Effectively defeating PWM on Jeep Wrangler

[Beckerton]

Hello all,

This is my first post here and I have to admit I know very little about electronics. I am a tinkerer at heart and a very quick learner though.

Now, on to my situation. I own a 2015 Jeep Wrangler that has horrendous halogen headlights. I recently switched them for LED projector headlights that work wonderfully except that they flicker. The Jeep uses PWM for the halogen headlights to extend the life of the bulbs. The PWM fluctuates between 8 volts and 12 volts from what I have read. There are "anti-flicker" solutions that use large capacitors to overcome the flickering but there is a noticeable delay when going from low beam to high beam.

Would it be possible to use a boost converter to set output voltage to 12 volts, regardless of the Jeep is outputting less than 12 volts at any given time?

 

[BobK]

Easier would be to simply bypass the PWM module that is driving the headlights. It must have 12V going in and 8-12V going out. Just detach the module and wire the in to the out.

Bob

 

[Beckerton]

I believe the PWM is built into the PCM of the vehicle. There is another harness out there that hooks right to the battery for 12 volts and triggers relays for the high and low beams. Problem with that is that I am in Canada and my vehicle is equipped with day time running lights. DRL are powered at half the voltage of the highbeam. By going direct to 12 Volt from the battery I have no way of controlling voltage to the DRL. I would essentially be driving around with full power high beams during the day and these LEDs are bright.

 

[(*steve*)]

I would go the cap. The PWM must be at a very low frequency if you can see noticable flicker.

Any solution looking for intermediate voltages will fail if PWM is used because the power is either on or off.

Be very aware that placing capacitors in parallel with LEDs can blow them up very effectively if a connection to the LED is intermittent or if they are connected with the capacitor charged.

it is better to connect the cap to the led than the other way around. This is a good logical solution, but the practicalities may may make it moot.

 

[Kiwi]

Would it be better to use the PWM to operate a relay?
The relay contacts could then supply direct battery voltage to the LED lights.

 

[Alec_t]

The Jeep uses PWM for the halogen headlights to extend the life of the bulbs.

Don't you risk shortening the life of the LEDs if you bypass the PWM? It depends on the LED modules/drivers you're using, of course.

 

[Kiwi]

Is the problem with the original halogen headlights poor output brilliance, or poor beam pattern?
Better quality bulbs may help to increase brilliance, but there isn't much you can do for crap design.

Halogen bulbs have a filament that heats up to give out light. The filament heats up and cools down relatively slowly, so its light output can probably be controlled with a fairly slow frequency PWM signal. LED's on the other hand don't have a filament and respond very quickly to voltage changes. I would guess that the Jeep PWM frequency is slow enough for you to see this rapid response as a flicker.

Might be interesting for you to measure the frequency with a multimeter, and also the battery charging voltage.

From memory, halogen bulbs are rated at 13.2v. Increasing the voltage by 5% will give a 20% increase in output, but at the expense of a 50% reduction in expected bulb life.

The voltage in 12v cars has been steadily increasing for some time. 14v was the charging voltage when I started my apprenticeship in the late 70's, but now some cars are charging at over 15v. It has probably got to the stage where it is affecting the bulb life. I believe some manufacturers are now using PWM to limit the effective bulb voltage and thus get acceptable bulb life and light output. It also allows them to dim the headlights so that they can be used as daylight running lights.

 

[Beckerton]

Is the problem with the original halogen headlights poor output brilliance, or poor beam pattern?
Better quality bulbs may help to increase brilliance, but there isn't much you can do for crap design.

Halogen bulbs have a filament that heats up to give out light. The filament heats up and cools down relatively slowly, so its light output can probably be controlled with a fairly slow frequency PWM signal. LED's on the other hand don't have a filament and respond very quickly to voltage changes. I would guess that the Jeep PWM frequency is slow enough for you to see this rapid response as a flicker.

Might be interesting for you to measure the frequency with a multimeter, and also the battery charging voltage.

From memory, halogen bulbs are rated at 13.2v. Increasing the voltage by 5% will give a 20% increase in output, but at the expense of a 50% reduction in expected bulb life.

The voltage in 12v cars has been steadily increasing for some time. 14v was the charging voltage when I started my apprenticeship in the late 70's, but now some cars are charging at over 15v. It has probably got to the stage where it is affecting the bulb life. I believe some manufacturers are now using PWM to limit the effective bulb voltage and thus get acceptable bulb life and light output. It also allows them to dim the headlights so that they can be used as daylight running lights.

You are spot on with all of this. I have already purchased the LED headlights so running them is my preference. Like I said in my initial post, my electrical knowledge is limited but I did think a boost converter could up the voltage when the Jeep's PWM limited it to 8v.

Judging by Steve's post my understanding of the PWM is incorrect. It sounds like the PWM switches off and on, which means a boost converter wouldn't have any input voltage to boost.

I am not sure if all PWM works the way Steve suggests but a fellow Jeeper put his Jeep on an oscilloscope and watched voltage range from 8V to 12.XX volts. I would imagine if voltage switched off and on that it would be visible on the scope.

 

[BobK]

The voltage might indeed vary from 8 to 12V, but that would not be PWM.

Assuming the LED headlamps are intended to by powered by the 12V auto electrical system, I would use a relay to power the lights, and use the "PWM" signal to drive the relay coil through an RC low pass filter.

Bob

 

[Beckerton]

The voltage might indeed vary from 8 to 12V, but that would not be PWM.

Assuming the LED headlamps are intended to by powered by the 12V auto electrical system, I would use a relay to power the lights, and use the "PWM" signal to drive the relay coil through an RC low pass filter.

Bob

Bob,

Can you break this down a little more simply for me? I have never even wired a relay but I aware of how they work. The lights are designed for automotive application. They are 14 LED array behind projector lenses. They are designed to run off of 12V.

Am I wiring the stock wiring that runs to the headlight to the relay? I will have to read up on what an RC Low pass filter is to gain a better understanding.

 

[BobK]

Here is a circuit that might work. The capacitor needs to be determined experimetally, probably in the range of 100uF to several 1000 uF. The relay should be an automotive relay with a high enough current rating for the LED lights. Too small a capacitor would still flicker, too large a one would soften the turn on and turn off (which might be an effect that you like).

 

[Beckerton]

Here is a circuit that might work. The capacitor needs to be determined experimetally, probably in the range of 100uF to several 1000 uF. The relay should be an automotive relay with a high enough current rating for the LED lights. Too small a capacitor would still flicker, too large a one would soften the turn on and turn off (which might be an effect that you like).

View attachment 27583

Bob, does it stand to reason that too large a capacitor may cause a delay between low beam and high beam? That is one complaint with the commercially available anti-flicker harnesses? Thank you for posting the wiring diagram. Much appreciated.

 

[BobK]

Hmm. I don't know how low / high beam is wired. Are there two connections to the LED headlight and two PWM outputs, or is the PWM signal itself changing from low to high? If the latter, you cannot use my solution, it would be on high beam always.

Bob

 

[Beckerton]

Although alternative solutions have been offered I would still like some input as to whether my original idea of using a boost converter would work. It makes sense to me but I am a complete rookie when it comes to electrical setups.

 

[BobK]

No. It has no advantage over the relay circuit I offered.

Meanwhile, you did not answer my questions.

Bob

 

[Beckerton]

No. It has no advantage over the relay circuit I offered.

Meanwhile, you did not answer my questions.

Bob

I didn't answer them because I am not entirely sure of the layout at this point. The headlight itself uses one connection on the back of it. I am assuming that the PWM supplies to low and high beams. I do know that those who have used relay harnesses in the past had to have their Jeep modified by using a $170 programmer to move the Daytime Running Light function from the headlight to the fog lights or turn signal markers. This is because the Daytime Running Light is the high beam at reduced voltage. By running the relay harness like you mention it provides the Daytime Running Light with 12v, in essence turning the DRL into high beams.

 

[JWHassler]

Here's a post that (if true- and it looks legit) might show what you're up against.
A PWM rate sufficient for an incandescent load is way too slow for LEDs

A fully nerded-out solution would:
- read incoming PWM frequency AND duty-cycle (which might be 100%, dunno)
- generate an output at a higher frequency and the SAME duty-cycle

A retriggerable one-shot ORed with the incoming signal would generate an output that would be on for any input: i.e., no high-beam/low-beam/DRL differentiation

 

[BobK]

Agree with JW. The only way to do this right and keep the high / low beam behaviour is to convert the PWM to a higher frequency. It is doable, but would involve some level of skill. Simplest, hardware wise, would be to use a microcontroller.

Bob

 

[Beckerton]

Well I have installed the anti flicker harnesses and now I have lost my passenger side headlight in all headlight stalk positions except "ON". DRL only work on the driver side. Also, when the stalk is in the "ON" position the highbeam on the passenger side will not illuminate.

I believe the anti-flicker harness is a large capacitor for each light.

I would love to try JWHassler's solution but we need to be spoon fed on how to accomplish it.