I have a batch of LM3435 LED driver boards where the green LED is not illuminating. On one of these boards it does work (although I already sent that out to a developer so I can't do a differential on it). The schematic was reviewed by TI.
I've checked a lot of things but I have some questions to help me debug better.
- I verified that the green control signal is coming in.
- I see a small voltage on iref
- I see output on the GLED pins although it is different than RLED and GLED
- I see matching waveforms on CR and CB, but a smaller one on CG
- Fault checking is enabled, no faults are reported.
- The path from VOUT, and CG to the green LED ohms out
- Registers 0x01, 0x02, 0x03 are all set to 0x5E. Setting 0x01 (Green) to 0xFF, and the other two to 0x00 results in no output on the green LED.
The voltage on VOUT seems to move up and down. I wasn't sure if this was expected, and part of the control scheme or if something was wrong? It's like a small square wave with a DC offset. The RED output voltage waveform is lower than VOUT and seems to match the BLUE output voltage. However the GREEN output voltage seems to follow the Vout waveform almost exactly except with a small voltage difference. It makes me think no current is flowing through the green output.
I know waveforms would be helpful but I was too tired to take them :)
So anyway I was curious if VOUT was expected to move up and down a little or should it be steady like a normal regulator output? edit: I read more and it seems like is does move to maintain efficiency.
I assume that to power the green LED current flows from Vout, through the LED, back to GLED pins, and then through a mosfet to GND? Does that use a specific GND pin? In other words is the output of the GLED mosfet bonded to only one or two GND pins? I'm wondering if maybe one of them might not be soldered well enough.
Any pointers on what I should look at or measure that might point me in the right direction?
Thank you
