stupid question, but gotta ask just for sure, as it produces positive square wave.

*edit
i should read more carefully posts
ok, so i use 555 to create pulse, but what transistor is good for this, i have 2n3055, some fets, and darlingtons, and how big current are we talking about here?....newer made these cind circuits before, only pwm circuit i have done is fan speed controller with fet and 555, and led dimmer

 

[MrAl]

Hi again,

Yes that article appears to show a buck circuit with a given topology. You dont have to stick to that exact topology, but that's the main idea. For example, you can arrange the circuit so that the collector of the transistor does all the drive work. To use a 2N3055 transistor you'd need to drive it with another smaller transistor which is then driven from the 555.
So you have the 555 which puts out a square wave, then you have a small transistor driven by the 555, then the 2N3055 driven with that. The rest of the circuit is similar to that which you see there. You need a load resistor too to draw some current so you might need a couple power resistors.

You'll want at least a 1 amp output pulse from the 2N3055, but the actual current will depend on load. If you can vary the load resistance somewhat you can control the current through the coil under test. If you have a very large coil you may want to pulse that with more current, but that will take a better power supply too...one that can handle the current obviously, and possibly a MOSFET switching transistor to handle the higher current level.

If you look at the buck circuit waveforms you can see the current ramping up and down. It's this ramping that tells you what the inductance is using v=L*di/dt, so you have to measure the ramp height and the time it takes to get to that height. The voltage v is the amplitude of the pulse. We can go into this in more detail if you think you want to build something like this.

 

[fezder]

thanks for reply again well i have irf540 fets too, can they be used in switching? or if i use 2n3055, could i drive it using tip122 darl.transistor?
in the formula v=L*di/dt, di=current change (A) and dt=time change (s) as inductanse is vs/a?

i would appreciate if you could go into more detail, i dont want to burn my components in vain

and what becomes in power supply, i have couple 12v/1.5a and one 19v/4.7A, think theyre powerful enought?

 

[MrAl]

Hi again,

I have posted two attachments, one circuit for testing the power inductors and one with the waveforms. In the circuit, D1 is a high speed diode with a current rating that is able to take the current required for the given load. Also shown is a capacitor across the power supply rails. This cap reduces the average current out of the power supply so the power supply does not have to put out the full peak current but only the average current of the load.

To get an idea how much current the component might take check out the lead wire gauge if possible.

In the circuit shown, a MOSFET is being used to do the switching and it is powered by a 555 timer ic.
If you look at the waveforms you'll see the current ramps up by 1 amp in 10us (each pulse is 10us long). Starting with:
v=L*di/dt

and solving for L we get:
L=v*dt/di

For the circuit shown and the waveforms shown we have:
v=5 volts
di=1.0 amps
dt=10e-6 seconds

Those are the three measurments we need which would be measured using the scope. Plugging them into the formula:
L=v*dt/di

we get:
L=5*0.000010/1

which of course equals 0.000050 which means it is a 50uH coil.

What happens when we increase the load to the point where the inductor starts to saturate i'll show next (see the third attachment).

 

[fezder]

hi,
thanks so much for this help, but i dont get this ''saturation'', does it saturate from the current?
do i even understand this graph right, but normally the inductors current rises when polarity of the voltage changes, in the graph it seems to me that the inductor current rises only when it changes from positive to negative, or did i get it wrong? and why d1 diode is backwards? is it to protect mosfet?

could you enlighten me more, if possible, this is quite new area to me...

i understand those calculations right thought

 

[MrAl]

Hi,

The basic idea behind saturation is that the core material reacts to current in a certain way, producing a magnetic field in response to the current level. The problem is that there is a limit to this process where at a certain level the core can no longer react the same way it did with a lower current level. When the current reaches a certain level the core stops being able to create more of a field and what this does is it makes the reactance go way down so the coil starts to look like a short circuit rather than a coil. Once it appears like a short circuit the current shoots up at a rate much higher than before which usually means something (the driver) blows out from an over current failure.

Another way to understand this is to think of the coil as a certain inductance L1 when the current is at some reasonable level. When the current goes past a certain point (although it is not that abrupt) the inductance goes down much lower. For example, say we start with 100uH at 1 amp. At 2 amps it may be down to 80uH, and at 5 amps it may be down to 20uH, and at 10 amps it may be as low as 1uH. What this does is it usually causes the external circuit to deliver more current which then blows something out. That's usually because the driver is putting out some constant (or nearly constant) voltage.

If we start again with:
v=L*di/dt

and solve for di we get:
di=v*dt/L

Now if v is constant and dt is constant and di is constant we have:
I=V*DT/L

and for a simple case where V=1 and DT=1 and starting with L=10 we have:
I=1*1/10=0.1 amps

and if the inductance drops to 1 we have:
I=1*1/1=1 amps

so you can quickly see how the inductance drop causes a higher current to flow, and that usually blows out the drive circuit.

So this is a major concern in switching power supplies and is something that we want to test for. The particular inductor either can or can not handle the required peak current. If the inductor can handle 10 amps without saturating then it might work just fine in a particular power supply, but if it can not handle 10 amps but instead deeply saturates at 5 amps then the current might go much higher than 10 amps. The other effect is it does not behave as an inductor of the required value anymore so we loose the property we needed in the first place.

For an ideal inductor, the current would ramp up in a straight line and then ramp down, so it would look like a saw tooth wave. But if it starts to enter saturation where the inductance drops, that ramp turns into a curved ramp where the current near the end of the cycle starts to curve upward. The curve is caused by the inductance dropping more and more as the current increases due to the inductance dropping, so it's a domino effect where the inductance drops, the current increases, so the inductance drops more and so the current increases even more, etc., etc.

 

[fezder]

that certainly helped me to undertsnd saturation, thanks so its wise to test inductors saturation, to know where it can be used safely without causing SC? and what purpose is that d1 backwards for? to protect fet? and one more question, what is good frequenzy to test coils, are there limit there?

 

[MrAl]

Hi,

Yes it is certainly good to know the sat current of an inductor so that you know what kind of power supply you can use it in later. If you build a power supply that requires a 10 amp inductor and you have a 5 amp, 10amp, and 20 amp inductor laying around you know which one would be best to use. If you didnt know the 5 amp was only a 5 amp'er, you might use it and blow up the drive transistor

The diode is part of the logic behind the buck converter. It conducts part of the time when the transistor is off and that is part of what delivers some of the energy to the output. If you are interested in this kind of power supply (and you probably should be at least to some extent if you want to test inductors this way) you can do some simulations using the free LT Spice program. You can then look at voltages and currents and see what's what.

The best frequency to test at is the frequency you intend to use in whatever you are going to build. You can probably get away with 50kHz for the power inductors although switching power supplies run from anywhere from about 10kHz to about 2MHz. 50kHz is a good place to start for your tester i think, but you can easily make it variable by using a pot to adjust frequency. There are 555 circuits that put out a square wave so you can start with that and do some testing and get a feel for how this all works.

 

[fezder]

ok, thanks for help again well i think ill get along fine with this ''project'' now, ill keep you up to date with my progress, this forum helped me again with my problems, thanks all
and yes, i know this LTspice software, one person recommended it for me, but i havent used it yet however.....

 

[MrAl]

Hi again,


If you try that software and get to know it you wont be sorry for spending a little time for that. You can do experiments right in the software without sacrificing expensive parts.

 

[fezder]

https://i48.tinypic.com/2h4hjll.jpg
https://i46.tinypic.com/u7bdh.jpg

hi, i assembled the circuit which you gave, but i wonder is it wrong however, say if something is screwed.
the scopes pic is taken from the coils voltage