Adjustable voltage regulator for car alternator

[marcus]

Has anyone come across a circuit for a manually-adjustable, or even
automatic regulator circuit
to control a 24V car alternator?
Application is deep-cycle solar battery charging using 3-5HP small petrol
motor coupled to 24V alternator.
Need to tweak the voltage up to about 29-30V to equalise the batteries.
Solar batteries like a 3-stage charging regime - boost, bulk & float. These
voltages are all diff. and also depend on battery type (chemistry), so the
field needs to have these adjustable set-points.
Would be nice to be able to sense battery voltage and feedback into
alternator field voltage, to make a set-and-forget circuit.

Here's some background (12V, and no feedback) :
http://www.homepower.com/files/mark8.pdf

Thanks to everyone in aus.electronics who had a look at this.

Jim Thompson : I notice from your web pages that you have designed some car
alternator regulation projects - wondering if you have any thoughts?

Thanks all

Marcus in outback Oz

 

[Jim Thompson]

Has anyone come across a circuit for a manually-adjustable, or even
automatic regulator circuit
to control a 24V car alternator?
Application is deep-cycle solar battery charging using 3-5HP small petrol
motor coupled to 24V alternator.
Need to tweak the voltage up to about 29-30V to equalise the batteries.
Solar batteries like a 3-stage charging regime - boost, bulk & float. These
voltages are all diff. and also depend on battery type (chemistry), so the
field needs to have these adjustable set-points.
Would be nice to be able to sense battery voltage and feedback into
alternator field voltage, to make a set-and-forget circuit.

Here's some background (12V, and no feedback) :
http://www.homepower.com/files/mark8.pdf

Thanks to everyone in aus.electronics who had a look at this.

Jim Thompson : I notice from your web pages that you have designed some car
alternator regulation projects - wondering if you have any thoughts?

Thanks all

Marcus in outback Oz

Of course. (Most of my early patents were automotive products.)

Early next week... I'm buried with work this weekend.

Flat voltage or do you need temperature compensation?

...Jim Thompson

 

[Jim Thompson]

Of course. (Most of my early patents were automotive products.)

Early next week... I'm buried with work this weekend.

Flat voltage or do you need temperature compensation?

...Jim Thompson

Also, Is field tied to A+ end or ground end of alternator?

...Jim Thompson

 

[default]

Has anyone come across a circuit for a manually-adjustable, or even
automatic regulator circuit
to control a 24V car alternator?
Application is deep-cycle solar battery charging using 3-5HP small petrol
motor coupled to 24V alternator.
Need to tweak the voltage up to about 29-30V to equalise the batteries.
Solar batteries like a 3-stage charging regime - boost, bulk & float. These
voltages are all diff. and also depend on battery type (chemistry), so the
field needs to have these adjustable set-points.
Would be nice to be able to sense battery voltage and feedback into
alternator field voltage, to make a set-and-forget circuit.

Here's some background (12V, and no feedback) :
http://www.homepower.com/files/mark8.pdf

Thanks to everyone in aus.electronics who had a look at this.

Jim Thompson : I notice from your web pages that you have designed some car
alternator regulation projects - wondering if you have any thoughts?

Thanks all

Marcus in outback Oz

I couldn't see the file you reference.
but
12 volt alternators can put out 14 volts at idle so with just a pulley
adjustment or running them faster the voltage goes up proportionately.
From that point on it is just a matter of a regulator.

I have a two transistor regulator that I scrounged from a Chilton's
manual back in '72. They didn't show component values, but I put what
I thought would work and it did in a BMW motorcycle and Toyota Land
Cruiser. Doesn't do anything fancy like float at a lower voltage, but
I could post the schematic if you want or email it.

Very simple device - NPN pass transistor is biased "on" with a
resistor (turning on the rotor) and a second transistor turns it off
when the Zener/ and potentiometer-setting voltage is exceeded. The
Chilton's manual showed fixed resistors - I used a military spec wire
wound pot to set the voltage.

The BMW reg was in the bike for 10 years (in the weather) and I had no
problems with it other than painting the transistors so the TO66 cases
wouldn't rust through - on the land cruiser I used a 2N3055 and it
never rusted in the engine compartment. No heatsink was needed - the
rotor would pull about 3 amps maximum.

 

[RW Salnick]

marcus inscribed in red ink for all to know:

Has anyone come across a circuit for a manually-adjustable, or even
automatic regulator circuit
to control a 24V car alternator?
Application is deep-cycle solar battery charging using 3-5HP small petrol
motor coupled to 24V alternator.
Need to tweak the voltage up to about 29-30V to equalise the batteries.
Solar batteries like a 3-stage charging regime - boost, bulk & float. These
voltages are all diff. and also depend on battery type (chemistry), so the
field needs to have these adjustable set-points.
Would be nice to be able to sense battery voltage and feedback into
alternator field voltage, to make a set-and-forget circuit.

Here's some background (12V, and no feedback) :
http://www.homepower.com/files/mark8.pdf

Thanks to everyone in aus.electronics who had a look at this.

Jim Thompson : I notice from your web pages that you have designed some car
alternator regulation projects - wondering if you have any thoughts?

Thanks all

Marcus in outback Oz

You can buy one for less than $20.

Are you talking about converting a 12V nominal automotive alternator to
24V? You won't need to make any changes to the pulley to make the
alternator deliver 24 V, but the the diodes and the diode trio will need
to be replaced with 24V versions.

And the new regulator, of course.

bob
s/v Eolian
Seattle

 

[John Popelish]

RW Salnick wrote:
(snip) You won't need to make any changes to the pulley to
make the

alternator deliver 24 V, but the the diodes and the diode trio will need
to be replaced with 24V versions.

(snip)

Are you seriously suggesting that a 12 volt alternator is
made with a diode trio that cannot handle 24 volts? I was
under the impression that few silicon rectifiers are made
with a breakdown voltage less than 50 volts.

 

[Carl Ijames]

Has anyone come across a circuit for a manually-adjustable, or even
The crude way is to use a voltage divider on the sense input. For the
GM 12 V 120 A alternators used on the 86-87 turbo V6 Regals, an example
diagram is at: http://www.gnttype.org/techarea/projects/gnalt.html The
idea is to use a throttle or boost actuated switch to kick up the output
voltage to give the ignition and fuel pump a boost at wide open
throttle. On these alternators the dash volts light is a 194 and it
goes from battery to the L pin on the 4 pin connector, to provide both
sensing input and idiot light. The internal resistance of this bulb is
crucial, you will blow things up if it is shorted and make things very
unhappy if it is open. The S terminal, on the other hand, is for
sensing only and can be connected directly to the battery. I don't know
the limits on output voltage but 16 V is what is commonly used in
racing - enough to make a real difference in ignition and fueling but
not quite start popping lightbulbs during a sub-12 second quarter mile
blast . This works on all the GM alternators I've seen from the 70's
and 80's. Somewhere in the 90's the engine computer started controlling
the alternator and I'm not familiar with the details on those. A
commercial external regulator (I think; it may just be a divider too) is
available from Precision Turbo & Engine in Indiana, USA.

 

[RW Salnick]

John Popelish inscribed in red ink for all to know:

RW Salnick wrote:
(snip) You won't need to make any changes to the pulley to make the


(snip)

Are you seriously suggesting that a 12 volt alternator is made with a
diode trio that cannot handle 24 volts? I was under the impression that
few silicon rectifiers are made with a breakdown voltage less than 50
volts.

They need to be able to handle *at least* double the intended voltage
delivery. A 24 V alternator will probably actually deliver 28V (just as
today's 12 V alternators are delivering as much as 16V).

bob

 

[John Popelish]

What PIV diodes do you think you need for 24 volts?

Need? 35 volts should be enough to work, but I would like
at least 100 volt.

What PIV diodes do you think they put in the trio?

 

[John Popelish]

You are forgetting that when the diode is off, it also has the
battery voltage added to the AC from the alternator, so the bare
minimum would be 24 VDC +(24*1.414=33.936-.6) 33.336, or 57.336 PIV, if
there are no spikes or inductive surges.

I'm not forgetting anything, that is not how it works.
there is a 3 phase bridge rectifier across the AC with the
negative side of the output to ground (so no AC swings can
go more negative than a diode drop below ground). There are
two positive sides to this bridge, connected in parallel to
the 3 phase AC. One set of 3 provides the DC output to the
battery, and the other 3 diodes provide the DC to the field
excitation regulator. But, since the 3 to the battery
prevent the AC from exceeding the battery voltage by more
than a diode drop, None of these 9 diodes sees more reverse
voltage than the battery plus a diode drop.

A load dump can be several
hundred volts, and the diodes have to survive.

This, I agree with, and this is why the main rectifiers as
well as the trio diodes all have PIV ratings in the hundreds
of volts.

What PIV diodes do you think they put in the trio?


The last set I replaced were unmarked. All they had was an OEM part
number. One of the trios opened on the way home from Orlando after
midnight one night. I was laying in wet grass and sand removing the high
output alternator from my stepvan, only to find it was too big to fit my
car. I had to strip both alternators and transplant the diodes to get to
work on time.

What does this failure tell us about the PIV rating of the
trio? Not much, I think. You don't know that thew trio
failed from excessive voltage. Could have been a bond
failure that overheated one of them.

 

[John Popelish]

All three diodes were open.

So what do you conclude caused this failure, if anything?

 

[John Popelish]

I have no idea. All I found was four lugs and burnt plastic ash.
There was nothing left to do failure analysis on.

I would suspect an intermittent short between the rectifier
output and ground. But, I guess, if a high enough voltage
transient occurred, it could have shorted a rectifier and
then that short blew the other two. Of a diode bond failed
enough to overheat one die, causing the same cascade.

But do you agree with my analysis of the normal voltage
applied to these diodes? You clipped it without comment.

 

[marcus]

Thanks to everyone who's contributed ideas to this thread.

It's my understanding of alternator operation that the voltage is
"built-in" by the number of poles, the number of turns on the rotor/stator,
and least of all by the regulator.
The speed controls current, not voltage.

If the regulator is removed (or the reg.'s ability to reduce the field
current is removed), then the alternator voltage will rise to over 100 V.
(OK for US -
you could get mains power straight from output - either AC or DC, at a high
frequency,
but I need 240VAC in Oz.) I believe a large alternator [200A] might produce
around 7200 Watts when driven this way. It would probably need 2 belts and
12 HP to drive it at this level, assuming bearings, heat, diodes, etc. were
upgraded.

What I want to do is charge a bank of solar batteries at up to a C20 rate
(about 40A in my case), with a 24V alternator driven by a 3- 6HP stationary
engine, with an adjustable regulator circuit that will provide up to 30VDC
and taper the charge as the batteries fill.

Great booklet about alternators :
www.1stconnect.com/anozira/SiteTops/energy/Alternator/alternator.htm

cheers all,

Marcus in Oz @ 28.4VDC (Sunny day!)

 

[default]

Thanks to everyone who's contributed ideas to this thread.

It's my understanding of alternator operation that the voltage is
"built-in" by the number of poles, the number of turns on the rotor/stator,
and least of all by the regulator.
The speed controls current, not voltage.

Speed controls voltage, not current - wire size determines current. I
think it was Homer that suggested you try to regulate engine speed for
maximum efficiency from a gas engine.

If the regulator is removed (or the reg.'s ability to reduce the field
current is removed), then the alternator voltage will rise to over 100 V.
(OK for US -
you could get mains power straight from output - either AC or DC, at a high
frequency,
but I need 240VAC in Oz.) I believe a large alternator [200A] might produce
around 7200 Watts when driven this way. It would probably need 2 belts and
12 HP to drive it at this level, assuming bearings, heat, diodes, etc. were
upgraded.

What I want to do is charge a bank of solar batteries at up to a C20 rate
(about 40A in my case), with a 24V alternator driven by a 3- 6HP stationary
engine, with an adjustable regulator circuit that will provide up to 30VDC
and taper the charge as the batteries fill.

Great booklet about alternators :
www.1stconnect.com/anozira/SiteTops/energy/Alternator/alternator.htm

cheers all,

Marcus in Oz @ 28.4VDC (Sunny day!)

 

[default]

On these alternators the dash volts light is a 194 and it
goes from battery to the L pin on the 4 pin connector, to provide both
sensing input and idiot light. The internal resistance of this bulb is
crucial, you will blow things up if it is shorted and make things very
unhappy if it is open.

The idiot light supplies the current to magnetize the field until the
alternator is producing power. The light goes from the positive of
the ignition circuit to the field. There's an additional set of three
positive diodes on the alternator that also go to the field - when the
alternator is producing its own power, the light bulb has positive
voltage on both terminals and doesn't light.

Good design to use the alternator itself to supply the field through
isolation diodes - a shorted rotor won't output enough current to
destroy the regulator that way.

Short the lamp and you may destroy the regulator or rotor or both -
too high a lamp resistance and it will change the point where the
alternator begins charging or keep it from charging.

The lamp serves two purposes - tells you the alternator is working and
supplies current to the field to start the alternator working.

 

[Franc Zabkar]

What I want to do is charge a bank of solar batteries at up to a C20 rate
(about 40A in my case), with a 24V alternator driven by a 3- 6HP stationary
engine, with an adjustable regulator circuit that will provide up to 30VDC
and taper the charge as the batteries fill.

Great booklet about alternators :
www.1stconnect.com/anozira/SiteTops/energy/Alternator/alternator.htm

cheers all,

Marcus in Oz @ 28.4VDC (Sunny day!)

Can you relax your requirement to just a current limited 40A, 28.4VDC
constant voltage regulator? If so, then the required circuit would be
extremely simple, ie a 40A 600mV current shunt, 3 or 4 transistors, a
few diodes and a few resistors. The current limiting circuit (one
shunt and one transistor) would ensure that the current never exceeded
40A, even at low states of charge.

Temperature compensation could also be provided by a string of 1N4148
diodes.

If you are happy with this, then I will draw up a circuit for you.

- Franc Zabkar

 

[marcus]

Frank ~
Thanks for your offer of a circuit - can I take you up on that?

thanks

Marcus, in the rain!! (first time in about 3 months....)

 

[Franc Zabkar]

Frank ~
Thanks for your offer of a circuit - can I take you up on that?

thanks

Marcus, in the rain!! (first time in about 3 months....)

This is the basic idea:
http://www.users.on.net/~fzabkar/alt-reg.JPG

If you have a grounded field alternator, then the field section of the
circuit needs to be modified. You also need to know the battery's
temperature coefficient before you can select your string of diodes.
Silicon diodes have a tempco of -2mV/degC, but zeners can have either
positive or negative tempcos that depend on the zener voltage. IIRC,
zener tempco is zero at around 5 or 6V. An alternative place to put
the signal diodes would be in the voltage divider.

I have added a momentary start switch just in case the residual
magnetism in the rotor is insufficient to self excite the alternator.
I have avoided using the "charge" lamp that normally provides this
function because the batteries would slowly discharge through it if
the motor were to stop.

One other thing you might like to consider is an overvoltage monitor
that trips a contactor if the regulator fails.

- Franc Zabkar