What is the power of the brake lights in cars?

[treez]

What is the power level in Watts of individual LED brake lights in cars, such as the Bentley Continental....

https://www.bentleymotors.com/models/continental_gt/

..if power not known for this , then please advise on power for any car that you know.

 

[godfreyl]

The required brightness is given here (in candela).
Maybe you can work it out from that?

 

[betwixt]

Normal incandescent UK brake lights are 12V/21W each.

Brian.

 

[chuckey]

Incandescent brake lights at 21 W each would give about 4W for LEDs.
Frank

 

[treez]

Thanks, it always amazes me how low power brake lights are.....I mean, with the brakes on, the rear lights are three times as powerful as with just the rear lights on...so the rear lights must be just 1.3W in LED terms?
I wonder why led brake lights are always done with linear regulators?....I mean, I understand that they have to come on quickly, but a switch mode driver could surely manage this?........I hear one reason is that they can sometimes flash on/off/on in quick succession, and so a switcher, with its LC filter, could give rise to damaging oscillations, wheras linear regulators don't really need LC filters.

 

[Mattylad]

Simply pick your car and look up replacement led bulbs.

You will soon get your answer.

 

[treez]

ok thanks, but im pretty sure that often, led rear lamp bulb replacements don't exist......you have to take the car into the garage to get them replaced...because it involves removing the entire heatsink assembly....specially in something like a Bentley.

I think on some cheap cars you can get the blinkers as replacement led bulbs, but the rearlamp/stop & tail lights tend to be in a heatsink assembly that is more complex.

Ive certainly seen the brake/rearlights for a car like a Bentley, it was a big pcb with about 20 linear current regulators on it, and some 80 low power leds....it wasn't something that could be replaced quickly....

 

[Mattylad]

Nah!

Most cars have them available as direct plug in replacements.

Do you drive a Bentley? posh sod !

 

[treez]

The below is the rear lamp PCB (brake and rear) for an executive car. It has 4.95W worth of led power...and 4.29W of losses in the linear regulators when car battery is 14V.......if you ask me, 4.29W of losses is too much, do you agree?

The rear/brake is the bottom circuit on the page...I did two pics as it was on A3 and I couldn't fit it all on one pic. (as you see, its 12 strings of 3-red-leds-in-series with 55mA in each string)

Also, shown is the whole circuit together, ..'stitched' together "rearlamp total"

 

[Dan Mills]

50% losses sounds bad, but actually when you consider that a brake light is only on for maybe 5% of the time, your 4W of losses turn into 200mW average which is completely irrelevant.

I actually quite like that circuit, it has several interesting features that may not be immediately obvious:

It has significant fault tolerence, a single string failure will not stop the brake light functioning.
It will work at full brightness with a severly flat battery.
It has fault reporting, a string failing will result in the appropriate 'monitor' line dropping and that can be used to signal a fault to the ECU.
It has effectively zero standby power consumption (A big deal if you leave your car in a airport long stay for a few weeks and expect it to start).
It is resistant to transients and even has modest resistance to reverse polarity, getting here with a switching regulator is not trivial.
It should be easy to get this through EMC qualification, certainally compared to a switcher on a single sided board (That this probably is).
It is cheap, cheap, cheap, there is nothing expensive on that board.
Component obscelecence should be a non issue fo a long time, with the LEDs being about the only thing that stands any risk of needing to have a replacement part qualified.

I have seen MUCH worse in superficially more sophisticated designs.

Regards, Dan.

 

[treez]

ok thanks, but do you think the LMV431 has been connected in wrong?....I mean....the reference pin of the LMV431 surely should be connected to the cathode of the LMV431?

 

[Dan Mills]

Nope, the LMV431 servos the bias point of the transistors to maintain 1.24V across the emitter resistor, see the current sink application on page 15 of the TI datasheet.

Possibly not quite what I would have done, but it will work just fine.

Regards, Dan.

 

[treez]

sorry I didn't tell you that all those bjts are not very close to each other, and are just same part number bjts, not particularly matched, other than being same part number.
I would be unhappy that all those separate bjts were not adequately matched.......they were spaced fairly far apart and certainly wont be well thermally coupled.

 

[betwixt]

Agreed, it's working on the principle that constant voltage across the resistor = constant current through it.

Nrian.

 

[treez]

Yes I see your point, though as you know, we don't know that the vbe's of all those bjts is going to be the same?....BeTwixt I think you may have answered before my #13 post came out.

To make firmer the point about how far the bjts are apart.....when this circuit was first done, there were none of those base-emitter caps in....and the linear regulators each oscillated at about 20MHz.....this of course resulted in an average led current of around 19mA, instead of the required 55mA.

 

[Dan Mills]

But the vbe drop is only mildy temperature dependant (~-2mV/C), that corresponds to about 1 part in 620 per degree C given that the voltage across the resistors is somewhere around 1.24V , and LED current is just not that critical.

As long as the LED current is within prehaps 10% of target it will all be fine, and this is quite good enough (I would probably have just used a few foward biased diodes in place of the '431 and it would still have been more then good enough).

Regards, Dan.

 

[betwixt]

When you say 'linear regulators each oscillated", which regulators are you referring to? There is only one with any gain in it's feedback circuit and that's the '431. Adding individual capacitors across each B-E junction isn't a bad thing because it reduces immunity to RF pick-up but the circuit shouldn't inherently oscillate.

Brian.

 

[Dan Mills]

Actually it can happen, particularly if there is a bit of inductance in the emitter leads, you can end up with a kind of hartley circuit built in the parasitics....
Been there, done that, got the EMC fail to prove it (Mine was a microprocessor reset circuit, embarassing and tricky to find (it vanished when you applied a scope probe)).

Regards, Dan.

 

[treez]

It has effectively zero standby power consumption (A big deal if you leave your car in a airport long stay for a few weeks and expect it to start).

sorry but I don't understand this point, there are no standby losses in any car lights when they are off, because when they are off, their relay is off and no current can flow to the light?

 

[Audioguru]

My Chevrolet Malibu uses 3157 (1157 with a different base) incandescent rear light bulbs that has 7W for the rear light and 27W for the brake and turn signal light. Someone posted an LED equivalent that has 99 LEDs and it is not as bright as the original incandescent bulb, maybe because the "white" LEDs lack much red light.