I have a 74HCT14N based ''RC'' square-wave oscillator. The produced signal then goes to PMIC gate-driver and finally to a mosfet which drives the primary coil at resonant frequency. (about 930-1000kHz) It is a hobby project, and I'm learning new every day.

I'm having problems with the stability of the oscillator/power supply. I'm powering the entire thing from ATX PSU with 5V, 12V voltages and ground from molex adapter.

In the schematic below all GND pins are eventually connected to single wire. This includes the secondary coil ground wire. However without that connection the circuit didn't work. Also U3 and U4 PMICs are powered from same +12V PSU rail as the primary coil.

Observations: The circuit draws about 3-5Amps from 12V when there are sparks and few 10mA from 5V rail. When I probed the mosfet gate the signal looked okayish.. when drawing an arc there is sinusoidial ringing on top of the square wave. Also the oscilator may eventually glitch out if I pull big enough sparks or move objects near it. (like a finger)

So the question is how do I isolate the oscillator and PMICs enough that the primary coil+secondary coil doesn't interfere with them? I was thinking that maybe adding LM317T + capacitance multipler to supply U2 with clean +5V would help?


  • \$\begingroup\$ provide a photo of your circuit, so the GROUND paths can be examined \$\endgroup\$ – analogsystemsrf Sep 9 '19 at 5:30
  • \$\begingroup\$ I should have probably mentioned the circuit is currently only assembled on breadboard. GND pins are tied together on single breadboard rail. \$\endgroup\$ – JATothrim Sep 10 '19 at 3:59

I don't see a bypass cap on U2 -- or any of the chips, for that matter. You should put 100nF caps on all three, positioned as close as you reasonably can to the power supply leads.

If that doesn't work, figure out the current draw of U2 and put a series resistor between the +5V supply and the chip, sized to drop no more than 100mV.

| improve this answer | |
  • \$\begingroup\$ I have tested putting by-pass caps and they don't seem to help with stability too much. I'll try that resistor trick tough. \$\endgroup\$ – JATothrim Sep 10 '19 at 4:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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