rectifier for an 8V 225A electrolysis machine

Thread Starter

bootlegengineer

Joined Dec 5, 2016
60
Hi! I'm trying to build an array of low-loss rectifiers for my electrolysis machine that can support 5V and 360A. I have the transformer set up already (primary 240 windings [12 gauge] and secondary 10 windings [0 gauge]) but am having trouble with the rectifier design. I don't want to use diodes so I plan on using my surplus of fdp047an mosfets. The only problem is that I read online that I need a special controller for this to work, which I don't have. Is there any way I can bypass this requirement with a different IC or maybe just a simple modification to the rectifier itself?

q6ti1id.gif
 

DickCappels

Joined Aug 21, 2008
10,153
From the transformer turns ratio I guess that you want to rectify a sine wave.

You need the control circuitry if you are rectifying waveforms that are not rectangular and charge capacitors. Otherwise instead of looking like diodes, the MOSFETs look like low value resistors driving a capacitor - not much in the way of rectification action.

The control circuitry can be as simple as a comparator and a few small parts but since the power supply represents a significant investment I would go with the expertise embedded in the Linear Technology (now Analog Devices) part.

If you are only driving a resistive load or an averaging choke ahead of the capacitors you can simply switch half of the MOSFETS on at the beginning of each half cycle and the other half of the MOSFETs on the other half cycles. The problem hiere is that you get average rectified voltage across the capacitor instead of the peak voltage you were expecting. In the case of a resistive load you should get the same RMS value as the incoming AC signal and I suspect the same average absolute current that you would get by passing the sine wave through the load.
 

DickCappels

Joined Aug 21, 2008
10,153
Yes, you could drive your MOSFETs with it but you would still have the problem of the MOSFETs being on for a fully half cycle. Think about when current flows through a diode when rectifying a sine wave.

upload_2019-4-28_6-27-36.png

You would only want your MOSFET to conduct while recharging the capacitor, not for the entire half cycle.

If the MOSFET were to conduct for the entire half cycle a good part of the half cycle (when the input voltage is below the output voltage) would be spent discharging the capacitor. If you are charging the capacitor through a choke, then the voltage across the capacitor would be the average of th input voltage.
 

crutschow

Joined Mar 14, 2008
34,280
If you want to use all N-MOSFETs in the bridge, then you need a circuit like the LT4320 to provide the correct gate signals, including a charge pump to generate a gate voltage above peak input voltage to fully turn on the MOSFETs.

If the output is resistive with no capacitors (as the electrolysis load may be), then you don't need a controller, as Dick stated, but you would need to use both N-MOSFETs and P-MOSFETs.
Below is the LTspice simulation of a circuit that will do that:

upload_2019-4-27_17-52-30.png
 
Top