Charge 12v lead-acid battery via pulse width modulation

[Larry Bradley]

I need to charge the 12 volt engine start battery in my sailboat from
the 12 volt house battery in the boat. The house battery is the one
connected to the alternator. I can't just parallel the two batteries
for charging, since the house battery is a long cable run away from
the alternator, and when a heavy charge is going into the house
battery(50 amps or so), the voltage that the start battery would see
could be as high as 15.5 volts. I only need to supply about 3 amps
when the engine is running, just enough to run the fuel pump and
ignition.

I can use a linear regulator, of course. But I thought perhaps I could
use a pulse width modulator driving a switch (FET ot BJT) to apply the
15 volts to the battery in pulses, with the average voltage being
around 13.6 volts. (Sort of a switching regulator without the
inductor).

If I had the battery home with me, I could test the idea. But it is
100 miles away.

Any battery experts out there who would care to comment?

Thanks
Larry Bradley VE3CRX
Remove "removeme" from my e-mail address for direct mail
Ottawa, Canada

(use the e-mail address above to send directly to me)

 

[Larry Bradley]

Yes, using remote sensing, so the regulator sees the actual house
battery voltage.

The peak current was what I was afraid of. I'm no switching regulator
guru, although I've read quite a bit. Probably the best way to go.
Any suggestions as to where to get an inductor suitable for the 5 amps
or so I need?

Use a real buck regulator with an inductor, and avoid nasty peak
currents in your pass device.

Question: If the "house" battery is so far from the alternator, are
you using remote sensing?

...Jim Thompson

Larry Bradley VE3CRX
Remove "removeme" from my e-mail address for direct mail
Ottawa, Canada

(use the e-mail address above to send directly to me)

 

[Jim Yanik]

Use a real buck regulator with an inductor, and avoid nasty peak
currents in your pass device.

Question: If the "house" battery is so far from the alternator, are
you using remote sensing?

...Jim Thompson

why would the voltage to the house battery be 15.5V?
Isn't that TOO HIGH?

ISTR that the charge voltage shouldn't be more than ~14v.

And why not just run a string of Schottky diodes to drop the charging V to
the start battery?

 

[Rich Grise]

I need to charge the 12 volt engine start battery in my sailboat from
the 12 volt house battery in the boat. The house battery is the one
connected to the alternator. I can't just parallel the two batteries
for charging, since the house battery is a long cable run away from
the alternator, and when a heavy charge is going into the house
battery(50 amps or so), the voltage that the start battery would see
could be as high as 15.5 volts. I only need to supply about 3 amps
when the engine is running, just enough to run the fuel pump and
ignition.

Somebody at the boat store could probably give a better answer than anyone
here; _somebody_ with a boat must have had a similar experience.

(or try to find a boating or RV newsgroup.)

Good Luck!
Rich

 

[Chuck]

Somebody at the boat store could probably give a better answer than anyone
here; _somebody_ with a boat must have had a similar experience.

(or try to find a boating or RV newsgroup.)

Good Luck!
Rich

I'm a little confused, Larry.

First, the long cable run conveys no information. What is the wire
size? I assume you're not on an aircraft carrier so the run length is
maybe 20 feet round trip?

Restoring the energy from starting the engine would probably not be
affected by the cable run provided the wire size is not ridiculously
small (e.g., 18awg or smaller).

Providing the 3 amps for ignition and fuel pump could actually be
handled by 18awg wire.

It seems the only real issue is that 15.5 volts is likely to lead to
gassing in the engine start battery unless the temperature is really
cold, in which case you're in the wrong cruising waters.

But 15.5 volts is really on the high side for lead acid, AGM, or gel
batteries as I recall. You certainly should not see these batteries
drawing 50 amps with a terminal voltage of 15.5 volts. Are you seeing
that?

So my suggestion is use the proper regulator for your house battery
and it may solve your start battery charging problem. Ideally, both
batteries are the same type. If your house battery is an AGM or gel
type, then you will not damage your lead acid starting battery with
the setup you already have, provided the house battery regulator
limits charging voltage to whatever your AGM or gel battery requires.

Chuck

 

[Larry Bradley]

I always get caught up like this - I ask a specific question, but
don't provide enough background, so I end up sending peple off track.
Sorry about that.

The alternator, of course, is on the engine. The house battery is 25
milliohms of resistance away (32 feet round trip of wire, plus various
connections to switches, terminal posts, etc). This is all
unchangeable, unfortunately.

The alternator regulator is a 3-stage regulator, and senses the
battery voltage at the battery terminals, so the battery never gets
more than 14.4 volts (the nominal gassing voltage). But 50 amps and 25
milliohms results in about 15.6 volts at the alternator. Another
poster asked about using diodes to drop the voltage - fine when the
charge rate is 50 amps, but when it drops down to 5 amps and charging
at 13.5 volts, the start battery doesn't get charged at all.

The start battery is connected by heavy gauge wire (via a switch) to
the alternator. It starts the engine, of course, but also provides
power for the electric fuel pump and the electronic ignition - about 3
amps. The brief load of starting the engine doesn't discharge the
battery much, so there is no real need to charge the start battery
with a large current. Being able to provide 3 to 5 amps is plenty. And
I have a separate 115 volt charger for when I'm in the marina.

So the problem is "How to limit the voltage from the alternator, which
can be as high as 15.6 volts, and as low as 13.5 or so, to around 13.2
volts for the start battery?"

Three alternatives came to mind.
a) a linear regulator. No problem, I can easily do this. Built lots of
them before.
b) a switching regulator. Never done this before, but I know what is
involved. Getting and inductor might be a pain.
c) Just use pulse width modulation (a simple FET or BJT switch) to
drive the battery. The switch when on would put 15.6 volts to the
battery for a short time. But the average voltage would be 13.2 (or
whatever). A previous poster suggested that this could result in the
current during the on time being quite large.

There is a 4th alternative. You can buy a device that does exactly
this - it will charge the start battery from the house battery - I
presume it is just a DC-DC converter - maybe a SEPIC arangement to
enable it to handle input voltages both above and below the output
voltage. But they are expensive, and I like to brew up my own stuff at
any rate.

Thanks for the interest and ideas.

Larry

In my original post, I was just asking about the feasability of
alternative c).

I'm a little confused, Larry.

First, the long cable run conveys no information. What is the wire
size? I assume you're not on an aircraft carrier so the run length is
maybe 20 feet round trip?

Restoring the energy from starting the engine would probably not be
affected by the cable run provided the wire size is not ridiculously
small (e.g., 18awg or smaller).

Providing the 3 amps for ignition and fuel pump could actually be
handled by 18awg wire.

It seems the only real issue is that 15.5 volts is likely to lead to
gassing in the engine start battery unless the temperature is really
cold, in which case you're in the wrong cruising waters.

But 15.5 volts is really on the high side for lead acid, AGM, or gel
batteries as I recall. You certainly should not see these batteries
drawing 50 amps with a terminal voltage of 15.5 volts. Are you seeing
that?

So my suggestion is use the proper regulator for your house battery
and it may solve your start battery charging problem. Ideally, both
batteries are the same type. If your house battery is an AGM or gel
type, then you will not damage your lead acid starting battery with
the setup you already have, provided the house battery regulator
limits charging voltage to whatever your AGM or gel battery requires.

Chuck

Larry Bradley VE3CRX
Remove "removeme" from my e-mail address for direct mail
Ottawa, Canada

(use the e-mail address above to send directly to me)

 

[Chuck]

I always get caught up like this - I ask a specific question, but
don't provide enough background, so I end up sending peple off track.
Sorry about that.

The alternator, of course, is on the engine. The house battery is 25
milliohms of resistance away (32 feet round trip of wire, plus various
connections to switches, terminal posts, etc). This is all
unchangeable, unfortunately.

The alternator regulator is a 3-stage regulator, and senses the
battery voltage at the battery terminals, so the battery never gets
more than 14.4 volts (the nominal gassing voltage). But 50 amps and 25
milliohms results in about 15.6 volts at the alternator. Another
poster asked about using diodes to drop the voltage - fine when the
charge rate is 50 amps, but when it drops down to 5 amps and charging
at 13.5 volts, the start battery doesn't get charged at all.

The start battery is connected by heavy gauge wire (via a switch) to
the alternator. It starts the engine, of course, but also provides
power for the electric fuel pump and the electronic ignition - about 3
amps. The brief load of starting the engine doesn't discharge the
battery much, so there is no real need to charge the start battery
with a large current. Being able to provide 3 to 5 amps is plenty. And
I have a separate 115 volt charger for when I'm in the marina.

So the problem is "How to limit the voltage from the alternator, which
can be as high as 15.6 volts, and as low as 13.5 or so, to around 13.2
volts for the start battery?"

Ahh. Thanks for the elaboration, Larry. It all makes sense now.

You might use a zener in conjunction with an automotive relay (30A or
so), or for possibly greater efficiency, a zener with a power mosfet.
Or a zener, pot, mosfet and relay. With a handful of parts, virtually
no losses, and no switching hash, you can set the trip voltage so the
starting battery is connected to the alternator only when the
alternator voltage is below the level you set.

Once your house battery soaks up some charge, the voltage drop (across
the 25 mohms) will decrease and alternator output will be at an
acceptable level for the starting battery. I think I'd be surprised if
you were sustaining a 50A charge rate for much more than an hour.

If you decide to go with the pwm approach, you might look into the
TL494 chip driving a power mosfet.

Good luck

Chuck

 

[Rich Grise]

I always get caught up like this - I ask a specific question, but
don't provide enough background, so I end up sending peple off track.
Sorry about that.

The alternator, of course, is on the engine. The house battery is 25
milliohms of resistance away (32 feet round trip of wire, plus various
connections to switches, terminal posts, etc). This is all
unchangeable, unfortunately.

Is it physically feasible to clip a pair of #16 or #14's to the house
battery itself, (16 o4 14 because it has to go all the way back to the
motor) and let the start battery just float-charge? Lead-acid is
lead-acid, you know; they all have the same voltage, and when the battery
is fully charged, it will draw negligible current.

Have Fun!
Rich

 

[Larry Bradley]

Chuck:

I actually have somthing like that now.
A few years ago I built my own amp-hour meter using a PIC 18F452 - it
measures both house and start battery voltage and current.

It is programmed to actuate a solenoid relay to connect house and
start batteries together when the engine is running (it monitors the
ignition switch).

Last year I added a large diode in series with the relay contacts - it
drops the voltage by about .8 volts to the start battery. When the
house battery charging current gets down to a reasonable level, the
program bypasses the diode via a small relay to apply the full
voltage.

It's a kudge, so I thought the regulator trick might be "neater".

50 amp charge is about 1 hour, then tapers off.

Larry

Ahh. Thanks for the elaboration, Larry. It all makes sense now.

You might use a zener in conjunction with an automotive relay (30A or
so), or for possibly greater efficiency, a zener with a power mosfet.
Or a zener, pot, mosfet and relay. With a handful of parts, virtually
no losses, and no switching hash, you can set the trip voltage so the
starting battery is connected to the alternator only when the
alternator voltage is below the level you set.

Once your house battery soaks up some charge, the voltage drop (across
the 25 mohms) will decrease and alternator output will be at an
acceptable level for the starting battery. I think I'd be surprised if
you were sustaining a 50A charge rate for much more than an hour.

If you decide to go with the pwm approach, you might look into the
TL494 chip driving a power mosfet.

Good luck

Chuck

Larry Bradley VE3CRX
Remove "removeme" from my e-mail address for direct mail
Ottawa, Canada

(use the e-mail address above to send directly to me)

 

[Larry Bradley]

I'll have to take a closer look at that idea. I have room in the
conduit to run probably 12 gauge wire from the house to the start
battery for charging purposes. Good thought. Thanks

Is it physically feasible to clip a pair of #16 or #14's to the house
battery itself, (16 o4 14 because it has to go all the way back to the
motor) and let the start battery just float-charge? Lead-acid is
lead-acid, you know; they all have the same voltage, and when the battery
is fully charged, it will draw negligible current.

Have Fun!
Rich

Larry Bradley VE3CRX
Remove "removeme" from my e-mail address for direct mail
Ottawa, Canada

(use the e-mail address above to send directly to me)