# FET Drive for 200A Battery Charger Question

I have 12V, 650A-hr house batteries I want to charge from the alternator when traveling. The alternator has a continuous rating of 240A @ 14.6V. The alternator regulator drops the voltage some.

The charger will be 3 stage, max 200A (by shunt feedback). I will sometimes want to operate the FETS at less than "full on", somewhere is the 4 - 8Vgs range. When I do this, the voltage out to the house battery will be lower and the battery will accept fewer amps. At Vout = 13.2V, the house batteries will be float charging, consuming about 10A total.

Because this is a one-off product, the additional cost for FET count overkill is not an issue, nor is the cost for generous heat sinks and multiple NTC temperature sensors (not shown) that I will attach to the FET copper bus bars.

I am seeking comments whether the attached circuit will work for what I intend?

• Why not throttle the alternator in various ways (control the generator RPM, control field yourself i.e. replace the internal regulator)? Saves a huge pile of trouble trying to make a linear regulator circuit. Alternately, you might as well double down on complexity and make a multi-phase interleaved buck converter; switching gives high efficiency. Commented Oct 29, 2023 at 2:49
• Messing with my engine's alternator is out of the question for me. I did investigate buck/boost converter, but quickly gave up. The first obstacle was specifying and obtaining the necessary inductor. Also, not particularly interested in doubling down on complexity. I'm not an EE. The topology of the brute force back to back P and N FETS is pretty simple to understand. It's the implementation of the FET drive I am asking about. Commented Oct 29, 2023 at 21:30
• The problem is more rudimentary than you think: as shown, you've drawn the PMOS backwards, so that the alternator is dumped into the battery regardless of control, through the body diode (which these symbols show twice: the triangle in the middle is the channel-substrate (body) diode structure, and the parallel diode reinforces this symbolically). Correcting for this, you still have no means of enforcing sharing between devices: naively in parallel, one will take almost all the current, and overheat in a runaway condition; nor implementing proportional control. Commented Oct 29, 2023 at 23:22
• You may still find a charge controller unit for such ratings, and power input; or perhaps you are willing to contract someone for design of this. But I would suggest playing to your strengths, and finding the simplest possible solution, whether mechanical or electrical. For example, simplest of all would be to find an alternator that simply can't supply more than the specified current, and just accept that you won't have adjustable range (it's always set at max; but no more than that). Commented Oct 29, 2023 at 23:24
• Explaining in more detail what you want that cct to accoplish would help. I do not know what the NFETS are for or how they function. || Without feedback cointyrolling that cct in otyher than off on is not viable (once actual cct issues are corrected). || I'd prtefer an off / on contyroller with either a capacitor to smooth the mean effective switch resistance OR makje it a buck regulator with an inductor and diode. That is less hard than you may imagine and FETs are MUCH better treated. Inductor can be hand wound on an old eg microwave over transformer core or ... . Commented Oct 30, 2023 at 6:44