IR2153S Power Supply for Piezoelectric Actuator

Status
Not open for further replies.

jabbar81

Newbie level 4
Joined
Mar 27, 2006
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
Seoul, Korea
Visit site
Activity points
1,371
I am making a Piezoelectric actuator driving circuit using Half Bridge. I have to make it work at 220Vrms, 60Hz line voltage. Currently I am testing it at 100Vrms, 60Hz voltage. The line voltage is rectified and converted to DC link Voltage. IR2153S is used as driving chip to drive piezoelectric actuator at 74kHz, using square output wave.

I am getting burning MOSFETs and IR2153 above 100Vrms main input.

The oscilloscope screenshot of PCB is attached.
Blue-Channel-1: is the Hi-side MOSFET Gate Voltage w.r.t GND
Green-Channel-2: is the Lo-side MOSFET Gate Voltage w.r.t GND
Purple-Channel-3: is the input voltage to Piezoelectric Actuator.(input capacitance of Piezoelectric 0.3nF)
Green-Channel-4: is the input current to Piezoelectric Actuator.


Problems

(point-1): I have dead time when Hi-side MOSFET is getting On.
(Point-2): No dead time, when Hi-side MOSFET is turning OFF, though I can see little transition in voltage when hi-side MOSFET should turn off.
(point-3): High output current pulses/spikes at switches ON/OFF, I think they are responsible for damaging the switches.
(Point-4): ringing in the Lo-side MOSFET gate voltage.

Without load, the dead-time of the gate voltage is correct.

How I can solve the no dead time, current spikes and ringing issues. Any suggestion?
I tried some sunbber circuit, diode in parallel with gate resistors, changing C6, 22-ohm R5, but couldn't solve the any problem specially dead-time problem.



 

Attachments

  • driving_circuit.png
    31.8 KB · Views: 339

There's no deadtime problem, only a misunderstanding of oscilloscope waveforms. Trace 1 is not the high-side Vgs, it's Vg against DC-. The highside transistor is switched off, but the output node is staying at high level til the lowside transistor switches on.

Ringing is a matter of unsuitable probe connection and possibly inappropriate circuit layout.

Output current looks like a capacitive load, not a piezo electric transducer near it's resonance frequency. Can't say more without knowing the transducer parameters. A series inductor may needed to reduce current spikes.
 

Status
Not open for further replies.

Similar threads

Cookies are required to use this site. You must accept them to continue using the site. Learn more…