440V 3phase to 0-17V/1000A rectifier design help

using 6 diodes instead of 3 gives better utilisation of the sec windings, a smaller lighter transformer, and you get to spread the heat over 6 diodes. Using only positive pulses out of a Y winding is not the flashest way to run a transformer...

- - - Updated - - -

with a decent output choke the current is square pulses, the phase angle power factor is OK, and the true power factor is not too bad either, a lot bigger choke needed for the 3 diode setup ...

FvM said:

Three versus six pulse rectifier has a large impact on transformer rms current and respectively transformer losses and minimal size.

Regarding harmonic currents, the design misses yet the expected output choke. I presume that you didn't make actual calculations for imagined filter capacitor or estimated current waveforms?

Click to expand...

No, I did not calculated the capacitor value yet. I'm collecting materials from my local market. I'll make one trial first, then final one.

no output caps needed for electroplating - a good sized choke is quite useful tho'

With emphasis in "good sized"

But I agree, the choke will significantly improve the PF.

1. 6 diode will give you full wave and less ripple; you will not need any filter (cap or inductor).

2. 17V is bit high for electroplating but the bath acts like a big capacitor and a resistor.

3. Three phase full wave rectification will give very little ripple and the capacitance of the bath will be enough.

4. If I assume that the typical operating current in the bath is 50% of the max specified, the important point to see that the current is more or less uniformly distributed within the bath.

Hi

Three versus six pulse rectifier has a large impact on transformer rms current and respectively transformer losses and minimal size.

Click to expand...

I didn't simulate it...but i think with three diodes each phase is conductive for 120° per fullwave. 30° ... 150°
With 6 diodes it is two times 60°: 60°...120° and 240° ... 300°. 2 x 60° isv120°, too.
I wonder if the RMS current changes that much. Both systems are 120° per full wave. So the conductive time doesn't change.
For sure there is a lot more voltage ripple.
The 3 diodes: 50% peak to 100% peak
With 6 dides: 86% peak to 100% peak.
With an ohmic load the current ripple is equal to the voltage ripple.
With a series inductance the current ripple will be reduced....in ideal case to about DC...

Where does the large impact on RMS current come from?

****
I do understand that on a one phase system there is a big impact on RMS current, because the conductive time on a half wave rectifier is only 50% of the conductive time of a full wave rectifier.

Klaus

I do understand that on a one phase system there is a big impact on RMS current, because the conductive time on a half wave rectifier is only 50% of the conductive time of a full wave rectifier...

Click to expand...

For a purely resistive load, at const load power utilization, the RMS current is same for both half-wave and full wave rectification; because;

1. you need to consider full cycle for calculating RMS;

2. Transformer current will be same as the load current because the load is resistive;

3. At const power, the load RMS current will be const

4. The transformer RMS current will not be different at const power.

The story will be different at constant resistance. But the power output will be obviously lower.

From the point of view of the load, it only cares for a DC voltage, but a 6 diode full wave 3 phase rectification will give lower ripple. That is a major point from the load side point of view.

For a purely resistive load, at const load power utilization, the RMS current is same for both half-wave and full wave rectification;

Click to expand...

It isn't. With resistive load, current form factor is about 1.05 for 6-pulse and 1.47 for 3-pulse, requiring nearly 50% higher transformer VA for the 3-pulse circuit. In addition, the large DC current bias might request a different transformer design further increase the core size.