Identifying and Improving Oscillation in Ultrasonic Cleaner Circuit

Discussion in 'The Projects Forum' started by Mike951, Mar 13, 2015.

  1. Mike951

    Thread Starter New Member

    Mar 13, 2015
    Hi, I am currently working on a project to basically build an ultrasonic cleaner. My design involves powering 3 pairs of transducers (each pair connected in parallel) with 3 ultrasonic generators, for a grand total of 300W at roughly 500V and [hopefully] 40KHz. What I have bought can be found here. That being said, I have built the device and epoxied the 6 transducers to the bottom of a stainless steel kitchen pot (a few gallons in size) and powered the whole setup. It sort of works, but the the transducers run at 36KHz instead of 40KHz, the system is very loud (lost energy), and it only just barely makes tiny holes in aluminum foil after about an hour of cleaning. This is compared to a industrial one of similar size (maybe more power) I was able to use which can make holes in aluminum foil almost immediately. So my problem is one of either increasing efficiency or power in some way to make this system actually useful.

    I have been reverse engineering the ultrasonic generators to see if I could figure out a way to improve them, but after determining the schematic and getting values for all the components, I have fallen into a slump. I am not entirely sure how the circuit works past a point. As far as I understand, the diodes on the top left are being using to rectify the voltage, after which a transformer increases the voltage and the transistor setup causes the circuit to oscillate using feedback from the load (my research tells me ultrasonic transducers are similar to an LC circuit). Note the load (2 parallel transducers) connects to the two pins labeled "out" on the right side of my schematic.

    Particularly, I am curious if some of you are familiar with the "push-pull" (i'm assuming) power transistor design that is being used in the middle of my attached schematic. Is it possible for me to tweak some resistor or capacitor values to get better performance? Thoughts and suggestions are welcome. I have verified all the connections, part numbers, and sizes to the best of my ability. I am considering making my own digitally controlled circuit to get the frequency I want instead of this 36KHz output I see from a scope.
    Last edited: Mar 13, 2015
  2. Mike951

    Thread Starter New Member

    Mar 13, 2015
    It appears I can no longer edit my post. Regardless, also note that my tests indicated that the total power I was using was about 17oW to power 6 transducers rather than the 300W I was hoping for. Transducers generally running 36KHz during this time-as mentioned previously. The voltage output from each separate generator was usually different from others by about 20V.
  3. Dodgydave

    Distinguished Member

    Jun 22, 2012
    The circuit is a self oscillator using a split power rail provided by C2,C3, which is a Live psu, used alot in Atx psus.

    The output comes from the transformer T1, terminals 1,2 to feed the transducer, so increasing the number of turns on that will boost the signal,

    you could try to put a transducer across the terminals directly.
    Last edited: Mar 13, 2015
  4. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    The advantage of a self-oscillator is that it automatically adjusts to suit the natural resonant frequency and acoustic loading of the transducers. Transformer T1 is crucial to success. In particular, its core must not saturate; so trying to get more juice out of it could be difficult.
    Hypatia's Protege likes this.
  5. Mike951

    Thread Starter New Member

    Mar 13, 2015
    Interesting and good to know. Dodgydave, when you say that increasing the turns will boost the signal, are you referring to voltage? I will spend some more time looking into self-oscillators. I find it odd that its automatic adjustment takes it to what I believe is a lower harmonic of the resonant frequency of the transducers. It's also odd that each transducer is only using about 30 of its advertised 50 Watts since each generator is supposed to do 100W, but i'm getting 180W total. This lends a little more credence to one my theories that perhaps the tank mounting may be non-ideal, leading to decreased output. Something I have noticed with my ultrasonic generators is that they used to output around 500V, but after using them for an hour or two at a time, they're running at 400V with a similar tank load.
  6. Paulussie

    New Member

    Jul 21, 2015

    I was wondering, did you get any further with this? Today I reverse engineered a ultrasonic generator to be able to modify it so I can tune the frequency. Apart from some extra filtering in the AC side and some difference in parts (parallel transistors, component values) the schematic looks the same as the one you took apart.

    I found that if I replace the 510R resistors with (for the time being) trimmers, I am able to influence the frequency somewhat, but not to a point I was confident I understand what happens. A search found me this thread, so I was wondering, did you get any further in understanding this self-oscillator thing?