Hi colleagues,

I have read interesting research titled "Instantaneous Voltage Measurement in PWM

Voltage Source Inverters" authored 2007 by Prof. Todd D. Batzel, of Pennsylvania State University, Altoona College, USA. (DOI: 10.1109/ACEMP.2007.4510497)

Anyway, the method he proposed to overcome the distortion introduced in the inverter output due to its inherent non-linearity (current-dependent device forward drop, the dead-time effect, etc) is simply based on integrating a scaled-down signal of the actual pole voltage (a, b or c) of the inverter along a complete PWM cycle. Then using the acquired value of the integration (volt.sec) and the PWM cycle data to restore the instantaneous mean (dividing the volt.sec by the duty cycle returns the volts). This can be accomplished by a combination of fast switches and integration op-amp. My question is: has anyone used IVC102 (a chip by BURR-BROWN) successfully to meet this requirement?

Does anyone know about a more cost-effective method to do the same task? (this chip is sold @ USD11 and an application may need 2 chips at least to capture the three-phase balanced voltage)