Poor man's 15 Amp battery charger?

[Teh Stork]

I've worked some more, with this as a start.

I've decided to make a buck with transformer isolation after I've finished my boost - psu charger. This is the fast scetch I've made on this circuit: (It's missing a lot of resistors, capacitors and ground symbols)


It's duty cycle will look something like this:


Imo this could be a fast and safe charger. The transformer is a 1:1 ratio. It's just there to provide isolation. The switching and deadtime is controlled by a microprocessor. It will use mosfets capable of withstanding 400vdc, the positive is that its load is more or less continous (less EMI) - the negative is increased losses since these mosfets has a bit higher on-resistance. A current sensor is utilized to sense the current - and adjust dutycycle accordingly.

If anyone would like to start, I could help out with the avr software, given that you prototype it with a xplained 1284p AVR (these are cheap and nice to prototype with).

[Teh Stork]

As a "proof of concept" I built this boost converter to charge my battery. It takes everything from 12-20V input - and can charge based on either input voltage sag or current sensor feedback. It monitors output voltage and cuts off once it's reached.

It uses a irfp2907 fet, a small mosfet driver, a homewound toroid (75uH, 6mOhm) and a bridge rectifier being used as a diode. Planning to use my experiences to try to make a buck - ac connected charger.

[bearing]

That's a good idea. I've been toying with the idea of making a multi phase buck directly from mains. If you have more phases than the ratio between battery voltage and peak mains voltage, then the converter will conduct continuously, which means it will be easy to filter the switching noise from the mains, so that it doesn't pollute the grid.

I think this stuff is worth it's own thread, since this thread is about the passive capacitor method.

[IBScootn]

But I wonder if the SCR or motor-run capacitor is even needed. I just found out that the J1772 stations can be instructed to deliver a max constant current. Setting the value of the resistor inline with the proximity switch determines what that current will be:


total resistance PP-PE 1500 Ω 680 Ω 220 Ω 100 Ω
current capacity 13 A 20 A 32 A 63 A
wire cross section 1,5 mm² 2,5 mm² 6 mm² 16 mm²
[edit]

So you can tell it to deliver 13A, 20A, 32A, or 63A based on a low-power resistor. That would seem to make the SRC or AC cap no longer necessary to limit current.

What do you think?