Hi Konstantin! Great video!! I relived the memories of my work from twenty years ago. The intention of the large L2 is to be the dominant inductance in the LC tank circuit formed with the ceramic. Thus, the energy of the LC tank circuit does not return to the primary circuit through transformer T2 (since L2 is uncoupled). This way, you can inject energy into the tank through T2, and this energy keeps bouncing back and forth between the ceramic and L2.
This topology is called a Jensen converter. It's principle of operation is based on the saturation of the current transformer that feeds the base's of the power transistors. When the converter is first turned on, one of the power transistors turns on before the other (due to gain mismatch) and the high value pull-up resistors to the positive rail. This is what initiates the oscillation. Once one switch turns on it causes a current to flow in the current transformer switching the "on" transistor on harder and switching off the other transistor. The current will continue to flow until the volt-second product of the toroid is reached, whereupon the core saturates and ceases to be a transformer, thereby removing the gate drive to the "on" device. The energy stored in the core causes the polarity across the windings to reverse (Lenz's law) switching on the opposite transistor and the cycle repeats. Thus the flux density in the CT core flips between saturation in each direction twice per cycle. The load I expect capacitive and is series resonant with the big inductor. So at resonance the two reactive components will cancel out and the equivalent circuit will simply be a resistance that represents the losses due to the heating of the water in the bath. This equivalent loss resistance needs to be included in your simulation to get a better result. From a design perspective the saturating core cross sectional area is adjusted so that its resonant frequency coincides with that of the LC of the inductor and peizo transducer. Incidentally, this topology was also used in "electronic transformers" used for low voltage halogen globes in past years. It was also used in some HF florescent tube ballasts.
Hi Roger, Thankyou for the Thorough Explanation of the Working Principle. I searched on Google and found that Jensen is also called Royer or Jensen-Royer, bcs I never heard it called Jensen. The Circuit used, looks like a mixture of a Royer and an HalfBridge due to the Capacitive Divider connected to the one Side of the Power Transformer.
@@KonstantinGrigoriadis The Jensen is an improved version of the Royer. The Royer saturates the main power transformer to make it oscillate. This causes huge current spikes at turn-off and makes the design of the transformer more difficult. The Jensen only saturates the base drive transformer and can use a magnetic material that has a more square B-H loop characteristic than most ferrites. This gives the designer an extra degree of freedom by seperating the power transformer design from the oscillator design. More modern designs use a semiconductor oscillator and you can optimise the power transformer design with having to worry about saturation. The Royer and Jensen topologies are very crude by modern standards but I find it fascinating that you can create a usable power oscillator with only passive components.
Hi Roger, Thankyou again for the Info ! Topologies like the Royer, ZVS and co. need a good Understanding on the working of the Passives used, but as you mentioned with only a minimum of Active Parts you have a Powerful Oscillator. On the other side they are more difficult to control. For example, to make a Adjustable CC-CV DCDC based on a ZVS or Royer Topologie. This is a Project I am on it currently(as a Proof of concept) and use a ADJ Buck to feed the Input Voltage to the ZVS while getting Feedback to the Buck from the ZVS Secondary HV Output, so the CV Parts is more or less(restrictions on the Input Voltage and the Bias of the Mosfet Gates) done but with quite an effort... The CC Part with Foldback is open Yet. Cheers, Konstantin
So the circuit doesn't resonate based on the piezo's reactance, but instead they just cancel out the capacitance with the large inductor? I thought for sure the frequency would be determined by the piezo's characteristics, not the other passive components. If that is the case, then why use a Jenson circuit at all instead of a 494 or another IC to set the frequency? Wouldn't a ZVS driver work even better for this application?
Excellent video thank you. I have one of these that stopped working or putting out as much power as it should. I have always been suspicious that the resonate circuit needed tuning somehow but I have no idea how and your circuit confirms that it is setup at production somehow. I am surprised there isn't some way to tune it to match variations in components. Im certainly not going to connect my scope to it to investigate :)
Hello Campbell, Thankyou ! If you use a scope, use either an Isolation Transformer or a High Voltage Differential Probe or a Battery Powered Scope. The whole PCB is Hot(mains) + the Higher Voltage Output to the Transducer. Cheers, Konstantin
I have one of those and I can assure you there is a lot of voltage that thing produces. I shocked myself bumping a clip wire attached to the transducer and I definitely felt it. Be careful with that circuit, its dangerous for sure. Luckily i only briefly touched the wire with the back of my hand. It definitely lands a punch. I learned my lesson. I had no idea what voltage it was at when I was assembling my test rig, but I found out quick, its enough to guve you a bad day if you aren't careful
Hello, the whole Driver Board is under Mains and rectified mains what makes it much more dangerous, the Transducer Voltage is an Additional Problem. Cheers, Konstantin
If you have a ultrasonic cleaner with multiple driver boards, the output of the driver boards have to be in sync to not create destructive interference right? So what if you replace the self oscillating circuit with another one that you could control the frequency (maybe with an arduino or esp32 for the ease of use and programming)
Hello, theoretically possible, but not feasible due to the higher complexity and costs. These are standard industrial boards, used in most of the cleaners. Check the comments of this video, there are some very good explanations about the synchronicity of the transducers. Cheers, Konstantin
Sir, whats the best way to tune this circuit for other frequency? I understand its with the gate driver tri-filar transformer, but is only with LTspice?
Hello, The Frequency is specified by the used Transducer, to change Frequency the transducer has to be changed the L2 Inductor and the T1 & T2 need to be recalculated eventually.
Thank you sir, I've been playing around with ZVS driver now for a couple of years now powering CRT transformer for fun. So yes I have 100khz transducer and was wondering if lt spice is a good software to simulate the circuitry if I plan on rewinding the torroid, the choke and the pulse transformer.
Hallo :-) No I didn't made a Destructive Tear down to get the winding data of the Transformers. I did not measured the Core Size etc either bcs I didn't thought it would be of interest. Grüsse aus Wien, Konstantin
if given 25khz input, can it be done? or does it have to be modified? and is it DIY? if you buy it, can you ask for a purchase link? sorry for asking a lot😁
1) Frequency is fix, matched to Transducers, 2) DiY !?!? that's a fully reverse Engineered Design from a Commercial Ultrasonic cleaner, please read the Description of the Video.
All ultrasonic cleaner circuits must be like this; they all hum when operating. I expected them to be silent when operating, because, well, ultrasonic means inaudible. But they all hum.
Hi John the Audible noise is bcs of the 100(120)Hz that are superimposed over the 40kHz Transducer Signal due to the non existing Smoothing after the Rectification. Btw. in my video I Recorded while the Thing was Operating on Purpose, only bcs we do not hear the High Frequencies, that does not mean they are not existent :-) The Cleaner Produces many many dB's of Ultrasonic sound. The microphone of my Mobile was so overwhelmed from the High dB Ultrasonic, that you could not hear my Voice even i was 20cm behind the Mic. (I use both the front an back Mic's of my Phone)
Hi Konstantin! Great video!! I relived the memories of my work from twenty years ago.
The intention of the large L2 is to be the dominant inductance in the LC tank circuit formed with the ceramic.
Thus, the energy of the LC tank circuit does not return to the primary circuit through transformer T2 (since L2 is uncoupled). This way, you can inject energy into the tank through T2, and this energy keeps bouncing back and forth between the ceramic and L2.
Hello, Thankyou and Thankyou very much for the Info !
This topology is called a Jensen converter. It's principle of operation is based on the saturation of the current transformer that feeds the base's of the power transistors.
When the converter is first turned on, one of the power transistors turns on before the other (due to gain mismatch) and the high value pull-up resistors to the positive rail.
This is what initiates the oscillation. Once one switch turns on it causes a current to flow in the current transformer switching the "on" transistor on harder and switching off the other transistor. The current will continue to flow until the volt-second product of the toroid is reached, whereupon the core saturates and ceases to be a transformer, thereby removing the gate drive to the "on" device. The energy stored in the core causes the polarity across the windings to reverse (Lenz's law) switching on the opposite transistor and the cycle repeats. Thus the flux density in the CT core flips between saturation in each direction twice per cycle. The load I expect capacitive and is series resonant with the big inductor. So at resonance the two reactive components will cancel out and the equivalent circuit will simply be a resistance that represents the losses due to the heating of the water in the bath. This equivalent loss resistance needs to be included in your simulation to get a better result.
From a design perspective the saturating core cross sectional area is adjusted so that its resonant frequency coincides with that of the LC of the inductor and peizo transducer.
Incidentally, this topology was also used in "electronic transformers" used for low voltage halogen globes in past years. It was also used in some HF florescent tube ballasts.
Hi Roger, Thankyou for the Thorough Explanation of the Working Principle. I searched on Google and found that Jensen is also called Royer or Jensen-Royer, bcs I never heard it called Jensen. The Circuit used, looks like a mixture of a Royer and an HalfBridge due to the Capacitive Divider connected to the one Side of the Power Transformer.
@@KonstantinGrigoriadis The Jensen is an improved version of the Royer. The Royer saturates the main power transformer to make it oscillate. This causes huge current spikes at turn-off and makes the design of the transformer more difficult. The Jensen only saturates the base drive transformer and can use a magnetic material that has a more square B-H loop characteristic than most ferrites. This gives the designer an extra degree of freedom by seperating the power transformer design from the oscillator design. More modern designs use a semiconductor oscillator and you can optimise the power transformer design with having to worry about saturation.
The Royer and Jensen topologies are very crude by modern standards but I find it fascinating that you can create a usable power oscillator with only passive components.
Hi Roger, Thankyou again for the Info ! Topologies like the Royer, ZVS and co. need a good Understanding on the working of the Passives used, but as you mentioned with only a minimum of Active Parts you have a Powerful Oscillator. On the other side they are more difficult to control. For example, to make a Adjustable CC-CV DCDC based on a ZVS or Royer Topologie. This is a Project I am on it currently(as a Proof of concept) and use a ADJ Buck to feed the Input Voltage to the ZVS while getting Feedback to the Buck from the ZVS Secondary HV Output, so the CV Parts is more or less(restrictions on the Input Voltage and the Bias of the Mosfet Gates) done but with quite an effort... The CC Part with Foldback is open Yet.
Cheers, Konstantin
@@KonstantinGrigoriadis It's a pleasure.
So the circuit doesn't resonate based on the piezo's reactance, but instead they just cancel out the capacitance with the large inductor? I thought for sure the frequency would be determined by the piezo's characteristics, not the other passive components. If that is the case, then why use a Jenson circuit at all instead of a 494 or another IC to set the frequency?
Wouldn't a ZVS driver work even better for this application?
Thank you for the Video. I was curious how it works.
Excellent video thank you. I have one of these that stopped working or putting out as much power as it should. I have always been suspicious that the resonate circuit needed tuning somehow but I have no idea how and your circuit confirms that it is setup at production somehow. I am surprised there isn't some way to tune it to match variations in components. Im certainly not going to connect my scope to it to investigate :)
Hello Campbell, Thankyou !
If you use a scope, use either an Isolation Transformer or a High Voltage Differential Probe or a Battery Powered Scope. The whole PCB is Hot(mains) + the Higher Voltage Output to the Transducer. Cheers, Konstantin
I have one of those and I can assure you there is a lot of voltage that thing produces. I shocked myself bumping a clip wire attached to the transducer and I definitely felt it. Be careful with that circuit, its dangerous for sure. Luckily i only briefly touched the wire with the back of my hand. It definitely lands a punch. I learned my lesson. I had no idea what voltage it was at when I was assembling my test rig, but I found out quick, its enough to guve you a bad day if you aren't careful
Hello, the whole Driver Board is under Mains and rectified mains what makes it much more dangerous, the Transducer Voltage is an Additional Problem. Cheers, Konstantin
I looked up the drivers a few months ago, and they seem to all be spec'd 800-1200v for the 60w versions. I bet you won't be doing that again! 😮
If you have a ultrasonic cleaner with multiple driver boards, the output of the driver boards have to be in sync to not create destructive interference right? So what if you replace the self oscillating circuit with another one that you could control the frequency (maybe with an arduino or esp32 for the ease of use and programming)
Hello, theoretically possible, but not feasible due to the higher complexity and costs. These are standard industrial boards, used in most of the cleaners.
Check the comments of this video, there are some very good explanations about the synchronicity of the transducers. Cheers, Konstantin
Sir, whats the best way to tune this circuit for other frequency? I understand its with the gate driver tri-filar transformer, but is only with LTspice?
Hello, The Frequency is specified by the used Transducer, to change Frequency the transducer has to be changed the L2 Inductor and the T1 & T2 need to be recalculated eventually.
Thank you sir, I've been playing around with ZVS driver now for a couple of years now powering CRT transformer for fun. So yes I have 100khz transducer and was wondering if lt spice is a good software to simulate the circuitry if I plan on rewinding the torroid, the choke and the pulse transformer.
Hello, LTSPICE it just a Simulation Software, yes you can use it to Simulate the Results of your changes.
Do you have more details about the transformers, e.g. core size and material, wire data etc?
Grüße aus Münster NRW nach dem schönen Wien!
Hallo :-) No I didn't made a Destructive Tear down to get the winding data of the Transformers. I did not measured the Core Size etc either bcs I didn't thought it would be of interest. Grüsse aus Wien, Konstantin
Hello, What is the working frequency bandwidth?
Hello, The working frequency is ca. 40kHz
if given 25khz input, can it be done? or does it have to be modified? and is it DIY? if you buy it, can you ask for a purchase link? sorry for asking a lot😁
1) Frequency is fix, matched to Transducers, 2) DiY !?!? that's a fully reverse Engineered Design from a Commercial Ultrasonic cleaner, please read the Description of the Video.
All ultrasonic cleaner circuits must be like this; they all hum when operating.
I expected them to be silent when operating, because, well, ultrasonic means inaudible.
But they all hum.
Hi John the Audible noise is bcs of the 100(120)Hz that are superimposed over the 40kHz Transducer Signal due to the non existing Smoothing after the Rectification.
Btw. in my video I Recorded while the Thing was Operating on Purpose, only bcs we do not hear the High Frequencies, that does not mean they are not existent :-) The Cleaner Produces many many dB's of Ultrasonic sound. The microphone of my Mobile was so overwhelmed from the High dB Ultrasonic, that you could not hear my Voice even i was 20cm behind the Mic. (I use both the front an back Mic's of my Phone)
Sir, why is it difficult in our country (Indonesia) to get (online, 1 month) sheet transducers while in other countries it's easy ?
Hello, unfortunately, I do not know!
RF and Magnetics to me are almost like vodoo and black magic.
Well, they are a little bit of Both :-) especially when it you go in the Microwave Frequencies and above.
Can I get the schematic?
Hello, Contact me via my Channel Email to send it to you.
Cheers