Varying the output frequency of an ultrasonic driver board

[Braeden Hamson]

Hey guys,

So I've got an ultrasonic system with a driver board and transducer. I would like to be able to vary the output frequency, as I've developed a horn that resonates at 26 kHz and not the 28 kHz the board outputs.

The circuit looks something like this:
https://cdn.hackaday.io/images/7856741430093121127.png

Not my drawing. But I think its fairly close, though I think that Q1 or Q2 is upside down. From what I can tell that is a Royer oscillator and one of those transistors should be opposite the other. Now I think that L1-1 is responsible for triggering Q1 & Q2. I think if I replace the signal from L1-1 with an Arduino I might be able to vary the output frequency. However this is what I refer to as "murder voltage" as its at times upwards of 300 V. I have the board connected to a switch that I can use to quickly turn it on and off. I don't take HV lightly but I think I'll be okay. Only one hand at a time!

My final question, I would like to attach my oscilloscope probe to L1-1, but I'm not sure where to put my ground connector. Any ideas?

 

[Harald Kapp]

I think that Q1 or Q2 is upside down. From what I can tell that is a Royer oscillator and one of those transistors should be opposite the other.

No. The Royer oscillator as shown e.g. here uses a center tapped transformer and switches the two ends of the transformer to steer the primary side current.
The circuit you show uses a center tap in the power supply (R1 | R3) connected to one end of the transformer and switches the other end to either V+ or V-. NPN transistors can be used in both cases as the base drive is floating via L1-L2 or L1-L3 respectively.

I think if I replace the signal from L1-1 with an Arduino I might be able to vary the output frequency.

That should be possible, but then this circuit doesn't act as an oscillator, only as a driver, which is o.k. Disconnect L1-1 from R5, drive this side of the coil via the arduino (a booster stage will be required to drive sufficient current through L1-1). That way the arduino side is insulated from teh high voltage side (insulation only as good as the transformer offers).

I have the board connected to a switch that I can use to quickly turn it on and off.

That is insufficient. If you really get shocked (which hopefully will never happen) you may not be able to throw the switch into the off position. At least a RCD on the high voltage primary side is required.

Only one hand at a time!

That too is insufficient. One hand only reduces the chances that you accidentally connect two points on the HV-side, but a single hand suffices to get shocked from a single point. Insulated gloves are a better option but awkward to use in a delicate electronic circuit.

I would like to attach my oscilloscope probe to L1-1, but I'm not sure where to put my ground connector. Any ideas?

A typical oscilloscope is grounded and therefore should not be connected to the HV-side of this circuit. With a bit of luck only a fuse blows, with less luck the oscilloscope is good for the junk yard. Use an isolated scope (e.g. battery powered or with an isolation transformer on the prmary side).
If you use the Arduino setup as explained above, L1-1 is on the low voltaeg side (disconnetced from R5) and the scope can safely be used.

 

[Braeden Hamson]

Thank you as always for you help. Yeah thinking on it I don't know why I even mentioned the switch, I can probably actuate it in 500 ms at best, which is an eternity and a half in the electronics world. The system is connected to a house hold wall outlet that is on a breaker circuit. I don't know if there is a GCI on the circuit, but this building is less than 4 years old. But I have no plans to have any part of me near it when its running. Funny that it has no HV warnings anywhere on it. So when I first got it I hooked the output to my oscilloscope. Which suggested that it was over 1000 V. That's when I took it seriously.

A booster stage you say, would this be a transistor and an external power supply? I'm not familiar with that circuit. Google just gives me rocket pictures. Also what would you suggest I do about connecting the ground of the Arduino and the ground of the power supply?

 

[Harald Kapp]

A booster stage you say, would this be a transistor and an external power supply?

Could be as simple as a single transistor, see e.g. here. But not in this case: L1-1 needs to be driven by an AC signal. An H-bridge driver circuit will be required.I named it booster in my first reply as L1-1 will probably require a hugh current (but low voltage), much higher than an Arduino can drive. Google"Arduino H-bridge module".

Also what would you suggest I do about connecting the ground of the Arduino and the ground of the power supply?

Two words: Don't (although I don't know whether that contraction counts as one word or two ). If you use L1-1 as suggested, the primary of L1-1 which is connected to the arduino is on safe low voltage potential and there is no need to connect this side to the HV-side.

 

[WHONOES]

What are you planning to use it for and how much power do you need?

 

[Braeden Hamson]

Now that I'm actually awake this makes much more sense I completely disconnect the winding if L1-1 and put my Arduino set up across this wire. I'm not connecting my Arduino to R5. Clever. And I'll need the H-bridge to drive current in an AC fashion correct? Thus far I've focused on digital electronics so I'm not familiar with driving a coil. Do I need to add series resistance?

EDIT: I Googled my question, and got a plethora of circuits to help me drive a coil. Heh, give a man an answer and he'll understand for a day. Teach a man to Google and he'll waste 5 hours and know 30 surprising facts about Tibetan sea snails. Wait something's gone wrong in my metaphor.

 

[Braeden Hamson]

What are you planning to use it for and how much power do you need?

I'm making a sonicator which is this nifty device that allows you to make emulsions without a surfactant or emulsifier. You can also make extractions from cells without additional chemicals. (everything is a chemical) I would like as much power as I can get a hold of, but don't we all. My board is an Ebay special rated at 100W, so probably 80W. In the future I may make my own board, something in the 500W range.

 

[73's de Edd]

Sir Braeden Hamson . . . . .

O.K. . . . . . Overall . . . . . initially, I can see this IS being a POWER oscillator circuit, with a failsafe fearure, in the respect that if there is no Xducer being connected or that it open circuits, you then have no feedback from L1 being inhibited.

Question . . . is this being your reverse engineered schematic from an actual unit ?

Since, I CAN see a pair of 13007's being capable of your specified power. . . . . but . . . . . not with the treatment of the circuitry, as is being shown in the designated yellow boxed areas.
Specificaly the power path of line pulsating raw DC from the FWB, in its getting to the pin 1 of T1 transformer.

The top drawing is being your original . . . .with markups . . . .and below your original I drew in one possibility of the wiring.
And with that even having the variant, involving the RED A to A' and the BLUE B to B' connections.

Thaaaaaassit . . . . . .

REFERENCING . . . . .





73's de Edd . . . . .

The automated factory of the future will have only two employees, a man and a guard dog.
The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment.

 

[Braeden Hamson]

Sir Braeden Hamson . . . . .

73's de Edd . . . . .

The automated factory of the future will have only two employees, a man and a guard dog.
The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment.

That's very funny BTW

I should've been more clear. That is not my schematic, I got it from hackaday. Someone else reverse engineered it. The board already exists, and is specified at 100W.

 

[Braeden Hamson]

I constructed the H-Bridge, it works as intended. Though I can't measure the "negative current" created by the bridge using my oscilloscope. No matter, I used a motor to verify the functionality. And I've just now realized I used N-channel MOSFETs for the whole bridge... I'll be heading out after this post to buy P-channel FETs. Arduinos aren't cool enough to run this circuit at 26kHz, so I used a 555 timer to make it happen. I use the output to trigger one half of the bridge and then the NOT of the 555 output to trigger the other half. Lastly I put two 5W 22 Ω resistors in parallel for ~10 Ω series resistance to the coil L1-1.

I'm switching a relatively high current 1.2 A @ 12V at 26 kHz. May be chump change for you but for me that's some power. The H-bridge has a square wave output which drives the coil I haven't done any kind of filtering to the output, and I'm getting some ringing. Its hard to measure whats happening as my scope doesn't like this kind of stuff. I've attached a scope capture with the blue trace being the output of the 555 circuit. This particular one is the active high signal. Then the purple trace is with the probe directly across L1-1. With the ground attached there. Not how probes are supposed to be used, I know. So the ringing I see is probably from the probe itself and should be ignored... Hmm... typing things out gives you clarity.

My question is what kind of filtering should I be using/ Obviously a sine output would be idea, I attempted to use 3 resistors in series with 3 capacitors going to ground to create a kind of sine wave. However, I got a big drop in amplitude of the output which isn't ideal.

 

[Harald Kapp]

My question is what kind of filtering should I be using/ Obviously a sine output would be idea, I

A sine cannot be well produced by an H-bridge. Too much power will be lost in the transistors.
You could try to improve the signal by adding a snubber across the primary side of the coil. But some ringing is almost unavoidable (unless you use sine wave drive) due to the switching operation. Make sure the power supply to your circuit is well filtered to avoid emission of the electromagnetic noise from your circuit into the mains grid.