Lighting PWM, No output

[Brian71583]

I am new to this and I have an operational reef-pi, For my lighting circuit build I cannot seem to get it right. I even rebuilt a second board with new components to start over and try again and have the same results.

I am certain I have it built to the specs on the guide. If I test at the output pin on the PCA9685 board I have 5V. If I then connect that output to my circuit board and test the output of the board rather than 10v I get about .1v

I'm lost

 

[Brian71583]

I tested further and realized I left out some details I’m sure people will ask.

The mosfet board for the 12v led strip works fine.

I have 5v coming out of the PCA9685 PWM modulation board and then obviously going into the board were the lighting circuit is built.

I originally built the circuit as shown in the drawing in the official guide and it didn’t work so I rebuilt as shown in the photograph. Same way just different layout. Neither way works.

I have my 10v and an operational PWM 9685 but am only getting about .10v coming out of the lighting circuit at the blue pins.

Also, with the functionality of the 12v circuit and ability to dim it through the UI I believe that tells me the raspberry and reef pi are set up correctly.

 

[theatrus]

Can you link the guide and any pictures of your setup?

 

[jolthoff]

Two most common problems with PWM are either a hardware or software problem.

Hardware-- make sure your wires are connected properly, *especially* your grounds. Ground from Pi to PCA9685 and ground from PCA9685 to lighting module PWM in. Make sure your SDA/SCL for I2C lines are connected properly.

Software-- I'm not using a Reef-Pi, but make sure you have the right drivers/software running and set appropriately. Make sure your PWM frequency is appropriate for your hardware. Again, not sure how ReefPi works, but make sure your I2C pins are set correctly to the ones you're using (if applicable-- sometimes you can't change the pins, but make sure the software doesn't set the i2c pins to ones you're not using).

If you have access to an oscilloscope, make sure that signals are being sent from the Pi to the PCA9685 when you'd expect (on startup/running any program to initialize the PCA9685). Don't have to debug, just see that they're being sent-- presumably you're not writing a driver/interface and don't have to debug that. Also, if you have an oscilloscope, you can also see if the PWM signal is coming out of the PCA9685 correctly.

 

[Brian71583]

Can you link the guide and any pictures of your setup?

Here’s some pictures. I went off of this guide…

reef-pi Guide 5: Light Controller

This guide will cover how to use reef-pi to create a light controller that can simulate daylight cycle like natural coral reefs. We'll cover two types (low current and high current) of light controller circuits.

learn.adafruit.com

I originally built the board exactly like in the illustration type picture below but when I couldn’t get it to work I rebuilt it like the white board pictures below. Both the illustration and the white board picture are from the same guide.

so the red board is mine and I believe is exactly like the white board in the guide (minus the fact I set it up for 2 light instead of 3)

the red wire going into the left of mine feeds 10v to the bottom rail.

the top rail feeds 5v which also jumps a 5v to the PWM board and also to the Pi.

PWM board has a ground to the Pi and the lighting circuit.

I have 5v at the green jumper going into the lighting circuit (from the PWM board)

I also have 10v going into the bottom rail of the lighting circuit.

ps. Ignore that melted connector on the PWM board. I was using a butane soldering iron and got a little too close to the exhaust.

 

[elysics]

You only say what the voltages are at 100%, correct? Have you looked at what the voltages are at 0% and 50%?

Maybe you just forgot to set the checkmark in reefpi for inverting the signal, which your circuit does. That circuit gives 0% (or 0.1V) if you input 5V, and should give 100% (10V) if you input 0V

 

[Brian71583]

You only say what the voltages are at 100%, correct? Have you looked at what the voltages are at 0% and 50%?

Maybe you just forgot to set the checkmark in reefpi for inverting the signal, which your circuit does. That circuit gives 0% (or 0.1V) if you input 5V, and should give 100% (10V) if you input 0V

I’ll check that tonight. I did verify that my 5v signal does dim but I don’t think I checked the blue header output while dimmed. Hopefully that is it.
Thanks.

 

[Ranjib]

Set the light intensity to 50% and take voltage reading at different circuit points. I know this could be frustrating but know for sure that if you are thorough you will be able to detect the issue. It will just take some time.
with 50% , you should get 2.5V at pca9685 pin, 2.5v at npn transistor base, 5V at npn transistor sink.
check and let us know where it differs..

 

[Brian71583]

We’ll I feel like an idiot but all is good Elysic’s suggestion was correct… I did not have it set to invert the signal. I didn’t see that in the guide anywhere I did not know this circuit inverts the signal.

Im using Fluval lights that aren’t readily controllable. The 10v circuit is kind of internal so once I modify them to work with my Pi there is no easily going back so Im trying to complete my Pi before mess with the lights themselves so for that reason it was never apparent I had voltage when it should be at 0. Problem would have been obvious if I plugged a light in instead of testing for voltage when I expected it.
Thank you everyone.