Independent PWM to two seperate LED arrays

Thread Starter

20_rc51_00

Joined Oct 30, 2014
14
Hi,

Just want to make everyone aware I have a very cursory understanding of what these components are doing and how they are interacting.I have pieced together this schematic with a lot of help from others and based in on what I can understand.

My goal is to have Tail/turn/brake function to two separate led arrays. Both arrays be PWM'd to have dim light-up, bright light up and have independent flashing bright light up while the other is dim or bright.

I have set this up on a breadboard and it works except....

Problem: Its only a minor problem from a functional stand point but I'm not sure if it's a hazard or risk to the components. When I have the right and left LEDs powered and dimmed with PWM function and I introduce turn signal input by providing 12V to the 4th pin the LEDs brighten as they should , but the other side which is dim seems to brighten just a little bit aswell. It's really subtle but it's there. I'm not sure how to better isolate the turn signal input from one side to the other? Is this causing damage to the PWM or anything else?

Any advice would be appreciated. Thanks.
PS. please excuse the weird way I had to show power and ground locations.
sorry I'm not sure why there are 2 images.
TAIL-LIGHTS SCHEMATIC 2.png TAIL-LIGHTS SCHEMATIC 2.png
 

ebeowulf17

Joined Aug 12, 2014
3,307
I'm not at all certain of this, but it seems like maybe diodes D3 & D4 should be between the resistors and the pwm output. That would prevent any positive voltage from one turn signal input from reaching pin 4 of the other side's regulator. I'm not sure if the diode is serving some purpose in its current position that would be lost in the move. Good luck!
 

dannyf

Joined Sep 13, 2015
2,197
but the other side which is dim seems to brighten just a little bit aswell.
The signal is cross through the two 2.2k resistor and the output impedance of pwm chip / controller.

One option is to use resistors bigger than 2.2k; Or to lower the output impedance of the pwm chip / controoler.
 

Thread Starter

20_rc51_00

Joined Oct 30, 2014
14
The signal is cross through the two 2.2k resistor and the output impedance of pwm chip / controller.

One option is to use resistors bigger than 2.2k; Or to lower the output impedance of the pwm chip / controoler.
Thanks for taking the time.
I tried with a 100KOhm but the result was the same.

I'm not sure how to change the impedance. Any advice on that?
 

dannyf

Joined Sep 13, 2015
2,197
try this:

1) replace the 2.2k resistor with a diode, from the pwm controller to ka378;
2) add a resistor (1k-10k, doesn't really matter) from ka378's pin 4 to ground.

Essentially, if the brakes are on, the leds are in full brightness; if the brakes are off, you can control the led via the pwm dimmer.
 

Thread Starter

20_rc51_00

Joined Oct 30, 2014
14
I'm not at all certain of this, but it seems like maybe diodes D3 & D4 should be between the resistors and the pwm output. That would prevent any positive voltage from one turn signal input from reaching pin 4 of the other side's regulator. I'm not sure if the diode is serving some purpose in its current position that would be lost in the move. Good luck!
Thanks for taking a look.

I switched the positions of the diodes and also added 100kohm resistors but it was the same.
 

Thread Starter

20_rc51_00

Joined Oct 30, 2014
14
try this:

1) replace the 2.2k resistor with a diode, from the pwm controller to ka378;
2) add a resistor (1k-10k, doesn't really matter) from ka378's pin 4 to ground.

Essentially, if the brakes are on, the leds are in full brightness; if the brakes are off, you can control the led via the pwm dimmer.

I removed the 2.2k resistors and in their place I put the diodes, I added a 4.5K resistor from pin 4 to G. When I powered it all up the LED's did not light dim or bright. Removing the resistor from 4 to ground allowed the LEDs to come back on dim as it did before. Touching a positive lead to pin 4 of ka378 to simulate the turn signal function results in brightening as before as well. It seems that the resistor from pin4 to G is a short and prevents all function. ( regarding the diode I tried both anode toward pwm and anode toward ka378, it only retains PWM function when anode toward ka378 ).

as a note: I am using the out- pin of the PWM for it's function, which is connected to the 4th pins ka378, not sure if you caught that (or if it matters..). I have seen other schematics where the out+ is used.....

Dannyf, In what way would it be supply related? Not exactly sure what that means.
Thanks!
 

dannyf

Joined Sep 13, 2015
2,197
What output impedance does the pwm controller have? Does it need a pull-up there?

I was suspecting that when the leds turn on, it may have dragged down the supply voltage and created a cross-link of some sorts.
 

dannyf

Joined Sep 13, 2015
2,197
I thought back then that the two sides got cross-linked through the pwm controller's output impedance: when one side's turn-signal is on, that signal goes through the diode, its 2.2k and the pwm's output impedance attenuated it and shows up on the other side's pin 4.

Maybe the following sim will help explain this a little bit more.

V2 is your pwm controller, R5 its output impedance. V1/V3 are the turn signals. V1 in this case is 0 so it should be dim. V3 is an on/off signal, simulating the application of a 12v signal periodically. The waveform is the signal showing up on the pin 4 of the left side led driver.

As you can see, the application of a 12v from V3 pushes up the signal on left side's pin 4. If the threshold is such that (between 1.5v and 2.7v in this case), the left side leds can turn on (50% duty cycle) when the right side's turn signal shows up.
 

Attachments

Thread Starter

20_rc51_00

Joined Oct 30, 2014
14
Yes I think I understand.

So, as I'm understanding it, it's the presence of the 12V signal from one side that changes the impedance of the PWM's output, which causes the slight brightening of the light on the opposing side. So then would it be possible to include something (a resistor I'm guessing) to keep it the at the same level when it gets a 12V signal on one side? And also keep it the same when the 12Vsignal is removed? Is there some way to have this added resistor recognize when it should be functioning and when it should deactivate?
 
Top