# automotive 12v regulator/stabilizer

#### forced-air

Joined Jun 10, 2009
4
hi i need a diagram for a cicuit to keep a car 12v supply constantly at 12v as accuratly as is possible. The input voltage 11.8 -15.8vdc is way to wide to give accurate engine managment opperation. The particular circuit i need to supply has a 10 amp fuse. im guessing this would also be the rated max throughput for the circuit. ive seen various 12vdc stabilizers on the market but they appear to be basically caps that take spikes out which isnt really what im after. I need this to be a constant 12vdc output voltage regardless of input from say 12vdc - 18vdc. any help would be greatly appreciated. please no maybe replies i need a definitive answer thanks in advance

#### Externet

Joined Nov 29, 2005
1,493
That would be a wrong thing to do.
The battery fed circuitry is nominal 12.6V Varying from ~10V at start to 14V charging. Outside that range, the alternator has problems. Fix that instead.
The engine management compfuser does have its own internal regulation, usually at 5 Volt to operate itself and all sensors.
Leave it, unless sensors do not get stable 5V. And if that happens, the compfuser regulation is faulty, not the "12"V system
Miguel.

#### AlainB

Joined Apr 12, 2009
39
Depending if it is for a permanent usage, maybe you could use an inverter and plug in it a computer power supply. You will have then your well regulated 12 volts.

Alain

#### forced-air

Joined Jun 10, 2009
4
what you say would be fine if it were true except none of it is that simple. for a start im getting swings from 11.8 - 15.8 vdc and whilst the ecu does have a regulator its not expecting to see swings that wide and the internal regulator is outputing a varied voltage, surficing to say this swing is effecting the performance . ie it runs much better when the lights heater and other ancilleries are switched on when the voltage in the system is lower . ive looked in the why this is and it appears that heavey duty batteries and aftermarket regulators for the alternator more often than not put out more than the oem parts. since alot of these oem parts are no longer available and the desire to use heavy duty batteries leads me to to want to regulate the supply. its not a question of what you think i should be doing but a question of how to do it

Last edited:

#### forced-air

Joined Jun 10, 2009
4
thanks alainb but i need 10 amps something like that would be a nice way to prove the theory if it wasnt for the amps issue but ive already done that by wiring the circuit to an indendent battery thats not on the charging circuit . i need a permanent solution but thanks anyway

#### forced-air

Joined Jun 10, 2009
4
Commentary removed by moderator

Last edited by a moderator:

#### Steve C

Joined Nov 29, 2008
88
oh I would LOVE to have something like this. I've got three processor power supplies @ 3 amps max each (6 amps will easily run all three) but two are REALLY sensitive to voltage. Given an input below 11 volts, even as low as 10 volts, I would LOVE to be able to provide a stable 13 or 14 volts to them. Including dropout protection from dips when things like lights are turned on, the auto up windows hit their peak position, etc.

I've found nothing like that. Was thinking maybe a 15 volt DC-DC converter that runs on nominal 12V in with an adjustable regulator to bring that down to say 13.5, but I've not done research yet to see if that regulator exists. And I know a 6 amp 12V DC-DC costs $100 from digikey or mouser or jameco and don't even want to think about solutions in the$100 price range.

On the other hand, though the computer PSU is a nice free solution assuming parts are laying around (and oh they are!) my solution ideally wouldn't be larger than a wall wart. computer power supplies are huge! And contain huge amounts of extra unneeded stuff. Not a slick solution!

I'd really like my solution to be thinner than a pack of cigarettes, and a footprint not much larger than the two larger dimensions of the pack.

And less than \$100!

#### David Bridgen

Joined Feb 10, 2005
278
Voltage regulators need a supply whose minimum value is about 3V higher than their output.

You can't expect to regulate a source which varies down to 12V, never mind a bit below that, to 12V.

There is no easy/simple way of achieving what you want.

#### Steve C

Joined Nov 29, 2008
88
I noticed that, figured I'd have to step the voltage up to 15 before regulating it.

Not cheap

Are there other ways to step the voltage up cheaply? I think telecom operates at 48 volts, doesn't it? My local surplus store has TONS of brick DC-DC converters that take 48 volts, but nothing with a 12 volt input. So Surely there must be other ways to get our starting voltage up higher than 12 so we can regulate, yes?

Seems inefficient and stupid, but perhaps an inverter with wall wart? This would be cheap, small, and if not noisy would work, but needs an idle current draw of less than 500 mA, and preferrably WAY less.

no good? I'm not sure what the idle draws are for inverters and wall warts.

#### David Bridgen

Joined Feb 10, 2005
278
An interesting chip, Bertus.

However, in their Typical Application diagram and graph they show Iout/load to be 5A, yet in their table of Absolute Maximum Ratings they show a peak output current, for less than 10 microseconds, of 3A.

Lower down, they have a graph of load regulation which shows current up to 5A.

Most confusing.

#### GS3

Joined Sep 21, 2007
408
Voltage regulators need a supply whose minimum value is about 3V higher than their output.

You can't expect to regulate a source which varies down to 12V, never mind a bit below that, to 12V.

There is no easy/simple way of achieving what you want.
The best way to have a stable DC output with an input wich swing above and below the output is to use a SEPIC (switching) topology which will do just that.

#### David Bridgen

Joined Feb 10, 2005
278
The best way to have a stable DC output with an input wich swing above and below the output is to use a SEPIC (switching) topology which will do just that.
I don't doubt it. But can it can be called simple or easy?

#### DC_Kid

Joined Feb 25, 2008
796
if the alternator runs at 14+ volts then i guess you could place a few LM2737's in parallel with some big heatsinks to regulate the voltage to ~12v. one LM2737 can handle 25 amps. but like others said, these need a few volts higher than output to regulate.

but i think we need this question answered: why does the voltage drop to 11.8 while engine is running?? thats at or below a good battery.

Last edited:

#### SgtWookie

Joined Jul 17, 2007
22,210
An interesting chip, Bertus.

However, in their Typical Application diagram and graph they show Iout/load to be 5A, yet in their table of Absolute Maximum Ratings they show a peak output current, for less than 10 microseconds, of 3A.

Lower down, they have a graph of load regulation which shows current up to 5A.

Most confusing.
The output current is being handled by the external power MOSFETs aka "output switches".
The IC itself sources/sinks current to/from the MOSFET gates.