Car Headlights, PWM
Printed From: the12volt.com
Forum Name: Lights, Neon, LEDs, HIDs
Forum Discription: Under Car Lighting, Strobe Lights, Fog Lights, Headlights, HIDs, DRL, Tail Lights, Brake Lights, Dashboard Lights, WigWag, etc.
URL: https://www.the12volt.com/installbay/forum_posts.asp?tid=131759
Printed Date: May 23, 2024 at 10:11 AM
Topic: Car Headlights, PWM
Posted By: jasonlipka
Subject: Car Headlights, PWM
Date Posted: July 06, 2012 at 11:57 AM
Hey guys, need some expertise here. I have a vehicle that has fancy new headlights, called "SmartBeam". Basically, it's an automatic high beam system. It worked really, really great before I started modding my vehicle. I installed HID's (high intensity discharge) lights. Each light is powered off its own fused +12V bosch relay.
Here's where the trouble began. Come to find out, the Smartbeam system doesn't just turn the high beams on and off. When it senses a car far off in the distance, it first begins dimming the high beams with PWM. When it dims them down to a nominal (and unknown) level, it shuts them off entirely. The problem is the PWM section. Obviously, my relays don't like this PWM at all. They are erratic on when they finally shut off and they make a terrible buzzing noise during the entire PWM cycle as they open and close a billion times a second.
I'd like to get rid of the PWM signal. There isn't a discrete wire in the car that I can pull off of thanks to the new CANBUS wiring system. I even tried the high beam indicator in the dash but it is a module. I have tried SSR's and they don't work. So I need some help. How do I do this?
I thought maybe a transistor as its saturation voltage is much lower than the pull in or hold in voltage of the relay coil. Would that work? I don't know how a transistor would respond to having a PWM signal applied to its base.
The other option is to filter it out with an R-C network right? Any ideas there?
Now for the bad news. I don't know the duty cycle, decay rate, or frequency of the PWM signal. I don't have an oscilloscope either. I have a DMM and am quite handy with it.
Replies:
Posted By: i am an idiot
Date Posted: July 06, 2012 at 4:21 PM
Parallel a 4700 microfarad 35 volt capacitor across the coil of the relay. If that does not totally take care of the problem, place a 6 amp diode in series with the wire feeding the coil of the relay.
Posted By: faja
Date Posted: July 15, 2012 at 11:27 AM
check your owner's manual or with the dealer, there should be a way to simply disable the "smartbeam" option.
Posted By: jasonlipka
Date Posted: December 12, 2012 at 9:17 AM
i am an idiot wrote:
Parallel a 4700 microfarad 35 volt capacitor across the coil of the relay. If that does not totally take care of the problem, place a 6 amp diode in series with the wire feeding the coil of the relay.
I will give that a try. Any chance I will need to play with capacitance value? I only ask because it would save me money to just buy a couple different capacitor values using your guidance now than to keep buying them through trial and error.
Also, for my edification, what purpose does the diode serve? Not doubting you, I am just really curious how the diode affects the circuit.
Thanks for your input.
faja] wrote:
check your owner's manual or with the dealer, there should be a way to simply disable the "smartbeam" option.
I know how to disable them, they are currently disabled. What I am saying is that the system functioned perfectly in automatic mode before I swapped to HID's. The HID ballasts are powered through relays. The relays don't like the PWM. It can be avoided by just disabling the smartbeams, but they function so well in automatic that I'd like to try getting them to function with relays in the circuit.
Posted By: oldspark
Date Posted: December 12, 2012 at 4:25 PM
Remove the relays. Relays should not be PWM'd; only the lights. And since the Smartbeam is already supplying the power for the HIDs, the relays are unnecessary (unless BEFORE the Smartbeam).
The PWM should also be after the HID ballast depending on the ballast (some don't like the on-off PWM cycling) but it would have to be a grounding type (in case it can't handle the higher HID voltages).
Posted By: i am an idiot
Date Posted: December 12, 2012 at 5:08 PM
I see no reason that the 4700 will not do what you are trying to do. But you can order a 1000, and a 2200 while you order. Try the 1000 by itself, if it does not do what you need, parallel the 2200 also. That will be 3200 total. If that still does not work then try the 4700. I am fairly confident that the 4700 along with the 6 amp diode will do the trick.
Posted By: oldspark
Date Posted: December 12, 2012 at 7:50 PM
In that case I'd be tempted to remove or bypass the Smartbeam system altogether since its PWM dimming will be defeated - only its automated switch-down to low beam (below 27kph etc) would remain.
Other sources also reckon the PWM causes havoc with HID ballast. (Though as I say, that depends on ballast design, though shorter life may ensue even if they do work.)
Posted By: jasonlipka
Date Posted: December 13, 2012 at 1:17 PM
oldspark wrote:
In that case I'd be tempted to remove or bypass the Smartbeam system altogether since its PWM dimming will be defeated - only its automated switch-down to low beam (below 27kph etc) would remain.
Other sources also reckon the PWM causes havoc with HID ballast. (Though as I say, that depends on ballast design, though shorter life may ensue even if they do work.)
I tried hooking the ballasts directly to the high beams and the lights still flcker. The PWM must be removed from the circuit.
Posted By: jasonlipka
Date Posted: December 13, 2012 at 1:19 PM
i am an idiot wrote:
I see no reason that the 4700 will not do what you are trying to do. But you can order a 1000, and a 2200 while you order. Try the 1000 by itself, if it does not do what you need, parallel the 2200 also. That will be 3200 total. If that still does not work then try the 4700. I am fairly confident that the 4700 along with the 6 amp diode will do the trick.
Okay, I have them on order. Not gonna be here until next week.
I ask again, what purpose does the diode serve?
Posted By: i am an idiot
Date Posted: December 13, 2012 at 9:36 PM
For the times that the PWM is at 0, the diode keeps the cap from trying to keep that voltage up. It will use all of it's capacitance to keep the relay energized.
Posted By: jasonlipka
Date Posted: December 14, 2012 at 8:05 AM
i am an idiot wrote:
For the times that the PWM is at 0, the diode keeps the cap from trying to keep that voltage up. It will use all of it's capacitance to keep the relay energized.
Got it.
Should I be concerned that excessive capacitance may keep the relay pulled in after the computer has signaled for the high beams to be off?
Posted By: i am an idiot
Date Posted: December 14, 2012 at 2:38 PM
The time will be minimal.
Posted By: jasonlipka
Date Posted: December 14, 2012 at 3:47 PM
i am an idiot wrote:
The time will be minimal.
Are we talking minimal like on the order of milliseconds? I only ask because with the way the system functions, being pulled after the signal is gone can get counter-productive.
If I find the time to be excessive, can I keep dropping the capacitance down to reduce hold-in time? Kind of "fine-tune" it until I achieve an optimal balance?
Posted By: i am an idiot
Date Posted: December 14, 2012 at 8:43 PM
Yes you can reduce capacitance to shorten the hold time. I do not remember how long a 4700 will keep a relay energized, but I am thinking around a second.
Posted By: jasonlipka
Date Posted: December 17, 2012 at 8:49 AM
i am an idiot wrote:
Yes you can reduce capacitance to shorten the hold time. I do not remember how long a 4700 will keep a relay energized, but I am thinking around a second.
Yikes. That is way too long. When you are talking about dimming high beams for oncoming traffic, a second is like a lifetime.
I guess I'll have to play with capacitance. I ordered some 220 uF caps, too. Will that be enough? If I'm not mistaken, I only need enough capacitance to keep what I will call the "effective voltage" above the hold-in voltage of the relay's coil, correct?
Any guesses at how much capacitance will be needed?
|