Regulate Car voltage?

Hello!
I have a question regarding regulating the car's voltage to protect LEDs.
Voltage spikes beyond 12V (up to 14-15 volts) can happen upon ignition of the car and during driving.

I was wondering what is the simplest way to regulate the voltage to provide constant current to power (a 12 volt LED strip) as an example?

Can someone please suggest which parts to get and how to wire it in the circuit?
Cheers,
Michael

Cars generally run 13.8 volts. What if you went with LEDs made for RVs/campers? They should handle the voltages without any regulation.

Or this: http://www.radioshack.com/product/index.jsp?productId=2062600

Hey!
Thank you for your reply.

I had the ML7812 in mind. However, the input voltage required is usually a few volts above 12V. My voltmeter showed 12V from interior light connections. The LEDs typically flickered on ignition and some significantly deemed. Thats why i'm looking to construct some sort of a voltage spike protection mechanism.

My idea was to run the wires from the parking lights to the foglights and some parallel connections to the interior.

You start the conversation discussing "spikes" in voltage from a cars charging system, but you quickly switch to wanting to control current. In this case since you are talking about flexible led strips that run series/parallel circuits you want a voltage regulator. The strips will include current limiting resistors every ~3 leds or so (this may depend on color).

The simplest way as you ask is to purchase a premade 12V regulator circuit. There are some switch mode regulators from sureelectronics that I am partial to. You can find them on ebay or direct from the sellers website. They also carry some premounted linear voltage regulators. Next simplest would be to purchase a regulator IC and mount the components yourself. What you need depends on voltage headroom and current draw.

Something like the LM7812 you mention probably has too high a dropout voltage for 12V output from a car. It requires ~2V of headroom. A popular alternative with the car crowd is the Sharp PQ05RD21 it has a max .5V dropout at a full 2 amp output.
The LM1086 has a max dropout of 1.5V at 1.5A, so that is marginally acceptable. And, one of my favorites is the LM2941ct. .5V max at 1A

Hi, the simplist way is to buy a driver from taskled.com, as George already takes auto voltage spikes into consideration. BTW, the spikes can be MUCH larger than 15 volts - more like 30 - 50. There are snubbers for this purpose, and frankly a tendency to use linear voltage regulators ( LDOs) just for this reason.

Thanks Harry, I was hoping to learn and do it as a DIY project rather then buy ready parts thats why im asking the experts in the forums

@VegasF6, I suppose you are right, maybe my question is actually about current regulation.

Harry wouldn't you say the parts VegasF6 recommended such as LM2941ct act like "snubbers?". They take it higher voltage and regulate it not to go above 12VDC (provide constant voltage).

I'm not very familiar with these terms but i'm willing to learn and been doing some reading. My goal is this: feed a 12 volt connection from one of the interior lights and prevent voltage spikes (LODs). However, I want to keep it 12volts and prevent the voltage drop. Maybe 0.5 volts is the best I can do. Can you link me with a part and where to buy? Will this task be difficult as a DIY project? i'm fairly good with soldering, just missing the electronics background because i'm in healthcare.

P.s. I called a local electronics store and they told me I can put in a diode (1n4007) in series. Would that achieve what i'm looking for?
Thank you!

Several PC mounted DC-DC converters exist for 12V regulation. Most come in the form of switching supplies rather than linear due to the need for a sufficiently high overhead for normal operation.

For loads around 1A, I usually go for something like this: http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_1953161_-1
Vin 9-18V, Vout = 12V. Good ripple control, fully isolated, encapsulated and wirewrap ready. Its a little pricy but for their performance I have no complaints. Wiring is easy, 2 pin in 2 pin out.

erm, 1N400X rectifier diodes are not suitable for automotive surge suppressors. What might be considered is TVS or snubber diodes used for buffering line surges. What you may have to take note in addition to surges is harmonics and electrical noise. Depending on your application line filters may be necessary to ensure long life of your drivers.

If you are seeking for current regulation LED driver without the use of a two step process involving a DC-DC converter for power prior to the LED loop, I'd look into Lux-drive's powerpuck or buckpuck line of led drivers, which has a voltage input of 5-30V. The use of such devices will require special layout considerations to prevent blowouts such as this: http://www.candlepowerforums.com/vb/showthread.php?208756-Buckpuck-3023-dn-1000-Blew-up-burned-up

Most 12V LED fixtures really do not need any additional current regulation, at minimum few I've seen requires it. Many I've used takes advantage of the LEDs own foward voltage/resistance to regulate current. Their only prerequisite is the stability of the voltage supply.

A simple way to going about it, [assuming of course you are mounting this on a huge heatsink and that efficiency loss is not a big concern] is to use an 7812 LDO and a fat cap on the output end [10,000-30,000uF] the LM7812 package should give you at least 1A of foward current at 12V. A pass transistor such as 2N3055 or similar PNP power transistor may be used in conjunction for more current. Since the battery's operating voltage is around 13.8V. There is enough overhead voltage to ensure stable operation. huge caps on the output end not only diminishes ripple but also lengthens the time necessary for the output voltage to sag below what its rated voltage specifies.

Let me add the lm2940CT-12 by the way, another low dropout regulator. .5V max.

What you are describing is indeed a voltage regulator, not current control. This will step down a variable input voltage to a fixed output voltage. For the above IC to work it will require at least 12.5V in, which you should have at pretty much all times, even when not running, with a healthy battery. Voltage spikes from the charging system or from capacitive loads like power window motors, door lock solenoids etc are handled by a capacitor across the input for the regulator. This is your snubber. Without it your voltage regulator could be damaged. These are called transients.

A diode could be used for wiring one of the capacitive loads I mentioned, that prevents spikes from that equipment when shutting it off damaging other sensitive equipment. It also causes a forward voltage drop however, not something want or need in this case.

Perhaps if you told us more about your load, but the led strips I am thinking of are only a few watts. In this case a linear regulator is more than good enough. The efficiency losses due to heat will be minimal and you will just need a small to-220 heatsink for a buck and a half, a scrap piece of copper or aluminum, or possibly even no heatsink at all. The Sharp regulator for instance is capable of dissipating 1.4 watts on it own (depending on temperature.) Probably under most circumstances your cars charging system won't go above 13.6V and you may experience some voltage loss through small wiring to for instance a dome light or tail light.
The equation for power dissipation in a linear regulator is (Vin - Vout) * Iout
Since we know the voltage differential is only about 1.6V and we know the regulator can handle 1.4 watts, that tells us we could have up to about .875 amps out and probably not even need a heatsink at all. Just for safetys sake lets just call it .5 amps. That's still a LOT of leds when talking about low power led strips that maybe only use .050 amps for every 3!

The resistor is your (imperfect) current limiting device in that situation, but it works best with a fixed voltage, that's why you want voltage regulation. You can not use a current regulator designed for power leds like the ones from Taskled or any other constant current regulator in this case.

And yes the led strips I have seen protected with only a resistor do indeed require a stable voltage to run within spec.

Also, just FYI, almost all interior lights are electronically controlled (that is how they get them to stay on a few seconds after the door is closed), so reading voltage off that may not be accurate.

I used a sharp PQ12RD21. I think they are discontinued, though.

HarryN said:

Hi, the simplist way is to buy a driver from taskled.com, as George already takes auto voltage spikes into consideration. BTW, the spikes can be MUCH larger than 15 volts - more like 30 - 50. There are snubbers for this purpose, and frankly a tendency to use linear voltage regulators ( LDOs) just for this reason.

Click to expand...

Indeed, the spikes can be much larger than 12v, but even 50v is too conservative. There are also negative voltage spikes, which is a reason to put a series rectifier diode upstream of the voltage regulator.

If you have access to a good technical library, look up SAE J1113/11. this is written by the Society of Automotive Engineers, and is a standard test for battery powered systems and defines the standard pulses that the device will be tested to.

Anyway, I'd recommend protecting against brief pulses that could be up to 150v, both positive and negative. There are some long duration pulses that are lower voltage but pack more energy. I would suggest putting a 0.1uF, 100v cap and a 18v transorb (maybe a SMB package) right at the point where battery power enters the device. Follow that with the series rectifier diode (to block negative transients), and then maybe another 0.1uF cap. After this, install your regulator, whether it's a voltage or current regulator.

by the way, "LDO" stands for "low dropout regulator", and describes a regulator that requires very little voltage differential from input to output. A typical linear regulator like the LM317 needs about 2.5v, where LDO's can get by with a few tenths of a volt.

regards,

Steve K.

Good info from steve K there, and IIRC isnt there provisions for a 24v jump start in the standards, something like 24v for 5 mins to represent a using a tow trucks 24v system to quickly rush some charge into a dead battery an get the system started? Obviously some equipment can just go into shut down (radio, some gauges, etc) but critical stuff for getting a vehicle started should be aple to run with the voltage. Sounds like a dangerous practice (abusing the 12v battery in such a manner) and possibly antequated now, but I thought I had read something like that in the past.

I've heard of using welders to jump start large machines with 24v battery systems. Not a great thing, so I wouldn't be shocked to hear of a 24v battery vehicle jumping a 12v vehicle.

Along that line of thinking, there is a reason that a lot of switching power supply IC's are being designed to operate from voltages up to 80v or so, and that's for automotive applications. They do make it easier to reduce power dissipation and still operate over the 9v to 18v that might exist on the battery bus, as well as some of the long transient pulses.

I wish I knew of a small protection module that non-EE's could just add to their design. Until then, I'd use my earlier suggestion, and try to use a big TVS (transorb, or transient voltage suppression diode). Might want to include a fuse too, just in case. The parts don't cost all that much, and are good insurance for $20 (or more) of LEDs and current regulator circuitry.

Steve K.

mikey99 said:

P.s. I called a local electronics store and they told me I can put in a diode (1n4007) in series. Would that achieve what i'm looking for?

Click to expand...

Oh dear, the dangers of listening to sales people.

A diode will reduce the voltage by 0.7 volts. 12 volts in will give 11.3 volts out. 22 volts in will give 21.3 volts out ! ! !

VegasF6 said:

. . . . And yes the led strips I have seen protected with only a resistor do indeed require a stable voltage to run within spec.

Click to expand...

I have retrofitted LED Parking Lamps and interior lights that simply use dropping resistors - years later they are still working.

You can often get away with over driving LEDs. It doesn't mean they are running within spec. Short duty cycle will help, but by doing so you are reducing the life cycle and affecting the luminous output.

assuming you have enough headroom and arent using 0.5ohm resistors or some ridiculously small value like that, dropping resistors should be fine for low-power LEDs. Just remember to use a resistor for EVERY series string. If your calculations tell you that you need a very small resistance to limit current, then you might want to go ahead and add a voltage regulator on the front end.

Aren't at least some (if not all) car's circuits already regulated? It seems to me that the radio for example, must have a fairly stable voltage to work properly. And what about the GPS and all those electronic gadgets? There must be some kind of power conditioning in there... I'm just wondering.

It would be nice if cars offered a nice clean 12v bus for solid state devices to operate from, but I haven't heard of this happening. Part of the problem is just the basic issue of adding and removing loads from the battery bus. For instance, when you turn your headlights off, the alternator suddenly loses a significant load. The regulator circuit in the alternator takes time to respond (there's a lot of energy stored in the alternator windings that has to go somewhere), and the battery can only absorb so much so fast. This can produce a voltage spike of many ten's of volts that can damage unprotected circuits.

Another source of nasty transients is simply turning off the blower fan. Assuming that it is a brushed DC motor, when it slowly spins to a stop, it is actually acting as a generator and may be producing some large voltage spikes.

The horn can be a source of voltage spikes too. The normal construction of a horn is a coil attached to the diaphram. The diaphram has switch contacts attached, so that when the diaphram has moved far enough, the current to the coil is interrupted, the coil de-energizes, and the diaphram moves back (which closes the switch contacts and starts the cycle over again). When the current to the coil is interrupted, the coil can produce large voltage spikes. A good design could snub out a lot of the voltage spikes, but that adds cost to what is considered a commodity product, so the horns tend to be a problem.

There's also the issue of needing to design electronics to survive some of the more usual failures on a car. For instance, if the alternator's regulator fails and allows the battery voltage to increase, it'll eventually kill the battery. Hopefully you'll see the "battery" warning light on the dash before this happens. Most electronics are designed to survive this high voltage condition, because you don't want to have to replace your ECU, your radio, your GPS, etc., just because the alternator failed.

Maybe this is a business opportunity for someone?? Design a good high current voltage regulator and see if there is a market for it? Be sure that you've reviewed the standard SAE (or EU style ISO) tests and can pass all of them.

regards,
Steve K.

Calina said:

Aren't at least some (if not all) car's circuits already regulated? It seems to me that the radio for example, must have a fairly stable voltage to work properly. And what about the GPS and all those electronic gadgets? There must be some kind of power conditioning in there... I'm just wondering.

Click to expand...

Any electronic device you buy, for home or car has a power supply section built in. Unless the power supply is external, like in a laptop.