Hi All,
Kind of following on here from a previous post regarding accurate regulation of high voltage outputs with 12-bit ADC's - I need to measure 0.1V of ripple on a 6000VDC output. 12-bit ADC's cannot achieve the required resolution I am looking for.
My high voltage output is made up of four rectifier and multiplier stacks, connected in series, each providing 1500V. Since each stack has two capacitors in series (doubler), each capacitor only sees 750V.
My question is, if designing a control system measure and regulate the voltage across one single output, how accurately will the other three track this voltage? My thinking is that these outputs are much lower voltage and most likely have higher ripple, so a 16-bit ADC could very accurately control the set DC level. All that is left then is second-stage LC low-pass filtering to achieve the desired ripple voltage, with the DC level controlled much more accurately at a stage prior to the filtering, which introduces a fair bit of phase shift anyway into the system.
I was also thinking of using both these measurements in the system - a 16-bit differential voltage measurement across one capacitor, and a 12-bit measurement just to check the 6000V output and ensure the voltages are being shared as desired across the series connected chains?
Any advice on this method or how it could be improved?
Best,
SIC
Kind of following on here from a previous post regarding accurate regulation of high voltage outputs with 12-bit ADC's - I need to measure 0.1V of ripple on a 6000VDC output. 12-bit ADC's cannot achieve the required resolution I am looking for.
My high voltage output is made up of four rectifier and multiplier stacks, connected in series, each providing 1500V. Since each stack has two capacitors in series (doubler), each capacitor only sees 750V.
My question is, if designing a control system measure and regulate the voltage across one single output, how accurately will the other three track this voltage? My thinking is that these outputs are much lower voltage and most likely have higher ripple, so a 16-bit ADC could very accurately control the set DC level. All that is left then is second-stage LC low-pass filtering to achieve the desired ripple voltage, with the DC level controlled much more accurately at a stage prior to the filtering, which introduces a fair bit of phase shift anyway into the system.
I was also thinking of using both these measurements in the system - a 16-bit differential voltage measurement across one capacitor, and a 12-bit measurement just to check the 6000V output and ensure the voltages are being shared as desired across the series connected chains?
Any advice on this method or how it could be improved?
Best,
SIC
