Automatic alternator cut out/regen braking

[NateTG]

What is the typical current and voltage production from an alternator?

[Randy]

The typical voltage is 14.4v, with the amperage being whatever it needs to maintain that... 8 or so amps just to run the motor, then add whatever else is on.

A normal internal-regulator alternator has 3 connections (plus ground)... the main output wire, the dash light wire, and the voltage sense wire. Current through the rotor's field coil generates the output. The regulator's job is to ground out the low end of the field coil just enough that the sense line is at 14.4v. The dash light provides a small amount of current to the regulator and field coil to bootstrap it.

Since the main output will keep rising until the sense lead reaches 14.4, adding resistance in the main line will increase the load, at least until the voltage maxes out. Just try it... you'll probably get 18+v at the output lug. Even if you cut the output lead, you're not at minimum load, as the field would be at full current. Cutting the field wire would fix that, but then you'd have to deal with the inductive spike from the field coil.

The sense lead is the input to the control loop... if you want to control it, that's the one to tap into. Higher voltage means lower output. So if you put a voltage source in between the output and the sense line, it would reduce the output by that much. Current would probably be less than 100mA, but I haven't tested it. In any case, adding in extra voltage is a lot harder than taking it away, even at low current.

That's what makes the Honda alternator nice: the extra input is positive-going, so lower voltages lower the output.

[NateTG]

Originally Posted by Randy

The typical voltage is 14.4v, with the amperage being whatever it needs to maintain that... 8 or so amps just to run the motor, then add whatever else is on.

Hmm, so the alternator is producing 115 Watts or so of electricity. I wonder if that's within reach for 'bottoming' off of radiator heat with thermoelectrics. (e.g.
http://www.allelectronics.com/cgi-bi...C_COOLER_.html
)

[jim-frank]

I'm a little skeptical about any significant savings, here. The alternator's load on the engine is directly proportional to its current output. That means that when it's just producing enough electricity to run the onboard electronics (once the battery is topped off) it's using far less than a horsepower, maybe as little as 250 watts, about a third of a horse. I can see some potential mileage improvements by going with the original idea of using it as a regenerative engine brake, but I really can't see how it could help under any other scenario.

[jpgmtech]

Originally Posted by jim-frank

I'm a little skeptical about any significant savings, here. The alternator's load on the engine is directly proportional to its current output. That means that when it's just producing enough electricity to run the onboard electronics (once the battery is topped off) it's using far less than a horsepower, maybe as little as 250 watts, about a third of a horse. I can see some potential mileage improvements by going with the original idea of using it as a regenerative engine brake, but I really can't see how it could help under any other scenario.

I am too, that's why it would be good to physically test this. A battery at saturation will present the highest impedance to the alternator, which will require the least current output from the alternator to maintain at the regulator set point (14.4 volts for this discussion). The battery (and other accessories) are the load, or impedance that the regulator "sees" by it's voltage measurement. (caveat: this applies to typical Delco alternator)

A little of Ohms law: I=V/R (I=current in Amps, V=volts, R=resistance in Ohms). The final voltage at the alternator will always remain constant, based on the regulator setpoint (which varies with temp) but we will assume 14.4 volts. If the battery is low and requires charging, its resistance (actually impedance, but for the simplicity of explaining this) is low. Assuming the regulator attempts to go to 14.4 volts, with a low resistance, that means current output is forced upward. If the battery is saturated, its R is high while the alternator is still shooting for 14.4 volts, that means the current has to drop correspondingly. The current is controlled by the regulator's control of field voltage. In order to test the minimum field voltage, a 1/4 ohm resistor is introduced in the main battery positive lead (which includes the voltage sense as well, and all accessories are turned off) This effectively simulates the impedance of a fully charged battery. The regulator requires less current to maintain 14.4 volts output, as only R has changed. The 1/4 ohm resistor is enough to drive the regulator to minimum field voltage even with a battery not fully charged. This is an actual documented test of the Delco regulator.

Why? It would be good to compare the minimum drag of the alternator while turned on (minimum field voltage) to the drag of the alternator while turned off in order to determine if there will be any fuel economy benefit.

[Johnny Mullet]

This was brought up on another forum and an idea popped in my head. One guy suggested hooking an alternator cutoff switch to react with the brake light for regenerative charging, but it was also brought out that most of us FE guys avoid the brakes and it would not be sufficient enough to keep a charge.

I thought about this "alternator cutoff" deal right when I popped the hood on my Chevy truck. The hood light came on from the momentum or position of the mercury switch and that's when it all seemed like a feasible idea. Why not implement a mercury switch and if you get it tuned correctly and adjusted properly, it should kick on the alternator current as soon as the vehicle begins to slow or the momentum of the car decelerating does it.

This might work, but I have no plans on ripping out the switch on my truck's hood light and hooking it to the Metro just yet.

[nefis]

I'm using a signal from my neutral safety switch(Park-Neutral-Position Indicator) to open a relay that kicks on my alternator. I've got an automatic transmission, so when i'm coasting, braking or stopped i have my car in neutral and my alternator is charging my battery(deep cycle battery in trunk). I've also got a switch/led indicator on my dash so i can have the alternator running all the time(if driving at night), and i monitor my battery/alternator output with a cigarette lighter voltmeter. My commute every day is only 26 miles round trip and my battery seems to be handling the load...Battery is recharged every night.

[nnachshon]

Originally Posted by Randy

I modified my car to not charge via a switch. All the 5th generation Civics have an extra wire on the alternator, which if you ground will reduce the voltage to 12v from 14v. It runs from the ECU to the alternator, which has an internal regulator. The ECU drives it with an open-collector circuit (at least on mine), so grounding it out with a switch doesn't hurt anything.

Does it help the MPG? I have no idea. It dims the lights, and reduces the load due to the lower voltage. It won't kill the battery because it's still above the discharged voltage.

If this wire wasn't there, shutting off the alternator wouldn't be so easy. Alternators self-excite as long as the regulator is in the circuit. Cutting off the field or output via a relay will cause spikes, possibly damaging the regulator. The best cure is a new regulator... a switching regulator would be more efficient than the standard linear type anyway. The other way to go is to add a clamping circuit to absorb the spike. This could be as simple as a diode, but you'd need to test it.

I just saw this thread (I'm new), which years are you refer as 5th generation?
I have Civic 2004, I do know that Honda control the alternator load throw measuring the current consumed by all the electrical loads (its even called ELD and it located underneath the Engine fuse box).
As I understand Honda do shut-off/on the alternator based on current consuming and of course ECU control.
Never tested it to see when its working on not, but my guess thats Honda know how to design cars


Thanks
Nir

[Clev]

Make sure you do some Googling to make sure your alternator can handle this. I had an alternator burn up on my '90 Accord after struggling to charge a nearly dead battery. I've since read that the alternators on these tended to overheat if asked to put out full power for extended periods of time. In my case, next time I need to replace the alternator, I'll go with a more powerful alt that can handle the load.

[RH77]

How far back does the honda electrical load system go?

I can feel the engagement of something during cruise -- like a muted AC compressor (on the '98 Acura). The batt is way overdue and I haven't determined that it's the cooling fan or this on/off alt load described. Whatever it is, robs power and kills FE for about 10-15 seconds and kicks back out for a few minutes. I drive w/ my headlights at all times for safety, which is the only constant load, aside from overall vehicle draw. The blower is in the Off position.

RH77