An attempt to make an analog capacitor ESR meter

Hello!

This is my first post here, and I hope I won't break any rules :smile:.

Anyway, I was trying to build an ESR meter from this LINK, and I was hoping that someone could help me and guide me on how to solve the problems I've encountered.

My current setup:



I didn't have the same components, so I made a few modifications: I replaced op amps with LM385 dual op amp, I used xenon flash trigger transformer (HV side as a primary), and a [0-1]mA analog ammeter.

The ammeter is the one that is giving me the headaches, I need to amplify the signal, and I did manage to do it, empirically, but I'm not satisfied with it. The output I get is not linear enough, and the scale saturates already at 5 Ohm input impedance.

I was hoping that someone could point me at which parts of the circuit should I pay attention to get what I want, since I have really little experience with AC signal processing.

What I did (my schematic):

-I decreased C5 and C7 from 10u to 1n. Made no sense to me, since Xc=1/(2*Pi*f*C), and lowering the C would attenuate the signal, but it worked - I didn't get the detection until I lowered the C significantly;
-I messed with the amplifier gain - settled at 33k (Av=30), for some reason. Giving it lower gain - meter wouldn't go full scale, and giving it a bigger gain - meter wouldn't go to zero with leads shorted;
-I also messed with the output resistor/pot, but that's irrelevant to the problem (I think).

I would like to make the scale somewhat more linear in the range of [0-10] Ohm, and ideally, I wouldn't want to use any more op amps.

Is that doable? Any suggestions? :smile:

The trigger transformer seems like it would be the biggest problem. I would put resistor values from 0.1 to 10 ohm on your test leads and watch the 200mV square wave with an oscilloscope and see if it changes.

Also look closely at the pictures on Ludens site. Both meters have new scales made of paper and fastened on the meter face. Also notice they are not linear.

I was hoping it wouldn't come to it, finding a gapless core is not an easy task. I thought I got away easily.

Unfortunately, I don't have an access to an oscilloscope.

Could you tell me why do you think that transformer is the problem?

I know it's not linear, but his meter does not saturate as fast as mine. His scale goes all the to the 20 Ohms, my scale won't detect anything greater than 5 Ohms.

The ignition transformer will have a lot more turns than yours. Typical 400vdc in and 8000 volts out, it seems like a huge mismatch and if it has a gap it is a completely wrong type.

USB scopes at $56 USD and up are usable. Here is a Ebay ad.

"Hantek 6022BE PC-Based USB Digital Storag Oscilloscope 2Ch 20MHz 48MSa/s TS 35BS"

Different oscillator frequency? How come?

The shop didn't have TL062, and no one sells signal transformers. I could order these for china, and I probably will eventually, but it'll take a lot of time to arrive, so I tried to do with what's available.

The output signal is somewhat useful, since I do get different values for measuring different capacitors, but the (1) is what I get, and (2) is what I want.

(Paint, you get the idea...)

So, it is doing what it should, but not on the desired range.

Your brief summary is not ok, since I was trying to adjust the whole thing to what I have in the first place (if possible).

I'm quite inexperienced with op amps and AC signals, and I know there are people who know stuff, since every google search finds these forums. All I wanted is to know whether it's doable with this setup, and if yes, what parameters should be tweaked and how.

It's not only the transformer. You also have different OP, apparently different oscillator frequency, different capacitors. I don't believe that you get any useful output signal with C5 of 1 nF.

In a brief summary: You changed almost everything and wonder why it's not working.

- - - Updated - - -

Different oscillator frequency? How come?

Click to expand...

O.K., it's not very different, about 60 kHz with TL062 and 40 kHz with LM358 according to simulation. Main reason is low LM358 slew rate.

I thought about the coupling capacitor point. When you observe it doesn't work with 10 µF at all, this means that the lower main inductance of the trigger transformer is loading the OP output too strongly, most likely stopping oscillation completely.

Thank you for your reply.

What simulation program are you using? I've had no idea that slew rate affects it that much. I've just checked the datasheet, I see now that it's a bad choice (assuming that I'm looking at the right thing ).

The R part of the transformer primary's impedance is 82 Ohm, no idea about the L and C parts. I don't have the part's datasheet, but I know that it should have a bunch of turns, so the inductance shouldn't be very low?

So, would adding a buffer at the output help? Or perhaps a resistor/impedance? Also, I have a few LF356s in my box, would they be more suitable for the oscillator part?

try to build this one : **broken link removed**
it uses standard components (very close to the one you're using)
and it does not have fancy difficult to find transformers ...
and it works very well as an esr meter !
you should'nt need an oscilloscope to make it work (although a scope is a nice tool to have like the esr meter)