Atomiser circuit - Changing resonent frequency

[eetaylog]

"Nothing" isn't a reading an oscilloscope makes. What does it show?

A flat line. I get the same reading when probing across C2 which suggests Q1 isnt being switched on and off by the signal generator?

 

[duffy]

A flat line.

Yes, I figured that much. There is something else it is telling you - care to share it with the guy who's trying to help you?

 

[eetaylog]

Yes, I figured that much. There is something else it is telling you - care to share it with the guy who's trying to help you?

Im not sure what youre getting at? Its reading a flat 0V as if there is no signal being read by the oscilloscope?

 

[duffy]

Yes - the VOLTAGE. And you say 0V. Put your finger on L1. Is it getting hot?

 

[eetaylog]

Is it getting hot?

No, not really.

 

[duffy]

Actually, there's no way you should get 0V unless there's a break in the +24Vdc line - the lowest voltage you should see even with Q2 on continuously would be around 2V. Check your power supply. Make sure your scope is set for DC input.

 

[eetaylog]

I know, its very strange. I know for a fact that the circuit is working just fine, but it just seems that as soon as i plug the holed transducer into it i lose all the voltage at the rail? Ive attached some images of my rig and the results im getting from both types of atomiser.

- Pic1 shows the oscilloscope reading from the collector of Q2 with the holed piezo (ignore the reading of 4.8V, its just a slight ripple that is being picked up)
- Pic2 shows the oscilloscope reading from the collector of Q2 with the non-holed piezo
- Pic3 shows my circuit (with non-holed piezo hooked up)
- Pic4 shows the chamber with the holed piezo in it (you can just see the hole through the water)

Im stumped, but can only imagine that the holed piezo is goosed?

 

[duffy]

Press the CH2 menu button, select "DC input", and try reading it again.

It will be higher than 0V. If you still see 0V, get a voltmeter (using the DC setting) and double-check it. We need to see where the collector of Q2 is sitting to figure this out.

BTW - you aren't trying to operate those 2 transducers in parallel, are you?

 

[eetaylog]

Press the CH2 menu button, select "DC input", and try reading it again.

It will be higher than 0V. If you still see 0V, get a voltmeter (using the DC setting) and double-check it. We need to see where the collector of Q2 is sitting to figure this out.

BTW - you aren't trying to operate those 2 transducers in parallel, are you?

Ok, im off to lunch now but will try it this afternoon and report back. No, im only soldering one piezo into the circuit at a time.

Thanks.

 

[alec_t]

Does your power supply have an overload trip? If Q2 stops oscillating (perhaps because of damping by the epoxy/glass) while Q1 is 'on' then the standing Q2 current will be excessive and drop the supply volts.
Have you tried the holed piezo in circuit without the sig gen and with J3/J4 shorted?

 

[duffy]

Even if the circuit won't self-resonate, it can easily be converted into a driver that you can use with your frequency generator to force some sort of vibration out of that holed transducer. You are almost there, you just need to trace the signal through the circuit.

 

[eetaylog]

Does your power supply have an overload trip? If Q2 stops oscillating (perhaps because of damping by the epoxy/glass) while Q1 is 'on' then the standing Q2 current will be excessive and drop the supply volts.

Ok, ive now tried reading the oscilloscope using DC coupling and i get 24V DC at the collector of Q2 with the holed piezo.

I think youre right about the transducer stopping oscillating due to the glass, and i think that as a result Q2 is on constantly because my power supply is reading 400mA being drawn by the circuit. Q2 is certainly getting hot when using the holed piezo so i can only imagine that this 400mA is probably going via the collector->emitter->L3 path where the reactance of L3 is the only thing preventing a short to ground.

I think that possibly the conclusion is that the piezo wont naturally resonate in this circuit whilst it has a hole in it, and that the only reason it responded during the frequency sweep is because it was being forced to, rather than (to use your analogy in a previous post) it being struck like a bell and left to resonate on its own?

Have you tried the holed piezo in circuit without the sig gen and with J3/J4 shorted?

Yes i have, the readings that ive just described are with the holed piezo in place with J3 and J4 shorted. With a normal piezo in place the collector of Q2 bounces up and down nicely as in pic2 in my previous post.

 

[alec_t]

If you've only got 45mm water pressure the glass plug is a bit OTT. If you could remove it and instead use a thin clear plastic or glass cover (microscope cover glass?) over the hole, on one face of the piezo, there would be much less mechanical damping. Assuming the piezo is being driven in thickness mode (rather than in a flexure mode), the plug is presently acting like a clamp to prevent vibration of the piezo. With the plug gone it might then be able to oscillate freely.

 

[eetaylog]

If you've only got 45mm water pressure the glass plug is a bit OTT. If you could remove it and instead use a thin clear plastic or glass cover (microscope cover glass?) over the hole, on one face of the piezo, there would be much less mechanical damping. Assuming the piezo is being driven in thickness mode (rather than in a flexure mode), the plug is presently acting like a clamp to prevent vibration of the piezo. With the plug gone it might then be able to oscillate freely.

Yeah thats what i think as well.

 

[duffy]

Yes, +24Vdc is more sensible. Q2 isn't being biased correctly. Hook the signal generator back up. Leave the scope in DC mode, crank the gain up, zero that offset, and follow the signal from Q1 down that chain of components D5 - L2 - R4 - R5 and tell me where where you lose it. Reduce VR1 and see if that improves the bias. If you can get .7V p-p at the base of Q2, you can use the signal generator to drive Q2 through Q1 and force the resonator to vibrate, and then let's see what's what. The kind of waveform it gives will convey information about the state it's in. A jagged mess would be harmonics from cracks you got while drilling it out. A regular but distorted waveform could be that loading you are talking about, which you might be able to adjust and watch in realtime. I'm hoping for a clean but very small p-p voltage sine wave, this one says you are in the ballpark and just using the wrong frequency.

 

[eetaylog]

Right, ive hooked the sig. gen back up and have it set on 100KHz square wave with amplitude turned up to max (20V p-p). The oscilloscope reading at the base of Q1 is as expected (100Khz square wave more or less). Following the emitter path from the output of Q1 i get:

Q1 emitter (top of C2) - 2Vdc
Anode of D5 - 1.3Vdc (as expected, due to 0.7v drop across diode)
L2/R4 junction - 1.3Vdc
R4/C3/C4/R5 junction - ~0.7Vdc
Q2 collector - 24Vdc

If i reduce the frequency down to 100Hz i can make the emitter of Q1 bounce up and down, so it looks like once the sig gen reaches ~100kHz C2 is keeping Q1 emitter constant DC due to the time constant being too short for C2 to discharge.

Am i right in thinking that if i change C2 for a resistor i can drive the resonator with the sig. gen as you suggested?

 

[duffy]

^ Yes! Try a 220Ω resistor in place of C2 and see if that gets it going at the high frequencies.

 

[eetaylog]

^ Yes! Try a 220Ω resistor in place of C2 and see if that gets it going at the high frequencies.

Right, just an update on this, ive got the circuit working with a holed piezo now. As i thought, the circuit was working the whole time because i was getting a mist form when i used an unholed one, but it turns out that the holed one was fubar'd all along which was why i couldnt get it ringing on its own.

Our customer has now said that the hole doesnt need to be as big as 5mm so i got our mechanical guys to drill a 2mm hole instead, which works just fine in my circuit.

Thanks for all your help on this duffy, it feels a bit like we were chasing our tails with it but we got there in the end.

 

[duffy]

Glad to hear it, and thanks for the update.