PCA9685 i2C pwm driver: 0.5V at output when off

I#M working on a 24channel LED Dimmer with up to a couple Amps per Channel an 12-24V.

For PWM generation I use PCA9685 I2C Controller.

I have now my prototype PCB with 2 PCA on it.

While I measure 0V at one PCA9685 channels when they are “off”, I measure 0.5V at the other PCA9685.
When they are on, everythings fine with 5V.

I added a 10k Pulldown, but that didn’t solve it. Voltage only Down to 0,3V.

I configured both PCA to Totem Pole (which is the default).

“The LED output driver is programmed to be either open-drain with a 25 mA current sink capability at 5 V or totem pole with a 25mA sink, 10 mA source capability at 5 V.”

But also with Open Drain there is no difference.

I checke with my breadboard test setup where I used breakouts for PCA9685 and there I also measure 0.5V when the channels are off.

I cannot explain the difference on my PCB (for reference see attachment).

But I can also not explain why there are 0.5V when off…
I don’t know if this causes problems, for now I’m just confused and curious.

KRS-XL_DIM-24CVLED.schematic.pdf (141 KB)

While I measure 0V at one PCA9685 channels when they are "off", I measure 0.5V at the other PCA9685... I checke with my breadboard test setup where I used breakouts for PCA9685 and there I also measure 0.5V when the channels are off.

You are saying that both on the PCB and on the breadboard, one pca chip behaves differently to the other pca chip? That sounds like you have a bad chip.

yes, but 3 of them show behaviour which I would call “strange” (0.5V at “OFF”) and on of them the behaviour I would expect (0V at "OFF))

And it seems the 0.5V at OFF do ot cause serious trouble on my breadboard setup.

SirSydom:
The LED output driver is programmed to be either open-drain with a 25 mA current sink capability at 5 V or totem pole with a 25mA sink, 10 mA source capability at 5 V.

You shouldn't use any code for that in your sketch.
The default setting are just fine for driving mosfets.

Diagram looks fine. Could it layout?
Did you put the decoupling caps near the VDD/VSS pin of the chips.
I have used many PCA9685 driving LED drivers, and never had issues.
Post a picture of the build.

I suppose you have measured that voltage with respect to board ground, not Arduino ground.
Leo..

I found the issue - the VSS pin at U2 was soldered badly.
The VSS Pin was at 0.5V.

I resolderd it, and now everything is fine at 0V.

And not, I did not measure it in respect to arduino ground, this would be stupid because they are isolated…

Another thing, that on my beadboard I also measured ~ 0.5V while off. The reason here - I also measured not in respect to the VSS pin, but on some GND point on the breadboard. There connections on the breadboard are so bad, that they also produced VSS to be at 0.5V…

Thats the device, btw…