I am trying to build an Inductivity meter from websites.

Not much remains from the original schematics.
Except a comperator is used as oscillator.

Most of these schematics on the web all seem to be the same for the oscillator- a LM311, some 100K resistors, as well one 47K, and 2x 10uF caps.

It does not work at all with the 10F200 comperator.

Which in turn goes to T1CKI of a larger PIC. I am setting the TMR1H to 0xff each interrupt, so I am getting 1:256 divider right now. Resulting in audible signals. Each time I toggle one I/O for a small piezo speaker.

Here the schematic. Only using 2 components, a resistor, and one capacitor.

I have tested 22uH and 680uH so far, both result in oscillation. That is about the range I'd be interested in.

Smaller inductors don't produce oscillation for instance one toroid seem to have only a few uH. But, they change the freq. of course if they are put in series with other inductors.

What to use as formula is not clear to me.

Seems to be I have to test a series of inductors, get the numeric frequency on a display, and apply calculus maths to get the formula.

Maybe it is the same as for the original circuit only in a different form.

No idea what the relay should be good for. Except it could be used to add Lx in series to a fixed inductor, and shorten it for calibration.

Still a lot of work until I get a numeric display, but I made it oscillating.

Most of the components are unneccessary however unfortunately it needs a stable voltage or the frequency will change. Maybe I'll use a dc/dc booster with an AA battery.

Here the original schematic:
http://electronics-diy.com/lc_meter.php

I don't need a cap. meter since this is built into my DMM.

Any ideas for a good coil based oscillator?

I was thinking to try a TTL inverter but can not simulate it properly in LTSpice.

 

The 4.7k and 47nF are in no way based on calculations. I had these around here. Maybe the values could be changed.

I have not used comperators before.

 

The basic approach, is after you get your oscillator working correctly is to write a frequency measurement routine. That frequency measurement can be the used to calculate the inductance.

That comparator-based oscillator is simple and works very well with high Q inductors, but huge errors occur when measuring inductors with high resistance. Let me know if you want Chris Krah's formula that shows the effects resistance has on frequency.

Good luck with the TTL based circuit, but why aren't you using an actual comparator? They are not expensive and it would be very hard to improve on the simple comparator based oscillator.

If you are going to measure inductance only, you won't need a relay; you can leave the resonating capacitor in the circuit all the time. For automatic zeroing, you can just store a constant to correct for small errors in the reference capacitor.

For a formula, if you are using the comparator oscillator, the inductance can be found with the classic formula below.

​

Its basically multiple precision multiplication and division.

Appears that you have a lot of options to work through. Hope this helps at least a little.

 

The TTL oscillator was just an idea. I have built the circuit with a 10F200, and soldered a row connector to plug in carrier PCBs.

Floating point is not a problem even if the 16F74 does not have much memory. I hope at least.

Don't have comperators here at all, sure I can order from the UK and it arrives within 2 or 3 days. Except some 10F206, but they do not work with 100K resistors. Also have the original circuit on a carrier PCB.

Yes if you have that formula available? Most of the inductors I use have fairly low resistance, only some RF inductors here with 1000uH and a few 22000uH.

The relay was already added to the circuit board, however it is a problem, the frequency drifts even if I only turn on a LED.

Eventually I want to leave 22uH or maybe 10uH in place, and put the Lx in series. The circuit does not oscillate with smaller value inductors.

16f74 has 4K FLASH, about 3.5K remaining. 3 digit LED display.

 

Here is what I got so far. I get a hex display of the frequency.
I will replace with a 4 digit display later on, not a big deal to do it.

The code is inaccurate as the display update is done in the ISR.
Also I would have to add the actual TIMER1 value (between 0xff00 and 0xffff), as the display update is done via TIMER0 each second.

I can do that later.

Also maybe I should use a higher oscillation frequency.

I don't need so much accuracy but I want to be able to know if it is a 4.7uH inductor or 2.2uH

 

​

I must say this formula is incorrect.

Also the resistor was 47K not 4.7K
I have replaced it with a 100K adjustable.
There is an operational area maybe 25K to 40K where the frequency does not change at all. Below 25K it is increasing.

I changed the cap. to 1nF (from 47nf) with little change in frequency. Even electrolytic caps work. It seems to be only working as capacitive divider, not influencing the frequency.

I have also succeeded making the circuit oscillate with smaller inductor, about 4.7uH (it is an unknown toroid). By adding a high-voltage ceramic cap (4700pF) in parallel.

I guess the coil is worked in self-resonant mode.

Changed the board as well to 4 digits.

Next I need a voltage booster to get a constant voltage for the comperator.

 

Found a comperator as well:

uPC339

http://vintageelectronics.betamaxcollectors.com/mitsubishivhsvcrmodelhs-421ur.html

Examining old PCBs and taking off some parts is a hobby of mine

It is also useful for learning how commercial electronics are designed.

 

Can you provide the correct formula?

 

Can you provide the correct formula?

Not yet. I am pretty sure it is not correct since the value of the capacitor has little influence on the frequency.

Using a 1:256 divider (PIC timer), I made the oscillations audible.

Even a 100uF electrolytic cap. works.

With the 100k adjuster set to 10K or so, it even oscillates without inductor. RC oscillator.

Then if I increase, it practically stops, about 8 oscillations/second.

Adding the inductor into the circuit then, again I get oscillation.

A few Khz * 256, so 150 to 500 KHz.

Some kind of quenched RC oscillator with added inductor.

Remember I've set the comperator to inverting mode via 10F206 "firmware" (just a few lines).

Maybe using a capacitor from the comp. output as feedback is not a regular circuit? I have not tested with 100uF cap and without inductor.