Hi,

I've been trying to make a working LC meter for a while now, and I'm pretty close to finishing one now (at least I think). The frequency part works just fine, but the LC part won't do it's job. For the LC part I chose a famous LM311 configuration, but I was reckless enough to skip prototyping it and went on and made a board. Now it won't work of course (LC part, F works just fine). I tried to measure the frequencies with and without the calibration capacitor, and got 0Hz for both of those, so my guess is that the oscillating circuit is somehow faulty.

I've been looking at it for hours, I checked the board for cold joints and I wasn't able to figure out what could possibly be wrong.

Could someone more experienced try to find where I went wrong with the LM311 part?

This is the simplified circuit of the oscillator and frequency counter input circuit.

PS: Don't know if it's relevant, but my design heavily relies on Paul's FreqCount library and uses Arduino instead of PIC.

 

Have you poked it with a scope at all?

 

If I had a scope I wouldn't be making this Is there any other way to find out whether the circuit is oscillating?

 

Yes. Make a half-wave voltage multiplier and connect its input to the output of the LM311 and check the output with your voltmeter. You should see between 4 and 5 volts DC if the LM311 is oscillating.

By the way, that's a nice looking board.

 

Thank you sir

I did what you suggested using 100n capacitors and BAT46 diodes, and THIS is what I got, so I guess the oscillator works, right?

So, now I have a situation where frequency counting part works, oscillator works, and rotary switch in between them also works (I checked for continuity), but altogether they do not work.

Huh...

 

Something's going on and its probably oscillation. What frequency did you expect?

 

Something's going on and its probably oscillation. What frequency did you expect?

I made the oscillator circuit on the breadboard, using the same components, and it gave about 1,5MHz, but when I hook up the onboard circuit to F meter it gives 1-2 Hz. I guess I messed up somewhere on the board.

 

Yes, I think you did Your meter was flickering at about 1-2 Hz, so it looks like your oscillator is oscillating and your frequency meter is measuring. Consider that besides an wiring error, you might have a wrong component in the tank circuit. Being off by a factor 1,500,000 is a pretty large error, so a wrong part should be easy to spot. Wrong wiring not so easy.

 

What are the L and C values in your tank circuit that gave the 1-2 Hz result?

By the way, the next time you use this circuit, you can use the output of the LM311 to drive the microcontroller directly.

 

i once had similar problems before, hope you can find right answer here

 

i once had similar problems before, hope you can find right answer here

So, how did you solve it?

By the way, the next time you use this circuit, you can use the output of the LM311 to drive the microcontroller directly.

Is that a recommendation, or just an option that would work too? Because I dumped the old board and I'm making a new one tomorrow, and tonight I intend to improve the design a little. So, if it's not a must, I'd leave it this way, but if you think that the other way would be better, I'll make an extra effort.

L = 100 uH , C = 1 n

Btw, I had to reduce the 4k7 capacitor to 1k, it seems that 14k7 was too much for BF199 to sense.

 

Connecting the output of the LM311 to the controller input is an option, not a recommendation. Maybe the fact that the BF199 circuit is a potential source of problems and does not contribute anything positive to the immediate goal of getting the oscillator/counter circuits to work might be a good reason for leaving it out until the basic functions are working.

Using the formula below, I get 503 kHz as the oscillation frequency.

With this circuit, you need to use a high Q (low resistance) resonant components. Many who made L/C meters based on this circuit found very large errors occurred because they used an inductor with too high of a resistance.

Here is the relationship as determined by Chris Krah and verified experimentally.

 

So, how did you solve it?


sorry i really don't remember, it was long time ago. Try Capple's suggestion.

 

Connecting the output of the LM311 to the controller input is an option, not a recommendation. Maybe the fact that the BF199 circuit is a potential source of problems and does not contribute anything positive to the immediate goal of getting the oscillator/counter circuits to work might be a good reason for leaving it out until the basic functions are working.

Using the formula below, I get 503 kHz as the oscillation frequency.

With this circuit, you need to use a high Q (low resistance) resonant components. Many who made L/C meters based on this circuit found very large errors occurred because they used an inductor with too high of a resistance.

Here is the relationship as determined by Chris Krah and verified experimentally.
View attachment 83415

Hey, sorry for the late reply. I read your message earlier but didn't have time to answer, and I really appreciate your help.

Well I can say that working with my relaxation oscillator was everything but relaxing. But, I remade the entire board, carefully made a good connections underneath, fixed the code and now it's sensing the change in frequency just fine, and the result stands still once displayed. Only problem is - the result is very, very wrong. And I believe it's because of my calibration capacitor. It's a styroflex capacitor, and has "1000H" inscribed on it. I asked a several enthusiasts here and they all suggested I use styroflex capacitor because they have low tolerance. A guy gave me few of these capacitors, and told me it's 1nF capacitor, but I'm not really sure it is, and I cannot check it since I do not have the capacitance meter yet, as you can see. I was unable to find anything about "1000H" marking either. I'll try the circuit later with a normal, 10% capacitor I guess. Oh well.

And I believe that with the circuit that uses calibration capacitor and low resistance reed relay the error is brought down to minimum, and mostly depends of the capacitor's tolerance. Here are the equations, it's a brilliant design. Some circuits online skip the whole calibration step, I believe this one is much more accurate than those.

 

Hey, I did some more research and found out that "1000H" means 1000pf with 2.5% tolerance, which is great.

And L meter seems to work just fine, while the C meter goes a bit out of specs. Here are a few pics:
http://i.imgur.com/SZlgmts.jpg - 100nF 10%
http://i.imgur.com/JQPkIOz.jpg - 100uH 10%

And without any capacitor on:
http://i.imgur.com/813866j.jpg

 

Congratulations on your success! I am envious of our great looking case.