1
\$\begingroup\$

I am trying to power an ultrasonic transducer using an H-bridge inverter design to convert an Arduino signal into a higher power AC square wave (shown in the circuit below).

Inverter Schematics

It was working well, then suddenly I started getting high voltage spikes in the rising and falling edge of the square wave. I am using a high varying driving voltage of 15-60 V (VDC in the schematics) and I have seen that increasing this voltage distorts the output more. I am hoping to drive the transducer from 20 kHz to 30 kHz.

Distorted output waveform

This is a zoomed in picture of it: DIstorted output waveform zoomed in

You can see that the top of the square wave is collapsed slightly as well as the spikes on the bottom left of the picture.

I've tried rebuilding the circuit using new components on a different PCB but I'm still encountering the same problem. I am thinking it might have something to do with the MOSFETs (IRF4227), but I have changed them to a different type of MOSFETs (IRF540n) and the problem still continues. I have also changed the bootstraps capacitors to a larger value of 330 nF from the 22 pF in the schematics but it did not do anything for the distortion.

Thank you very much for any help you can give.

\$\endgroup\$
7
  • 3
    \$\begingroup\$ By any chance are you using a standard X10 probe ? And the EARTH clip at some annoying point \$\endgroup\$
    – user16222
    Commented Feb 15, 2019 at 8:50
  • \$\begingroup\$ You'd need ultra fast recovery diodes. Also 1k ohm gate/source resistors are not needed, IMO. From the picture, the scope is kind of DIY \$\endgroup\$
    – Marko Buršič
    Commented Feb 15, 2019 at 9:42
  • \$\begingroup\$ If it isn't just scope pickup, then stray parasitics can also be causing spikes, can you show the PCB layout? \$\endgroup\$
    – Unimportant
    Commented Feb 15, 2019 at 11:09
  • \$\begingroup\$ Yes I am using a cheap DIY Oscilloscope and no I don't think it's a standard x10 probe as it has alligator clips at the end. I have the ground clipped to the same ground as the 12v input which is common to the 5v. \$\endgroup\$
    – HDR
    Commented Feb 15, 2019 at 11:12
  • \$\begingroup\$ @HDR These clips in the picture? If so, that's probably a part of your problems. You need a low inductance probe like shown in this post. Also, trying to view a 30kHz square wave with a 200kHz "scope" won't give you enough harmonics to see the true wave. \$\endgroup\$
    – Unimportant
    Commented Feb 15, 2019 at 11:38

1 Answer 1

Reset to default
1
\$\begingroup\$

You don't have 22 pF bootstrap caps, you have a 10 μF plus 22 pF bootstrap caps! Remove the 10 μF twice. They probably cause the spikes below 0 V, as the cap pushes the voltage downwards when opening one of the lower FETs.
(depending on your layout)

\$\endgroup\$
5
  • \$\begingroup\$ Wait, why are the 22pF drawn as resistors? \$\endgroup\$
    – Huisman
    Commented Feb 15, 2019 at 11:06
  • \$\begingroup\$ In any case, 10uF is overkill and 22pF is too little. 100nF or so would probably be more in the ballpark. \$\endgroup\$
    – Unimportant
    Commented Feb 15, 2019 at 11:13
  • \$\begingroup\$ @Huisman Haha I think it's a mistake but I definitely have capacitors where they are. \$\endgroup\$
    – HDR
    Commented Feb 15, 2019 at 12:33
  • \$\begingroup\$ @Huisman I will try to remove the 10uF caps and see. \$\endgroup\$
    – HDR
    Commented Feb 15, 2019 at 12:34
  • \$\begingroup\$ @Huisman I will also try around 100nF capacitors for bootstraps. Thank you all! \$\endgroup\$
    – HDR
    Commented Feb 15, 2019 at 12:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.