The diodes I'm looking at are applied in 400Vdc and 800Vdc PV input to a VFD DC link, to block the VFD DC link from backfeeding the external PV array.
The full load amps are anywhere between 10Adc and 700Adc with a possible overload of 150% for 1 minute.
PowerModule construction is desired to screw the module directly to a backpanel with convection cooling , avoiding additional heatsinks and diode fans.
------- rationalizations -----
I've selected 1200V and 1600 V PIV diodes.
Solar PV panels have a limited short circuit rating as relative to its full load current.The panels I'm using spec the short circuit current is 6% higher than the full load current, so not too worried about the panel diode instantaneous current as they have robust instantaneous current ratings.
------ the question --------
Referencing diode DZ600N as an example. I for sure need to evaluate a family of diode currents.
The diode spec sheets list "average on state current" at xxx amps at 100C , yyy amps at 75C .
These ratings seem specified in conjunction with a rectifier, where the diode only conducts at most 180 degrees of a cycle. The spec sheet does chart temperature based on the firing angle when used with a SCR, and with 1 ph full wave, and 3 phase full wave bridges.
1) To use a power diode in a DC blocking application where it sees a continuous DC current , does the applied average current minimally need to be 50% of the specified average current (because the amps are with 360 degrees conduction and not 180 degrees conduction )?
2) The DZ600 also has a chart for overload currents based on a six pulse converter. The Overload on-state current IF(OV) for 1 minute when used at 50% average current looks to be a 1.0 multiplier, i.e. no overload left, so the DC applied diode average current needs to be based on the 150% overload capacity rather than 100% full load average.
Have I rationalized all the needed derates ?
Are these derates correct for applying a diode to block DC ?
Would the same diode DC derates be applied if instead of a bridge and capacitive load, a battery bank was used ?
I'm also assuming if a diode has all these derates, its not likely to need additional cooling, or maybe another 30% derate is needed to keep the temperature down ?