Change the frequency of voltage of C1

[Chuck88]

Dear All,

I am designing the circuit for an atomizer. The problem I am facing now is I do not know the method to change the frequency of the voltage applied on the capacitor of my circuit.

Can you provide me some information about the way to change the frequency of the voltage applied on the capacitance C1? C1 is indicated in on the figure attached.

I do not quite know the elements in the circuit which would determine the frequency of the voltage applied on the capacitance of C1.

 

[NascentOxygen]

I don't know either. Can you construct the circuit in such a way that you can experiment before selecting the final chosen value? Why do you want/need to vary the frequency? Do you need to just fine tune it, or might you want to double or triple the frequency?

For tweaking the frequency, I'd try making small changes in any of C5, R1, L3. large changes may cause oscillation to cease or be unreliable.

I'm not sure about these, I'm guessing.

BTW, what are you atomizing?

 

[Chuck88]

I don't know either. Can you construct the circuit in such a way that you can experiment before selecting the final chosen value? Why do you want/need to vary the frequency? Do you need to just fine tune it, or might you want to double or triple the frequency?

For tweaking the frequency, I'd try making small changes in any of C5, R1, L3. large changes may cause oscillation to cease or be unreliable.

I'm not sure about these, I'm guessing.

BTW, what are you atomizing?

Thanks for your reply. The capacitor C2 is actually a piezoelectric disk and I am going to use the ultrasonic wave generated by the piezoelectric device to atomize the water. The reaon I want to vary the frequency is that for different piezoelectric disk, the resonant frequency is different.

This circuit is not designed by myself and I just want to copy this circuit board for my own use. So I do not quite understand the working principles of this circuit design. What I want to do is to change the frequency of C1 to change the resonant frequency of the piezoelectric device.

 

[NascentOxygen]

If you can find the source of the circuit, then perhaps we can find there some clues to the way the circuit functions!

Are you an electronics student? If so, perhaps you could model the circuit in SPICE and investigate it there? The frequency may be dependent on supply voltage, so that might be another way to vary the frequency a little. I can see the need to be able to tweak the frequency to be resonant with the crystal.

 

[fbs7]

This is a modified Colpitts oscillator. I didn't particularly like this circuit, because experimentation during simulation led to some high voltages (1KV) and instability in oscillation. If you intend to build the circuit, be very - very careful. Those LC combinations produce some nasty shocks, specially if you're connecting directly to AC.

The circuit operation is highly dependent on the piezo. During simulation it only maintained oscillation on an inductive piezo - a capacitive one damped the oscillations.

Any number of factors can change the oscillation frequency; the one that seemed to be more impacting is changing slightly your C4. If C4 was smaller you could add a trimmer in parallel, but at 470nF you may not find a big enough trimmer; so you may need a bit of trial and error.

By the way, this circuit is suspiciously similar to

http://www.google.com/patents?id=XiEbAAAAEBAJ&printsec=abstract&zoom=4#v=onepage&q&f=false

Whoever you got the circuit from seems to have taken the patent and modified it slightly (replaced a resistor with what seems to be a current source. not sure exactly to what benefit - perhaps to control power? /shrug). I don't know enough about patents to advise if that's a violation or not, but whoever your source is he doesn't seem to be completely fair (unless he's the original inventor).

 

[NascentOxygen]

I notice that the power supply voltage is not specified on the circuit, nor the transistor types. What voltage would be needed across the terminals of the piezo element at resonance? Are we talking hundreds of mW of power to the piezo, or tens of watts?

 

[vk6kro]

That high voltage developed in the simulator may be because the simulator assumes perfect components unless you tell it otherwise. This is not the case in real life where inductors have resistance. You can guess at a reasonable value for the resistance and add it into the circuit on the simulator.

The odd looking component at the right is probably a bridge rectifier, but it is connected wrongly. The AC should enter it at the left.

This is a thinly disguised version of a Colpitts crystal oscillator and can hardly claim to be original.
I thought circuits couldn't be patented because they will usually work with minor changes to the component values and yet the circuit is visibly different.

The ultrasonic device will oscillate at its resonant frequency, just like a quartz crystal does, so you won't need to drive it at different frequencies.

It is hard to estimate the power of this device, but it would take quite a bit of power to "atomise" water. This is even allowing for the use of the word "atomise" to mean generate a spray of fine (but not atomic) particles.

 

[psparky]

Dear All,

I am designing the circuit for an atomizer. The problem I am facing now is I do not know the method to change the frequency of the voltage applied on the capacitor of my circuit.

Can you provide me some information about the way to change the frequency of the voltage applied on the capacitance C1? C1 is indicated in on the figure attached.

I do not quite know the elements in the circuit which would determine the frequency of the voltage applied on the capacitance of C1.

Kind of an odd question in my mind...because the reactance of a capacitor is controlled BY THE FREQUENCY...not the other way around.

A capacitor's reactance is 1/(JWC)...where W is radians per second or 2*pi*f...f being frequency. So as frequency changes...so does reactance.

However, if you are talking about the resonant frequency...that is set by this equation:

W=1/(√LC)

So the total reactance of all the inductors...multiplied by the total reactance of all the capacitors as seen by the source...take the square root...inverse...and then change to frequency...will determine the resonant frequency.

So in my mind...the FREQUENCY of your source should be your interest. So if you adjust the frequency input...you adjust the reactance of your capacitors and inductors. If you want a resonant frequency at a particular frequency...you will have to set your final capacitance or inductance.

Perhaps I'm not getting the question or the circuit...but I'm pretty sure what I stated is true.

That looks like Pspice btw...simply alter the freqency of your AC voltage source...device on the right side of the circuit with the ~ sympol in it. Altering the frequency of your source will alter the output of your circuit.

 

[psparky]

Also...like someone else said...that rectifier is odd right to the left of the AC source.

If you are going to put DC to the ciruit...the inductors will short and the capacitors will open to some respect. The DC component will have some type of frequency...but it's still DC. (THIS STATEMENT IS FALSE...I FIGURE IT OUT IN THE NEXT POST!)

Hmmmmm...

Also, you said you are designing a circuit...yet want to find the way to alter the freqency of a capacitor? You would have to be pretty advanced to design a circuit that complicated...Hmmmmm...once again...something doesn't add up. It sounds more like you copied a circuit...altered a thing or two and called it your own. That's not designing...that's altering. Just sayinig.

One more thing...the frequency in a circuit is the same throughout the entire circuit. If your voltage source is 60 hz...your entire circuit is at 60 hz. You can not alter the frequency of just one component.

 

[psparky]

What I just said about rectified DC raises another question in my head.

In a true battery...caps and inductors will open and short respectively.

But in rectified AC...there is a frequency...so I'm assuming Caps will follow the "W" in their reactance...is this true or false?

Thinking even more...in rectified AC...the current never changes direction...does this play a role in the reactance as well?

Someone help me out here...

OH wait...since the current thru a cap is C*(dv/dt)...this tells me that caps don't open in rectified DC...because there clearly is a change of voltage! Nothing like having a conversation with yourself and arguing points! lol

Which leads me to yet another conclusion...if you find the current thru a cap...using 1/JWC (using V=IR)...or C*(dv/dt)...you should come up with the exact same current...if not...you did something wrong!

 

[vk6kro]

The output of the bridge rectifier should go to a large electrolytic capacitor which will turn pulsing DC into pure DC. I don't see such a capacitor, but assume it is there. The oscillator should only get DC voltage from the power source, not 60 Hz or pulsing 120 Hz.

This is an oscillator circuit similar to this one (except it uses a FET) :

The component marked "10M" is a 10 MHz crystal and the circuit will oscillate at 10 MHz and produce a sinewave at the drain of the FET.
This is because the crystal produces a high impedance at its parallel resonance and, at this frequency, feedback from the source to the gate is possible.

The author of the original article apparently found that an ultrasonic resonator will work in the circuit in the patent in post #5. It looks pretty weird, though.

The other components in the circuit have only minor effects on the frequency of oscillation and using a different crystal (or ultrasonic resonator) will result in oscillation at the resonant frequency of that device.

.

 

[NascentOxygen]

The ultrasonic device will oscillate at its resonant frequency, just like a quartz crystal does, so you won't need to drive it at different frequencies.

Good point! https://www.physicsforums.com/images/icons/icon14.gif It shouldn't be labelled as a capacitor. The piezo device is the tuned circuit and (all being well) the circuit will naturally oscillate at the resonant frequency of the device, no fine tuning or tweaking necessary.

I get the feeling this is not the circuit for Chuck88. There are too many "unspecifieds" involved. A google search is sure to turn up better-explained circuits for the hobbyist, practically guaranteed to work. Surely the manufacturers would have application notes, too? Have you purchased the peizo device, Chick88? What are its specifics?

 

[NascentOxygen]

Here's an atomizer that provides all component values. http://www.8085projects.info/post/Multi-purpose-ultrasonic-atomizer-circuit.aspx

 

[vk6kro]

Good. It looks like the power supply would be about 50 volts and the circuit uses a power transistor.

The circuit can be redrawn like this:

http://dl.dropbox.com/u/4222062/Colpitts%20Osc.PNG

because the supply rails are bypassed so that either supply rail is the same as the other for the oscillation frequency.

This is a very conventional Colpitts oscillator.

 

[fbs7]

That high voltage developed in the simulator may be because the simulator assumes perfect components unless you tell it otherwise. This is not the case in real life where inductors have resistance. You can guess at a reasonable value for the resistance and add it into the circuit on the simulator.

The ultrasonic device will oscillate at its resonant frequency, just like a quartz crystal does, so you won't need to drive it at different frequencies.

Just two comments on this: the original circuit is for a 25-30W, 1-MHz transducer driven by a tuned circuit connected directly to AC. That can give nasty shocks. It seems dangerous to me, that's why of my warning.

About the operating at resonance frequency, you are right, provided the circuit provides the correct feedback for the oscillator to operate at resonant frequency. Now, a piezo element has two main frequencies: fr which is the resonant frequency and fa which is the anti-resonant frequency. One can drive the piezo at fr or fa, depending on the configuration, or at a multiple too.

But patent 3989042 explains that from their tests the nebulizer thingie produces more... err.. nebulizations neither at fr or fa, but on a frequency in between them (I have no idea why, I'd imagine that maximum mechanical vibration should produce more droplets - I can only guess that is some magic frequency defined by water, not by the transducer - and they don't disclose the magic frequency). Therefore the circuit on the original post is designed to operate off the transducer's resonant frequency.

 

[vk6kro]

The voltages produced in the circuit are not a problem if proper construction techniques are used. It can be enclosed in a sealed box and the voltages can be insulated.

However, with a 50 volt supply, the voltages won't be anything spectacular. Certainly nothing dangerous, despite the simulation forecasts.

High powered ultrasonics can cause heating in human bodies, particularly where flesh meets bone, so it is worth being aware of this.
However if it is in a metal enclosure, there will be no problem.

Crystals produce two resonances too. They are typically 1000 Hz apart and the series resonant one is at the lower frequency. Oscillation usually occurs at one or the other, although loading with capacitance can pull the oscillation frequency down while series inductance can raise the oscillation frequency.

From what you say, this applies also to ultrasonic resonators and the frequency may be pulled down by parallel capacitance.

 

[fbs7]

The voltages produced in the circuit are not a problem if proper construction techniques are used. It can be enclosed in a sealed box and the voltages can be insulated.

Vk6kro, I don't understand your point. You sound as if this is safe. The guy is a novice, doesn't understand the circuit and wants to tweak it around in order to change the frequency applied to the transducer.

I warned that a megahertz-range, 30W tuned circuit connected directly to AC is prone to give dangerous high-voltage spikes if you mess with it.

If this was a UL-tested, professionally built sealed appliance I would say nothing. If that was someone with experience I would say nothing. If you were on his side overseeing his safety I would say nothing either.

But as far as I know from the post this may be a lone kid playing in the garage. You think that's safe and doesn't deserve a warning?

 

[vk6kro]

You are quite entitled to give whatever warning you like, but in this case, it is not justified, although I never comment on other postings unless asked.

RF energy should not be touched, but, if it is, it runs across your skin instead of through your body. So, you can end up with a minor burn on your skin rather than a potentially fatal shock.
I've had them while adjusting antennas. It stings slightly, but that is all.

I like the term "connected to AC". How much AC?
Nobody said it is connected directly to the mains, did they?
AC also comes from a low voltage transformer winding.

You ran a simulation and got the impression that huge voltages are produced. You said 1000 volts.
Well, firstly that is not true in real circuits and, secondly, any voltages that are produced vanish pretty quickly when a human touches them.

But where do we stop with the needless warnings?
Don't touch hot resistors. They might burn you?
These are intelligent people and they have chosen a mildly dangerous activity, so they have a responsibility to look after their own welfare.

We say to put live circuits in a box and don't touch them. That should be enough.

I can think of some activities that I would draw the line at, but this is not one of them.

The actual hazard with this device is the ultrasonic energy. You can't hear it or see it, but enough of it can cause pain.

 

[fbs7]

You are quite entitled to give whatever warning you like, but in this case, it is not justified, although I never comment on other postings unless asked.
...
I like the term "connected to AC". How much AC?
Nobody said it is connected directly to the mains, did they?
AC also comes from a low voltage transformer winding.
...

Except you commented on every one of my posts, dismissing my warning.

You will notice that original circuit (from which the one in the first post is copied) is connected to 110V through a 1:1 transformer that feeds several such boards in parallel. The purpose of the circuit is to allow several such boards to work, each one with very little regulation, and it is designed to avoid one board from interfering with the oscillation of the other.

But, nevermind that, you clarified your advice: it's not justified to warn of danger of transients in AC-level high-frequency circuits, even if we have no idea who the original poster is or what is his experience.

That's fine advice.

But, I'm not going to change my warning. This is a public board and people with all kind of experience read it. There is no such thing as too much warning - specially if we don't know the other person's experience. Safety first, and I would appreciate if you didn't try to dismiss that.

That's the end of this discussion for me.

ps: just so that it's clear where I'm coming from, please read for me the voltage in the inductor in the simulation below. I don't claim the simulation is perfect (I have no idea the characteristics of the transducer he'll use), I only claim that the circuit originally posted can produce high voltages, specially if the poster will try to mess with it.

 

[vk6kro]

The patent data you published shows an input AC voltage of 48 volts on page 5, and an output waveform of 75 volts peak (150 volts peak to peak) on page 7.

The choice of input voltage is obviously optional. Another post gave an input voltage of 37 volts, as I recall.

My only interest in this is that you gave a warning based on a simulation where you apparently used an inductor of infinite Q (the 24 μH ) and as a result, got an output of thousands of volts.

Of course you got this result, but it doesn't warrant a general warning about the danger of building this device.

Your circuit doesn't show a transducer. How did you arrive at the substitute components?

I just did a simulation of a pi-network ( 1000 pF to ground, 2 uH in series then 300 pF to ground) and at 6.5 MHz I got a peak of 36000 volts when I fed it with 2 volts across the 1000 pF capacitor.
Adding just 0,1 ohms in series with the inductor brought the output down to 1600 volts and 1 ohm dropped it to 160 volts. So, you have to be skeptical of simulator results like this.