Hello again,

10uH is a little small but not that small, but why dont we look at the circuit you are using.
Can you draw the circuit you are using? Show scope probe connection too.
A picture or two might help too, of the entire setup. Of course they have to be clear so we can see all parts and wires.

 

Yeah sure,

The development of the probe setup is being outsourced right now, he's developing a probe setup with BNC connectors directly to the probes. For now i'm measuring with a breadboard and BNC cables. Either way, even when i put the capacitor and inductor directly on the BNC output of the function generator i get the same results. From which i can assume (i guess i can) that the breadboard and wires that are used right now are not of big influences. (see photos)

https://ibb.co/gPuPCJ
https://ibb.co/iB5xXJ
https://ibb.co/bZn0md
https://ibb.co/iRpD6d
https://ibb.co/n520md
https://ibb.co/cTsNzy

 

Yeah the datasheet says 1 percent. And the capacitor is 10%, sorry my bad.

These are the components:

https://nl.mouser.com/ProductDetail/810-FA20X7S1H106KRU6
https://nl.mouser.com/ProductDetail/807-4470-13F

I indeed couldn't imagine the influence could be that big. I just don't see what's wrong here. Another question, if i use crocodile clips for example and make sure the components are hanging free in air. Would this have influence? Right now i have the components just straigth plugged into the bnc connector.

I'm having a hard time to gain insights in these parasitic properties.

Inductor value is 10 uH 1%

Hello again,

Nice pics. Looks like a decent setup there. Note i am quoting a previous older post of yours however, the one with the part links.

Just one problem now, and that is that the data sheet that comes with that inductor part shows inductors with 10 percent (ten percent) not 1 percent (one percent) tolerance for the 10uH value.

The "Mouser" page reads "1 percent" (one percent) which could be a typo and should have been 10 percent. I've seen this happen time after time so that may be the problem. Vendor writeups do not always match the data sheet.

I think you should go back to the part you are using and double check this tolerance also just to be sure.
If that does turn out to the the case, if you can get a 22uH value you could check to see that the calculation you are doing gets BETTER by a little bit because the 22uH (and others) are 5 percent.
As i said though go back yourself and double check this to make sure we now know the right tolerance, i could be linking to the wrong data sheet other than the one for the part you are using. I did follow the Mouser page though, so let's see what happens. Let me know as i am curious now too

If the tolerance of both L and C are off by 10 percent in the same direction then we will see a frequency that is 10 percent off. That is hard to believe that could be the case here though.

 

Thanks for thinking with me, really appreciate it!

Yeah yesterday i noticed this as well, very happy to have found the problem at first instance.

But then i saw it's the F-type which indeed has 1 percent tolerance.

i'll let you know if i find any more irregularities or solved the problem!

 

Thanks for thinking with me, really appreciate it!

Yeah yesterday i noticed this as well, very happy to have found the problem at first instance.

But then i saw it's the F-type which indeed has 1 percent tolerance.

View attachment 155566
i'll let you know if i find any more irregularities or solved the problem!

Hello again,

Ok, then do you have the series resistance for that particular part too? We need that.

I see now this is a parallel circuit, from your breadboard. That means the series R is important because that might through the resonant frequency off quite a bit.

I see for the part on Mouser that the series resistance is 15 ohms, is that right or is it 1.5 ohms or 0.15 ohms or something?

 

Oh i didn't knew/realise. The ESR of the inductor is indeed 0.15 Ohm.

How do i take this into account within my calculation? or is it better to switch to a series lc-circuit?

TIA

 

Oh i didn't knew/realise. The ESR of the inductor is indeed 0.15 Ohm.

How do i take this into account within my calculation? or is it better to switch to a series lc-circuit?

TIA

Hello again,

Oh it is just 0.15 ohms? That is probably not significant.

I can give you a new formula, but i need to know for sure how you are driving this L and C.
It looks like you have the frequency generator connected in parallel to the L and C, which are also in parallel.
To see a peak or dip, you should put a resistor in series with the LC combination. That way you are looking right across the L and C and the Rs helps to see the dip or peak.
The frequency generator probably has some internal resistance, so it may not matter, but we need to know what that resistance is really i think. Maybe a 10 ohm resistor would help here.

The resonance we will be looking for will result in a peak or dip.

[LATER]
Did a quick calculation, it does not look like 0.15 ohms is significant.
You could double check the series resistance of the inductor with the ohm meter.

 

Hi,

Probably a stupid question, but how can i switch between connecting the function generator parallel and in series to the tank circuit? For this moment i don't really understand how you can switch between these two states, when all other components are already parallel.

Will try the resistor in a moment, see if this results in any changes.

I checked the ESR and the ohmmeter gives 0.3 ohm which as i understand is not sufficient.

Thanks

 

Frequency generator has an impadance of 50 ohm fixed btw.

 

Hello again,

Oh 50 ohms, that's good.

You can not connect a generator in series with a parallel RLC. However, you can insert a resistor in series with the generator and then the resistor is also in series with the parallel RLC. Since we know the gen has 50 ohms internal resistance, that helps.
Unfortunately i still dont see any reason why the measurement should be as much off as more than 11 percent. 11 percent would be the extreme if the inductor can cap where both off as much as they can be, and that would be hard to believe. The resistance of 0.15 ohms isnt enough to cause much difference, maybe around 1 percent.

Is the sine wave generator really putting out a clean low distortion sine wave?
Just guessing at this point.
Could the generator have some output inductance or capacitance too.

If you want to try a series circuit, connect L in series with C then in series with the generator.

 

Hi there,

i'm a big noob and in love with this forum.

At this moment i'm willing to measure capacitance with the highest accuracy possible. I'm gonna do this with an LC circuit and the material available (function gen. / oscillloscope / multimeters and so on). I've got the following inductors:

2.2 uH / 1% / 0.04 ohm
10 uH / 1% / 0.15 ohm
1 mH / 3% / 33 ohm
10 mH / 5% / 72 ohm

I'm willing to measure capacitors in a range of 100 nF to 10 mF with 5 percent accuracy.

For another part of my measurement setup (4 points resistance measurement with probe holder etc.) i've used the inner connections of BNC cables for all leads. In this case i can minimalize parasitic inductance/capacitance. At the measurement devices i short one of the inner connections with the outer shield so they come together in one 'connector'.
The capacitors to be measured are probably inkjet printed. So these will be measured with the same probe setup as well. I can simply raise the frequency of a sine wave to look for a peak in the amplitude of the voltage on the scope. In that way i can find the resonance frequency from which i can determine the Capacitance with a known L. My function gen. has an impedance of 50 ohms. I hope i've been clear enough about my situation, i'm a little bit in a hurry with all this stuff.

Now i've got a few questions:
- Is it in my case better to use a series or parallel circuit for this measurement setup? I know i can find a peak in the amplitude with a sweep in both cases. But which one would result in better accuracy. I also don't really understand what the influence of a resistors in series with a parallel tank circuit is.
- I find it difficult to approximate my accuracy. I don't know how to take into account the inaccuracy of looking for a peak on the scope. Is this a few hundred Hz, a few thousand?

The way i'm looking for this peak is setting the timebase very large so that the sine appears as a big dense line. Then i put the volts per division as high as possible and sweep manually around my resonance frequency until i found the best matching frequency (or so the highest peak visible).

- These inductors i've got have axial termination so i have to use a breadboard to develop this circuit. Can anyone help me how much influence things like this will have on my measurement accuracy. I can imagine that although i use BNC cables for the connection from the breadboard to the capacitor, these make no sense at all when using a breadboard and a inductor with those large leads going trough the breadboard. These must bring a lot of parasitic properties with them.

I will post some pictures of my measurement setup as well. Thanks in advance!!!

 

https://ibb.co/msNTxT

 

- Is it in my case better to use a series or parallel circuit for this measurement setup? I know i can find a peak in the amplitude with a sweep in both cases. But which one would result in better accuracy. I also don't really understand what the influence of a resistors in series with a parallel tank circuit is.
- I find it difficult to approximate my accuracy. I don't know how to take into account the inaccuracy of looking for a peak on the scope. Is this a few hundred Hz, a few thousand?

Without doing error analysis I would bet on parallel. That way Fgen looks like a real voltage
source to the network as you approach resonance.

Accuracy. mordern DSOs have peak detect. also cursor measurement. Maybe best way is use one trace as a horizontal cursor
and position it at top of waveform such that you can see waveform x either exceeds or is lower than trace of second channel
acting as a cursor. Eye may be better equipped to see this than just one trace on screen guessing what is peak.

Regards, Dana.

 

Thanks!!
Sorry for misplacing my last post.
Yeah indeed, right now when i'm close to the peak i change the position of the peak to the center of the screen. That way it's easier to notice small variations. I still got the idea this whole approach is not very accurate, but gonna test tomorrow morning.

Do you maybe got any idea about the following?

-These inductors i've got have axial termination so i have to use a breadboard to develop this circuit. Can anyone help me how much influence things like this will have on my measurement accuracy. I can imagine that although i use BNC cables for the connection from the breadboard to the capacitor, these make no sense at all when using a breadboard and a inductor with those large leads going trough the breadboard. These must bring a lot of parasitic properties with them.

With my short timeframe i'm not able to connect the inductor in a better way...

TIA

 

One small question.

If i'm testing the maximum current a conducting path can handle until it becomes destructive. Would there be a difference between AC and DC?
As resistance is purely resistive and has no complex part?

 

Have you tried using the scope in X-Y mode on the series RC circuit? Measuring the voltage across the resistor on the horizontal axis and the voltage across the capacitor on the other. Hopefully you can choose a frequency such that the reactance of the capacitors swamps its effective series resistance?

Another way might be to configure a well-calibrated current source and monitor the voltage on the capacitor as it charges. The ESR can be removed by using the voltage after switching the current to zero. This would also let you get a reading for the ESR and the leakage resistance, as well as a curve showing the linearity of your capacitors.

 

I have heard of a method that involves series resonance to measure capacitance (or even inductance). You just need to do a frequency sweep and see where the current is greatest. The nice thing is, ESR doesn't affect it at all. The one thing is, if L and C are too small, you get issues due to higher frequencies.

 

I have heard of a method that involves series resonance to measure capacitance (or even inductance). You just need to do a frequency sweep and see where the current is greatest. The nice thing is, ESR doesn't affect it at all. The one thing is, if L and C are too small, you get issues due to higher frequencies.

I don't know how this would help you find the capacitance. In order to work, it would require that all capacitors of a particular value have the same effective series inductance, regardless of the type of capacitor. That's seems completely unlikely.

 

Lead length inductance calculator -

http://www.consultrsr.net/resources/eis/induct5.htm

https://www.edn.com/electronics-blo...ing-wire---loop-inductance--Rule-of-Thumb--15


Regards, Dana.

 

I don't know how this would help you find the capacitance. In order to work, it would require that all capacitors of a particular value have the same effective series inductance, regardless of the type of capacitor. That's seems completely unlikely.

If you use a large inductance you can probably offset the ESL. And maybe you could measure the inductance with calculations and looking at mutual inductance.