LED driver
Maxim Jones , 07-27-2017, 09:34 AM
Hello every one,
Attached is the design for a LED driver. The circuit uses MAX16834 as a LED driver and is controlled by Arduino. We made the design according to the datasheet of MAX16834 and it has been used both for PWM dimming as well as Analog dimming. In order to make the checks easy to do, we have done all the checks for Analog dimming (ie., dimming by changing voltage)
In this circuit, for dimming with the input voltage, a PWM system has been used that after passing through a LOW pass filter, a variable DC voltage is created for the dimming. The LED panel works with 22V and 1.4 Amp.
The problem that you can see in the images is with the gate of the Q1 transistor.
As you can see, the pulse on the gate is not stable and in some of the cycles suddenly drops to zero. This happens once per several cycles and depending on the level of analog dimming it gets various shapes.
For example, within the voltage change of 5-7%, there is no problem, but when the voltage change is above this figure, the problem appears and 100ns after the rising edge of the gate, it suddenly drops to zero. As a result, the LED current never goes beyond 0.5A because some cycles are missed.
We have done the following checks:
Kind Regards,
Maxim
Attached is the design for a LED driver. The circuit uses MAX16834 as a LED driver and is controlled by Arduino. We made the design according to the datasheet of MAX16834 and it has been used both for PWM dimming as well as Analog dimming. In order to make the checks easy to do, we have done all the checks for Analog dimming (ie., dimming by changing voltage)
In this circuit, for dimming with the input voltage, a PWM system has been used that after passing through a LOW pass filter, a variable DC voltage is created for the dimming. The LED panel works with 22V and 1.4 Amp.
The problem that you can see in the images is with the gate of the Q1 transistor.
As you can see, the pulse on the gate is not stable and in some of the cycles suddenly drops to zero. This happens once per several cycles and depending on the level of analog dimming it gets various shapes.
For example, within the voltage change of 5-7%, there is no problem, but when the voltage change is above this figure, the problem appears and 100ns after the rising edge of the gate, it suddenly drops to zero. As a result, the LED current never goes beyond 0.5A because some cycles are missed.
We have done the following checks:
- Shorting R8 (CS pin to GND) but it did not resolve the problem
- Adding 0.22uF Capacitor to green points and GND, but it didn’t resolve the problem either.
Kind Regards,
Maxim
robertferanec , 07-27-2017, 10:01 AM
I would maybe check REFI (?)
Maxim Jones , 07-27-2017, 10:12 AM
yes, please
robertferanec , 07-27-2017, 10:46 AM
I mean, if I was you I would use an oscilloscope and measured the signal on REFI.
Maxim Jones , 07-27-2017, 02:14 PM
Actually I did it and the attached pictures are result of my debugging and testing that I took screen shot of my oscilloscope that has communication with my PC. Also I'm testing other probability and I will send you its result.
Comments:
Maxim Jones, 08-03-2017, 02:20 AM
Hi, I solved the issue. In my last testing, I added a RC filter ( one capacitor 1nF and one resistor 100 Ω ) to CS pin. The problem was solved. But I’m not sure it causes bad effect on the function of the pin because I didn't find such filters that has been used for this IC. But there is such filters that use for (max668 and max15005).
robertferanec , 08-04-2017, 08:08 AM
Awesome. Thank you for letting us know.
Use our interactive Discord forum to reply or ask new questions.