Is 1% feasible?
Medium pressure UV. Would be nice to measure 20kW but I'd settle for 6kW.
Thanks!
I was checking out the specs on Yokogawa's high end meter 3000 or something and its rated 0.014 x KHz % of reading+0.3% of range
this means you're going to be at +/- 1.7% at 100KHz. Is it liquid or air cooled?
With 20kW output, you have got to have good 1-2kW of ballast loss which would allow reasonably accurate thermal measurement not too far off from above.
You'd measure the input power at line frequency which can be done to fractions of percent with a relatively cheaply leased power analyzer and use the coolant flow rate or draw entire cooling air flow through mass air flow sensor and deduct this from the input power. You could alternatively put a pass-through water heating element downstream of ballast and feed power to the element through a variac and through a second channel of the power meter. You adjust the power so delta T in coolant across ballast and delta T across heater matches.
Deduct this value from total input and you have the lamp wattage.