FET BATTERY ISOLATOR

[kinarfi]

I saw a circuit for a FET full wave bridge and used it, so I figure a FET battery isolator would have the same advantages over a diode isolator, lower voltage drop. What I am seeing on Spiced is that once the FET are turn on, they stay on. I tried to set this up so it had to have alternator voltage and ignition voltage in series to turn on the power PFETs to pass current to the batteries, and if either the ignition or the alternator voltages go to zero, the power FETs turn off, isolating the batteries. Any ideas why the "junction" is not going to zero?
Please don't suggest a relay, anyone!!!! I want to make the FET isolator work.
Thanks, Jeff

 

[MikeMl]

You are forgetting about the intrinsic Source to Drain body diode inside the NFETs

 

[kinarfi]

Thanks Mike, not sure I under stand, would explain it for me and do you have a suggestion of how to make this work?

 

[crutschow]

Why do you want to connect in the ignition switch? Can the alternator have an output without the ignition switch being on?

 

[kinarfi]

Why do you want to connect in the ignition switch? Can the alternator have an output without the ignition switch being on?

Once either of the PFETs have turned on the JUNCTION will be high and gate that is tied to it will be high even when the alternator is off, so the ignition switch is supposed to turn off the other fet and allow the PFETs to be turned off so they can't back feed the JUNCTION. Similarly, if the ignition is on and engine not running, I don't want the to PFETs on until there is an output from the alternator, can you imagine what could happen if the PFETs were on during starter engagement? Also if one battery is dead and the PFETs turn on, it could drain the other battery down to where I couldn't start the RV, I saw that once on a vehicle that had a relay to tie the batteries together.

 

[alec_t]

How about this?


Note that on the rising and falling edges of V(alt) (exaggerated here for demo purposes) there is some cross-feed between the two batteries. I think this would be inevitable whenever there is a 'junction' rail.
R4 is included to limit Vgs.

Edit: The circuit would probably need spike-suppression components added.
Cross-feed between batteries could be avoided by feeding each via its own Schottky from the alty.
Edit 2: The modelled Schottky is rated at only 25A, so a beefier diode or two would be needed in practice.

 

[crutschow]

But wouldn't it be sufficient just to have the output from the alternator turn both FETs on and off?

Edit: As Alec's circuit does.

 

[alec_t]

have the output from the alternator turn both FETs on and off?

As in post #6

 

[MikeMl]

You guys have all missed my original point. How do you explain that the bus remains powered even when the Alt is off?

 

[alec_t]

Body diodes of M1/M2 explain that. But that 'bus' won't be driving other loads, as I understand the OP's intentions. It will only be driven by the alty. I assume all other loads will be drawn from the tops of the 0.1 Ohm resistors (R2,R6)?

 

[crutschow]

You guys have all missed my original point. How do you explain that the bus remains powered even when the Alt is off?

We didn't miss the point. It's just that the circuit was too convoluted to easily follow. Simpler to modify the circuit to do what the OP needs (not necessarily wants).

 

[kinarfi]

I got it Mike, Thanks, so even when the engine is off and both PFETs are off, the line to the alternator will have the same potential as the highest battery minus the voltage drop of the intrinsic diode of the corresponding PFET, but since the alternator has a diode output, there will be no power loss via the alternator and the intrinsic diode of the PFET for other battery is reverse biased, there will be no power loss there either, so this should work.
Thanks,
Jeff

 

[alec_t]

You will need a pull-up resistor on the 'gates' node, to prevent the gate voltage floating when the ignition is off: otherwise the FETs could turn on and allow one battery to drain into the other. Be careful that the Vgs max of the FETs is not exceeded.

 

[kinarfi]

I have it, it's labelled now, first mock up appears to work beautifully, with F1D NOT pulled down to ground, the junction voltage would go up to the voltage of the Main battery (I haven't done it with both batteries yet, tomorrow!) and if I turned power supply up to where it's voltage was higher than the battery's voltage, the PFETs would block the current until I took F1D low and when I let go back high, the current stopped. Voltage drop Vsd was .063 at 2 amps, if I remember right. There was no detectable heat except for the pull up resistor which did not have a high enough wattage rating.

 

[kinarfi]

It seems to be working just fine, how ever it is not a battery isolator, it ties the batteries and the alternator together which allows one battery to drain power from the other when the key is turned on but not started. That being the case, I decided to put in a 30 second delay so the starter battery would not drain the RV battery during start and so that the alternator was running when they are connected. There is a solenoid that also ties the batteries together for the purpose of starting in case the starter battery is down, but it's designed to handle the current which could damage the FETs.

 

[alec_t]

If you use the alty, not the ignition switch, to turn on the FETs as suggested earlier then you don't need a 30 second delay.

 

[kinarfi]

That won't work, the alty voltage is always high via the body diode of the PFETs , the diode you used in post 6 is integral to the alty, as I tried to convey by putting a circle around the voltage supply and the diode, which is needed for the sim so current can't flow backwards through the alty from the the body diodes in the PFETs.
Crutschow, what you mean, convoluted, I thought it was pretty straight forward, simple and easy to understand.
Thank for the help and suggestions,
Jeff

 

[alec_t]

the diode you used in post 6 is integral to the alty

Ah. That makes a difference .