Smoothing capacitor on 3 phase rectifier

[Submariner]

I charge a large 48vdc battery bank from the small micro-hydro plant. It drives a 3 phase induction generator (delta) -> transformer (delta - delta) -> full wave rectifier -> battery bank. About 100 amps at 60+ volts max power. The frequency varies according to load, but is in the range of 50-60 Hz.

I built the rectifier from discrete components, it's worked well for several years but after a recent experience repairing a single phase 36vdc golf cart battery charger, that stopped charging due to a loose connection on the smoothing capacitor, I'm now wondering if the efficiency of my 3 phase charging system could be increased. Clearly not as much as a single phase, since the ripple would be much less but maybe still worthwhile.

I believe the formula for required capacitance for a single phase full wave rectifier is: C = (I * t) / V. where t is the half wave cycle time.
I've been unable to discover the formula for 3 phase full wave though, seems clear there would need to be an adjustment for "t". Maybe "t/3" ? Seems too easy.

Any thoughts or advice would be greatly appreciated.

 

[KlausST]

Hi,

we need to see your schematic.
Without details I´m not sure you need a (big) smoothing capacitor at all.
A battery usually can stand the current ripple of a three phase system.

Klaus

 

[Submariner]

I'll try to dig out the schematic.

I have plenty of power most of the year, but in the hot summer months I have to dial the hydro way down, supplement it with solar (it's a bad area for solar), and it's still not enough. The issue is really the efficiency during this low power time, where even even 5% would be useful.

 

[Easy peasy]

Assuming Pb-acid batteries - for charging only - a bit of ripple is beneficial - if charging and suppling standing loads at the same time then no ripple is better, an LC ckt off the charger will reduce the ripple to whatever you want, e,g, 100uH 100A Cu/iron choke followed by a heap ( say 10 ) of 4700uF 80V electro's will reduce the 300Hz ripple from the 3 ph B rect to fairly low levels....

 

[crutschow]

In general reducing the ripple with a larger capacitor and without an inductor will likely reduce efficiency, not increase it, since the diodes will be conducting a larger current pulse over a shorter interval.

 

[Submariner]

Assuming Pb-acid batteries - for charging only - a bit of ripple is beneficial - if charging and suppling standing loads at the same time then no ripple is better, an LC ckt off the charger will reduce the ripple to whatever you want, e,g, 100uH 100A Cu/iron choke followed by a heap ( say 10 ) of 4700uF 80V electro's will reduce the 300Hz ripple from the 3 ph B rect to fairly low levels....

I'm charging and supplying. I have a couple of DC heat pumps and a DC well pump that are battery driven, and a pair of 5500W inverters supplying 120/240 VAC to the residence. Sound like I should leave it alone.

--- Updated Feb 23, 2022 ---

In general reducing the ripple with a larger capacitor and without an inductor will likely reduce efficiency, not increase it, since the diodes will be conducting a larger current pulse over a shorter interval.

Thanks you for that interesting bit of information.