LC resonator low amplitude?

[brashear]

Hi All,
I'm trying to drive a piezo ultrasonic transducer with a capacitance of 1.6nF with an LC circuit. Here is the circuit diagram:

http://min.us/icWNq4.jpg

The square function is at 60kHz. The Inductor I made myself by winding 20AWG enamel-coated wire around a 0.44in diameter x 0.75inch long plastic rod about 77 turns. It's done pretty sloppily - but I don't know if this will make a huge difference.

I measured the voltage on the collector of the PNP and am only getting a peak to peak of 10V. I played around with the frequency of the square function, and it seemed like 60 kHz gave the max amplitude from the LC circuit.

Is there a way I can increase the amplitude? Thanks and happy thanksgiving.

 

[brashear]

I also put a 100uF electrolytic capacitor between +5 and ground. I forgot to include this in the diagram.

 

[VEBill]

I'm not sure where to begin...

2N2222A is an NPN (Not Pointing iN), not a PNP (Pointing iN Please), transistor.

Transistor circuits tend to include the occasional resistor to control the currents in a voltage world. ;-)

I'm sure what your assumptions were, but voltage rise in a resonant (?) LC tank circuit requires Q and thus needs to lightly loaded. A parallel resonant circuit at resonance is essentially an open circuit.

...

It would better to start with the specs of your transducer and just start again.

 

[VEBill]

Correction: "I'm not sure..."

 

[brashear]

I think it's quite clear that I'm not an electrical engineer

Yes, I wrote PNP incorrectly. If you look at the circuit diagram, I used the correct transistor. I forgot to include a resistor between the signal and base in the diagram.

Anyway, here is the circuit that I'm trying to mimic:

http://www.micro-examples.com/pics/090-ULTRASONIC-RANGER-schematic.png

Which comes back to my question, how can I increase the amplitude?

 

[VEBill]

If you're using the same (or a similar) transducer as the Gavand schematic, and you trust him, then simply copy his circuit exactly. AND you would need to follow his guidance for the waveform emitted from the uC.

Otherwise, if you wish to make a different circuit, then you must start with the requirements ("...the specs of your transducer...") and work backwards from them. One cannot even begin to start a fresh circuit design without having the requirements (transducer) defined.

For Gavand's circuit, it seems (?) like he might be using a flyback effect where he 'loads up' the inductor, and then cuts off the transistor current allowing an inductive kickback to decay through the transducer. Maybe...

 

[VEBill]

PS: His inductor probably isn't an air core. The value (330uH) and application hint that his inductor has a ferrite (or maybe iron) core.

 

[brashear]

Yes I understand the flyback effect and that's what I'm using. It seems like the current is not shutting off fast enough, when the uC goes low, thus not generating a high enough voltage from the inductor. Maybe I can switch the transistor?

I do know the specs of my transducer. It has a capacitance of 1.6nF, which I wrote on the diagram. I added a large capacitor in parallel to decrease the size of the inductor to reach resonance.

I don't think an air core vs. ferrite core would make any difference on the voltage produced by the inductor. As I understand, air core should give higher voltages due to lack of hysteresis losses in the core.

I noted that the transistor got a bit warm if the incoming signal was continuous. I will try using a transistor with a higher slew rate and see what happens.

 

[VEBill]

"...I understand the flyback effect... ...I added a large capacitor....to reach resonance. ..."

Flyback effect .NE. resonance. Flyback effect is like an ignition coil; nothing to do with LC resonance.

 

[ijl]

0.4uF sounds a bit too large as the capacitor of a LC circuit at 60kHz. Try about 10...50nF, and a correspondingly larger inductance. The Q-value of the LC-circuit tends to go up, when the inductance is increased. You may also try to connect the transistor to a tap in (approximately) the middle of the winding. If the transistor is getting warm, try a shorter pulse.

 

[djs]

Most ultrasonic transducers are driven by a pulse, very fast rise and fall time, extremely short duration. The idea is to produce a sound pulse (one pulse) and then measure the time it takes for that pulse to return.

 

[ScottyUK]

Depends on what the transducer is doing. The transducers used in ultrasonic cleaners are quite different in purpose to what you are describing.

Of course the OP could always tell us what the application is and then we wouldn't have to guess...


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[VEBill]

An ultrasonic cleaner might need a bit more 'oomph' than could be provided by a 2N2222A. ;-)

 

[btrueblood]

The schematic shows he's trying to make an ultrasonic ranging device, i.e. "ping" a small speaker, then listen for the return echo using a seperate microphone. Agree with djs, the ping should be a single pulse, to minimize ringing and crosstalk to the pickup device. But whatdoiknow, just a mech.

 

[VEBill]

Remove the parallel capacitor.

Add inductance. Your homebrew coil is *about* 50uH (?), the schematic you're copying calls for 330uH. Consider using a core other than air; it's not common to see "kHz" and "air core" together (an overly generalized statement, but mostly true).

Better yet, just copy his (Gavand's) circuit exactly before fiddling with the design.

Good luck.

 

[brashear]

One ping is usually not enough to produce an echo that is loud enough for the transducer to hear - I guess because of resistance to vibrations in the transmission medium and the object reflecting the sound waves. I am sending 10 pulses. It's a trade off between increasing minimum detection range to a longer range.

I think Ijl nailed it with the lower Q factor of the lower inductance coil. I suspected it's because there wasn't enough inductance, even though I reached the 60kHz resonance frequency by increasing the capacitance. The problem is that I didn't have any inductors or a suitable core material, and I wasn't about to wind a 10^4 turn inductor by hand to reach mH inductance.

As VE1BLL pointed out, non-air core inductors are usually used for kHz frequencies because the core increases the inductance for the same number of turns of wire when compared to air-cores, allowing less length of wire to be used. Thus the L/R ratio is higher (Q=wL/R)

By using a smaller capacitor to reach the same resonance frequency, the Q factor of the capacitor is also larger. (Q=1/wCR)

Thank you for your help everyone, this is my first time experimenting with electronics. I must say that this is less intuitive to me than mechanical systems where I can feel and see what's going on.

 

[DHambley]

You can't get more than 10V peak-peak with that circuit. There is no gain in the transistor when the collector swings down past 0V.