LED being dimmed or not showing any light Issues

Thread Starter

justinvil1103

Joined Apr 6, 2016
41
Hello there Friends,

We are currently seeing some issues with a LED driver circuit. I am not sure about root cause because there are too many variables. During operation the LEDs are being dimmed or not light at all (After 2-3 months in operation ). Attached is the portion of the driver circuit for the LED, any ideas that would cause (Excessive electrical stress in the form of excessive current applied to the LED) the LED to be Dimmed or damaged. What would caused excessive current, beside a short we have too many LED being damage or go dim. Any help would be greatly appreciated

Thank You
WJLED Driver.jpg

LED SPEC.jpg
 

Papabravo

Joined Feb 24, 2006
21,158
If we neglect the voltage drops of Q10 & Q12, we have 7.5V voltage source, an unspecified Vf (forward voltage drop) [Pick 3.0V for the sake of argument], and a resistor of 121 Ohms
So: (7.5 - 3.0)/ 121 Ω gives a forward current of 37 mA which, sad to say, exceeds the ABSOLUTE MAXIMUM RATING for forward current. Also 37 mA * 3.0V = 112 mW, which is close to the ABSOLUTE MAXIMUM RATING for power dissipation.
We call this SLOW LED DEATH, and whoever did the design it would seem, either did not read or did not care to read the TFDS.

Tell me what the properties of the forward drop are and I can suggest a fix.

EDIT:
  1. What is the purpose of the DAC waveform?
  2. Where is the opamps DC path to Ground when Q12 is OFF?
 
Last edited:

Papabravo

Joined Feb 24, 2006
21,158
Why do we need to guess what is the very important power supply voltage for the opamp?
The schematic shows U13B. The power connections are probably on U13A, another opamp in the same package, in a part of the schematic that was not shown. Complete information is most helpful if the TS actually wants help.
 

LowQCab

Joined Nov 6, 2012
4,022
PWM for Dimming is preferred over a 0-5V analog signal,
along with a proper Current-Regulator instead of just a Resistor.

A whole "re-do" of the Circuit is in order.
.
.
.
 

Thread Starter

justinvil1103

Joined Apr 6, 2016
41
If we neglect the voltage drops of Q10 & Q12, we have 7.5V voltage source, an unspecified Vf (forward voltage drop) [Pick 3.0V for the sake of argument], and a resistor of 121 Ohms
So: (7.5 - 3.0)/ 121 Ω gives a forward current of 37 mA which, sad to say, exceeds the ABSOLUTE MAXIMUM RATING for forward current. Also 37 mA * 3.0V = 112 mW, which is close to the ABSOLUTE MAXIMUM RATING for power dissipation.
We call this SLOW LED DEATH, and whoever did the design it would seem, either did not read or did not care to read the TFDS.

Tell me what the properties of the forward drop are and I can suggest a fix.

EDIT:
  1. What is the purpose of the DAC waveform?
  2. Where is the opamps DC path to Ground when Q12 of OFF?
Thanks for your reply, I am not sure what you mean by "properties", but the LED Vf =2.4V. What is T.F.D.S stand for?
 

Papabravo

Joined Feb 24, 2006
21,158
Thanks for your reply, I am not sure what you mean by "properties", but the LED Vf =2.4V. What is T.F.D.S stand for?
The essential properties are the forward voltage Vf, at a particular current of interest. With a Vf of lower than 3.0 volts, that I assumed, it is likely that you are running too much current through the LED and also dissipating too much power. You need to go back and analyze the forward current and power dissipation under the worst case conditions of DAC output voltage when Q10 is turned on, AND when Q10 is off, the opamp has no DC path to ground. This may or may not be a bad thing. I don't know how you are actually working these two things.

T.F.D.S = The Freaking Data Sheet
 

Thread Starter

justinvil1103

Joined Apr 6, 2016
41
The essential properties are the forward voltage Vf, at a particular current of interest. With a Vf of lower than 3.0 volts, that I assumed, it is likely that you are running too much current through the LED and also dissipating too much power. You need to go back and analyze the forward current and power dissipation under the worst case conditions of DAC output voltage when Q10 is turned on, AND when Q10 is off, the opamp has no DC path to ground. This may or may not be a bad thing. I don't know how you are actually working these two things.

T.F.D.S = The Freaking Data Sheet
Thank you very much.
 

LowQCab

Joined Nov 6, 2012
4,022
There is no efficiency advantage for the same average current unless you add an inductor, so why use PWM in this application?
The circuit posted is a current regulator.
When on, the current value is the DAC input voltage divided by R59.
That's true, that there is no efficiency advantage without an Inductor,
but it's still better than varying the Voltage for Dimming,
which tends to make the Colors get wonky.

The Circuit is "kinda" a Current-Regulator, but with a Current-Sense-Resistor that
is large enough to be used as a generic "Fixed-Voltage" LED-Current-Setting-Resistor,
it could be somewhere around ~10-Ohms, or even less,
although it depends on how You want to go about spreading-out the Heat.
It also doesn't need the separate Transistor, or the FET.
.
.
.
 

crutschow

Joined Mar 14, 2008
34,280
The Circuit is "kinda" a Current-Regulator
So you've gone from it's not a current regulator to it's "kinda" a current regulator.
Define "kinda".
It looks like a good current regulator to me.
The size of the sense resistor is not a large factor (within limits) as to how good a regulator it is.
 

LowQCab

Joined Nov 6, 2012
4,022
You know You could build a far superior Circuit,
so why are You giving me a hard way to go ?

There are very few absolutes in this world,
but there's a lot of space between better and worse.
.
.
.
 

crutschow

Joined Mar 14, 2008
34,280
You know You could build a far superior Circuit,
so why are You giving me a hard way to go ?
How superior?
The posted circuit is a pretty stiff current source, determined mainly by the open-loop gain of the op amp.
So i really don't understand why you apparently think it is not, from your "kinda" statement. (?)
I apologize if you feel that is giving you a "hard way to go".
 

Thread Starter

justinvil1103

Joined Apr 6, 2016
41
How superior?
The posted circuit is a pretty stiff current source, determined mainly by the open-loop gain of the op amp.
So i really don't understand why you apparently think it is not, from your "kinda" statement. (?)
I apologize if you feel that is giving you a "hard way to go".
Thank you for your help.

@crutschow & @LowQCab would you both show me another example besides my posted circuit?

I would like to see both circuits and try it.
 
Top