Subaru Forester Owners Forum banner

1 - 20 of 26 Posts

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #1
The subject of dimming lights by altering the duty cycle brings to mind another question. Has anybody looked into replacing the DRL resistor with a chopper circuit? I've never been keen on turning electricity into heat for no good reason.
 
T

·
Guest
Joined
·
0 Posts
I drew up an adjustable chopper circuit to do it just before our winter trip last year, but didn't have enough time to work on it. Needless to say I'm back and it's on my low priority list or no priority list.

My DRL's are on the highbeam filament and are using the stock 0.3 ohm power resistor.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #3
Popular Electronics (now defunct) once had an article that did this with a 555 and a high current MOSFET. I think I lost my copy of that issue long ago. But it shouldn't be too difficult to reconstruct the concept. I guess we'd have to establish the proper duty cycle by experiment.
 

·
Premium Member
2014 Impreza Ltd CVT
Joined
·
3,319 Posts
I know what you are trying to accomplish. But that's a lot of engineering for what a simple resistor is currently doing. (ha, get the pun)

Also gating a junction off/on rapidly, with the current of a headlight WILL produce heat, heat that will have to be dissipated from the junction. You are controlling current in amps here, not ma as with the interior lights.

Here's a thought for you, if you want to get rid of the heat, Toyota's use a relay to connect the High Beams in series for their DRL's. You can use the DRL relay to drive another multi-pole relay to do just that. And no additional heat to dissipate.

But this is how things are accomplished......5% inspiration.....95% perspiration.
 
T

·
Guest
Joined
·
0 Posts
There's actually very little heat generated using mosfet's as long as you stay out of the linear region when biasing the gate. Using bjt's is another matter. These will heat up. Another thing is to use a lower frequency to switch the mosfet like 1kHz. And yes, I'm talking amps not milliamps. I calculated the worse case current to be around 12amps.

The plan was to use a uC (specifically the ATtiny15L) to control the mosfet. It has an ADC input that you can attach a pot to adjust the duty cycle of the PWM output to the mosfet. You can actually adjust the duty cycle of the PWM output on the 555, but using software simplifies the design.

Toyota, like many other manufacturers, are using their highbeams for DRL's. Most of them use a dropping resistor, like Subaru (except Subaru is using the low beams), to decrease the intensity. I read somewhere that a company was using PWM to decrease the intensity, but it was unreliable. There's no problem getting rid of the heat. We just short out the dropping resistor or remove it entirely from the circuit.

I guess we just hijacked this thread. BBottomley should start another thread. I'll add the conceptual designs (schematic) just to make it interesting. My initial design consisted of only 2 wires. I didn't find any time to test the theory though. Maybe someone wants to do it?
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #6
Yes, we (I) hijacked the thread. I'll see if the moderators can move our posts to a new one. - DONE

I once saw a circuit for achieving a high range of duty cycle with a 555 that just added a diode to the normal multivibrator configuration. I cut it out of the mag but now I have no idea where I filed it.
 

·
Premium Member
Joined
·
1,229 Posts
If it is heat dissipated in the DRL resistor you don't like, then bypass the resistor! This will give you full power to your low beams at all times. This will also make your bulbs stay brighter for longer life.
 
P

·
Guest
Joined
·
0 Posts
If you can supply me with a circuit diagram and a partslist i can supply the assembled circuit boards - www.psi3.co.uk is our second business our main company is www.auwell.co.uk PCB Design, manufacture, assembly and test ( all in house )
I will do this at cost price and not charge for the CAD and tooling.
Alternatively i am quite happy to just make the PCB's and you can assemble/source the components yourselves.
Not looking to make money just offering a service i can provide to fellow enthusiasts.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #9
MountainBiker said:
If it is heat dissipated in the DRL resistor you don't like, then bypass the resistor! This will give you full power to your low beams at all times. This will also make your bulbs stay brighter for longer life.
The rationale for doing some mods is like climbing mountains - because they're there!
 

·
Premium Member
Joined
·
1,229 Posts
bbottomley said:
The rationale for doing some mods is like climbing mountains - because they're there!
I hear ya!

But my suggestion may still be the better mod, at least in results. When you run halogen bulbs below their rated voltage they create soot, which ends up coating the glass, decreasing light output. I'm not sure how bad this is from the DRL resistor, but it certainly isn't good.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #11
As you noted, running them below rated voltage is what happens now with the stock circuit. I don't know the physics involved, but it could be that controlling the light intensity through pulse width modulation of the full voltage would be better, or at least no worse.
 
T

·
Guest
Joined
·
0 Posts
Here's the schematic. The idea is to replace the dropping resistor with this circuit.

D2 is the blocking diode when the mosfet shorts to GND (Gate on positive transition)
C2 provides the voltage (power) during the mosfet ON transition.
R3 varies the output of the PWM
R1, D1, and Q1 is just a voltage regulator.

All of this should fit in the palm of my hand. Probably smaller if we went to surface mount. C2 can't get any smaller (1000uF, 24V). Probably half of a AA battery.

What I don't know:

1. How much power will the components sink when "A" goes to GND when the mosfet is enabled. Will the size C2 be enough. My experience with the ATtiny15L that it's low power.

2. Turn On and Turn Off times.


Comments? Any other design ideas?

I guess I should try it seeing that I have access to all these parts.
 

Attachments

T

·
Guest
Joined
·
0 Posts
MountainBiker said:
But my suggestion may still be the better mod, at least in results. When you run halogen bulbs below their rated voltage they create soot, which ends up coating the glass, decreasing light output. I'm not sure how bad this is from the DRL resistor, but it certainly isn't good.
After doing some reading on halogens (more than I want to), the magic number is 10% over or under the rated voltage will start causing trouble. We'll use 14.4V as the operating voltage.

10% of 14.4V is 1.44V over or under the 14.4V.


With the low beams (55W):
55W / 14.4V = 3.82A ... 3.82A * 0.3ohm (dropping resistor) = 1.15V

With the high beams (60W):
60W / 14.4V = 4.17A ... 4.17A * 0.3ohm (dropping resistor) = 1.25V

With the dropping resistor Subaru is within the limits of the 10%.


Driving the lights using PWM (mosfets), the filaments will see the full rated voltage. There might be some drop across the mosfet, but it's minimal (I calculated 0.06V @ 55W or 60W).


PWM Plusses:
1. Minimal heat/power loss.
2. Ability to control light output while supplying full rated voltage.

PWM Minuses:
1. Not as simple as hooking up a resistor.


IMHO, PWM is better.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #14
More PWM Plusses:

3. Replaces a design that's offensive in that it uses power for no good reason.

4. Gives engineers a project to design and perhaps even build.
 

·
Premium Member
Joined
·
1,229 Posts
tooxtraxtra said:
Driving the lights using PWM (mosfets), the filaments will see the full rated voltage.
I like your analysis and research results! However, I think you are mistaken on this point. The PWM circuit is a very good way to accurately control the effective light output of a halogen bulb. But the filament in the bulb will still react as it would with a lower supplied voltage. This is because the filament cannot react quick enough to the 14.4 volts and then 0 volts, and then 14.4 volts pulses (For any readers that don't know, this is what is supplied by the PWM circuit). The filament kinda (technical term there!) averages out the 14.4v and the 0v pulses, giving you less light just like you get with a lower, steady state, voltage. If you are low enough in the effective voltage average, sooting will still happen.

However, I do agree that with the PWM circuit, no power is wasted in a resistor! And I hate wasted power!
 

·
Registered
Joined
·
192 Posts
ahem, the filiment will see whatever voltage you give it!

butt, and its a big but,

what we are really concerned with here is the temperature of the filament. if you make the bulb run at reduced output, but it lower constant voltage or some sort of high frequency switching arrangement, the filiment will settle at a steady state temperature correlating to the average power it is dissapating.

so the volts are different, but the average power is the same and the effect is also same as long as the switching is fast enough to be undetectable.

aren't there better mods to worry about? come on, there are resistors everywhere, get over it! if you want to save power and come out with a better car, build a grounding kit or something.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #17
Come, now. The ethic of this forum says that it's OK to discuss the merits of various options, but we stop short of telling people to get over it. We're interested in this because it's interesting.
 

·
Administrator
2018 2.5i Premium CVT
Joined
·
18,741 Posts
Discussion Starter #19
Appreciate it. Thanks.
 
T

·
Guest
Joined
·
0 Posts
MountainBiker said:
I like your analysis and research results! However, I think you are mistaken on this point. The PWM circuit is a very good way to accurately control the effective light output of a halogen bulb. But the filament in the bulb will still react as it would with a lower supplied voltage.
You say that PWM is a good way of accurately controlling the effective output, but can you give me that "lowered supplied voltage value" that you think the filament will still see?

MountainBiker said:
This is because the filament cannot react quick enough to the 14.4 volts and then 0 volts, and then 14.4 volts pulses (For
any readers that don't know, this is what is supplied by the PWM circuit).
Can you elaborate when you say, "the filament cannot react quick enough?"

MountainBiker said:
The filament kinda (technical term there!) averages out the 14.4v and the 0v pulses, giving you less light just like you get with a
Can you tell me what the filament is doing to cause the 14.4V and 0V pulses to average out and what value does it average out to?

MountainBiker said:
lower, steady state, voltage. If you are low enough in the effective voltage average, sooting will still happen.
Can you tell me what is the "low enough effective voltage average" that sooting will happen and what if I go above it?

It's good that you're contributing to this thread, but as an electrical engineer I need the hard facts (data) that will tell me (and others reading) why this won't work and not because you said so.

I did some more calculations and my values on Subaru's 0.3ohm resistor were incorrect (somewhat). I also did some for the PWM. My calculations tell me it will work and not soot.
 
1 - 20 of 26 Posts
Top