Newbie inductor question

OK. That explains some of it. But the real reason for my confusion is that when I go to the shop, the largest value available is 3.3mH--and it's as big across as my hand!

Is there another kind of inductor which can achieve 47mH without being huge? Or is this one simply intended to be a desktop bat detector? I am quite certain I'm missing something really basic and vital here. If a 47mH component is destined to be large, I'll just pick another design. I just can't stand not knowing. :)

OK, I'll do that.

Thanks,

Torben

when

across

huge? Or

certain I'm

destined

knowing. :)

The 47 mH could be correct. It sure would not be uH at that frequency. Look for some toroid inductors. To handle low current and low voltages you should be able to find some that are less than 2 inches in diameter. They look like small dounauts wound with wire.

Hate to follow up to myself, but never mind. Found some for sale on the net.

This

Anyway, thanks again for the help. If you know of anyone in Vancouver who carries them...:)

Torben

The inductor values are correctly marked in milli-Henrys (mH). For example, the filter cut-off frequency is calculated like so: 1/2/pi/sqrt(47e-3*47e-9) = 3.4 KHz

A 10uH choke across the input is like a short-circuit at ultrasonic frequencies - better to have no choke at all than one that is too small.

The best way to test it would be to get a second ultrasonic transducer and feed it from a small 40 KHz oscillator. I wouldn't expect to get any results from TV sets, dropping coins e.t.c. Alternatively, you could just inject a very weak 40 KHz signal directly at the input via a very small coupling capacitor.

Yes, that's the sort of thing. They're usually wound on small ferrite bobbins. I've got some in my junk box with values of 220mH and more in very small sizes.

It all depends on the current rating. Since this application is a very low current one, they can be quite small and low cost, because they can contain a large number of turns of very fine wire.

For example:

The inductor is not marked wrong. It is the proper value for a low pass filter of 3.39KHz which is correct. The purpose of the LPF is to eliminate the higher frequencies of the oscillator and bats from the audio stream. The thing works by hetrodyning or mixing a high frequency oscillator with the ultrasonic bat squeeks and detecting the resultant difference frequency when it is in the audio range. It is in essence a down converter with filtering and amplifying circuitry for the detected audio frequencies.

If you hear beat frequencies like a squealing radio when there are no bats, there are at least two high frequencies in the system interacting with each other, so something is wrong or oscillating that shouldn't be. Possibly a nearby radio station is interfering. Trouble shooting is going to be difficult without knowledge of electronics and a scope to observe wave forms.

As far as the inductor is concerned, it is the right value but is a poor choice for a low pass filter being large and expensive. A better filter would be made from an op-amp and a couple of resistors and capacitors, no inductor. It's called a low pass active filter, second order. Look up Sallen and Key. The 10mH inductor can be replaced as well with proper circuitry. I think the circuit needs revision and upgrading. Bob

Thanks for the detailed explanation.

I've now built a Sallen Key Butterworth LPF (my first) for the circuit, and it seems to have settled the whole thing down--now it just sounds like an audio circuit. Next project is a 40kHz oscillator to test it...and then something to detune the piezo, if needed.

In the circuit designer's defense, I believe that one of the main criteria for the design was low part count, and that he made compromises. After putting the LPF in and a rail splitter to power it, my bat detector has grown by about 50%. I have found 47mH coils online which are quite small (look about a quarter-inch across) so I'll see if I can't order a few.

Cheers,

Torben

use thinner wire, they're only carrying small signals not power like the inductors used in a loudspeaker crossover.

Bye. Jasen

Did you read to the bottom of the page? Under "Improvements", he says, "The value of the coil parallel to the transducer is not optimal. A value of 8.2 mH or 5.6 mH is probably better suited to broaden the frequency range of the transducer. See also the page about detuning."

And for inductors, you can get them anywhere, e.g,:

Cheers! Rich

Good Grief, what kind of shop? That sounds like a very high-current choke - is it used in a welder or something?

Good Luck! Rich

I did, and I've been going over the detuning page as well. Sadly, the largest available value was 3.3mH, and as I said, I could wear it as a bracelet (if I could find a matching hat). I was just stating the value from the schematic since I could deal with a suboptimal but close value; an order of magnitude off is too much though.

Right now, I've got a grab bag of chokes from The Source (was Radio Shack) but they're not marked or labeled, so NOW I'm working on the math to work out their values. There seem to be a lot of articles on the topic.

But for the 47mH coil in the low-pass filter: I've just pulled it out for now and replaced it with a sallen-key filter.

I have found suitable values on the net for sale, but the guys in the component shops here just kinda look at me funny and say that nobody in town is likely to stock anything in the mH range except, apparently, giant ones.

I must be asking them the wrong question, because it doesn't seem like it should be this hard. Maybe Canadians just build more welders than bat detectors. Dunno. But there are some local crazy repair guys with all kinds of stuff, local ham clubs, and so on.

Thanks for the replies! I know this is painful to watch for the old hats on the group, but at least this problem has forced me to learn more about inductance, filters, and heterodyning than I would have had it been easy.

Torben

Thanks for the explanation. Much easier to understand than a lot of the library texts. :)

Torben

Thanks again for the ideas. It hadn't occurred to me to just choose another LPF design, and it worked great. I got rid of the rail splitter by using an LM358 for the LPF--may not be the best opamp around, from what I gather, but it works well enough for my purposes, and I have a pile of them from a grab bag. A bit hissy, but hey.

The whole thing now works wonderfully. I have a bit of squeal when I tune the TCA440 to the low end of its range, but I think I have the equation to explain that so I'll see about fixing it. (I'm planning to raise the low end of the tuning range with the C*R off of the TCA440 pins 5 and 6. Wrong? Looks right to a newbie.)

The piezo I have seems to have a wide enough bandwidth without detuning. You said that detuning could be achieved by proper circuitry (not involving inductors), but from what I have found so far, that would be at the expense of response. If you have some hints in that direction I'd love it (I'm happy with the current response/bandwidth, but I'd like to know how to make it even better).

Troubleshooting is hard without a scope and knowledge, I agree. I have an old tube scope (which actually works); the knowledge, I'm working on. :) I'd love to go back to school for this (computer programming didn't satisfy) but for now, reading textbooks and asking for explanations will have to do.

Torben

When it comes to inductors, the size has everything to do with the thickness of the wire, which determines how much current it will handle and how much DC resistance it will have and possibly more important, what the core of the inductor is made of. Air core inductors or those wound on plastic or wood have much less of a value than those that are wound on ferrite bobbins or toroids.

James. :o)

Search for 807-4590-476k. This gets you the 47mH. Click on page 671 link for the whole page of them.

The 47mH are in stock for $1.98. They are .9" long by .455" diameter.

Get your 10mH, 8.2mH, and 5.6mH at the same time.

Good luck, John