Before I move on...

[Sral]

Great, that probably means that your chip wasn’t powered !

And the VCC to 3.3V was s good idea since you run your I2C directly from the PI. In that Case you should connect VCC to 3.3V indeed.

If you just have this one board, or maybe one or two more, it should still be fine to connect VCC with 5V but at some point the VCC (then at 5V) pull-up resistors on the boards will burn out the protection diodes on the PI, since it’s pins are made for 3V3 .
For explanation why: the I2C signal lines are always pulled-up (connected to) VCC with a resistor. On the Pi the signal pins are protected by diodes, so if the signal goes beyond 3.3V, the diodes start conducting this excess to the 3.3V line on the Pi.
Since the pull-ups are typically 10k, this means you’ll get less than 0.17 mA on the diodes (5V conducting to 3.3V through a 10k resistor:
(5V-3.3V)/10k Ohm = 0.17mA )

If I recall correctly, the datasheets recommends a max of 0.5mA through the protection (also called clamping) diodes.
Since you get much less than 0.17mA, you should be fine for some time.

 

[Hugo_Fish]

Great, that probably means that your chip wasn’t powered !

And the VCC to 3.3V was s good idea since you run your I2C directly from the PI. In that Case you should connect VCC to 3.3V indeed.

If you just have this one board, or maybe one or two more, it should still be fine to connect VCC with 5V but at some point the VCC (then at 5V) pull-up resistors on the boards will burn out the protection diodes on the PI, since it’s pins are made for 3V3 .
For explanation why: the I2C signal lines are always pulled-up (connected to) VCC with a resistor. On the Pi the signal pins are protected by diodes, so if the signal goes beyond 3.3V, the diodes start conducting this excess to the 3.3V line on the Pi.
Since the pull-ups are typically 10k, this means you’ll get less than 0.17 mA on the diodes (5V conducting to 3.3V through a 10k resistor:
(5V-3.3V)/10k Ohm = 0.17mA )

If I recall correctly, the datasheets recommends a max of 0.5mA through the protection (also called clamping) diodes.
Since you get much less than 0.17mA, you should be fine for some time.

 

[Sral]

Perfect, that indicates that your curcuit responds correctly on the address it should have (decimal 64).

Was that with the 3V3 connected to the VCC or the V+ ?

If that works, I would try to measure the voltage on one of the PWM pins vs GND. Watch the pin definitions, I think the board goes from 0 to 15, not quite sure right now if ReefPi counts from 0 to 15 as well or from 1 to 16. So try to measure all of them.

If that doesn't work I can try to write some code that tries to set the channels itself so youc an do furtehr testing.

Otherwise, you can always just try to switch out the board if you have several. Sometimes you just get a bad one.

 

[Hugo_Fish]

Perfect, that indicates that your curcuit responds correctly on the address it should have (decimal 64).

Was that with the 3V3 connected to the VCC or the V+ ?

If that works, I would try to measure the voltage on one of the PWM pins vs GND. Watch the pin definitions, I think the board goes from 0 to 15, not quite sure right now if ReefPi counts from 0 to 15 as well or from 1 to 16. So try to measure all of them.

If that doesn't work I can try to write some code that tries to set the channels itself so youc an do furtehr testing.

Otherwise, you can always just try to switch out the board if you have several. Sometimes you just get a bad one.

I just switched the boards out ... still have no voltage coming from the PWM to the breadboard . I plugged the 5v to vcc and left V+ unplugged

 

[Hugo_Fish]

well, what I can read is .9 mv.... but that's come from all the PWM pins to the breadboard.. I even took my multimeter directly to the pins .. and .9 mv is what I'm getting from there also .. I was also getting this same reading from the
original one

 

[Hugo_Fish]

If that doesn't work I can try to write some code that tries to set the channels itself so youc an do furtehr testing.

Otherwise, you can always just try to switch out the board if you have several. Sometimes you just get a bad one.

Could you write the code?.. I'm also going to switch the Pi's around.. its weird that all 3 of the PCA9685's are not working... I am also wondering if I didn't install I2C correctly, if that doesn't work .. format the SD and install reefpi again..

 

[Hugo_Fish]

let me also tag @Ranjib .. maybe there is something wrong that I'm doing..

 

[Hugo_Fish]

I wonder if its a software issue.. I read a old post that Reefpi 5.1 had the same issue and people downgraded to 4.1 and set the driver then upgrade to 5.1 and it worked.... Might try that as well..

 

[Sral]

Could you write the code?.. I'm also going to switch the Pi's around.. its weird that all 3 of the PCA9685's are not working... I am also wondering if I didn't install I2C correctly, if that doesn't work .. format the SD and install reefpi again..

Well, you get the answer from the prescaler, so I2C seems to work. Did you use @robsworld78’s install script from this thread ?
That should set up I2C correctly. Personally I’m using ReefPi version 6.0, and my PCA9685 worked. Another tried and tested version is 5.3

I’ll try setting up the code

 

[Hugo_Fish]

Well, you get the answer from the prescaler, so I2C seems to work. Did you use @robsworld78’s install script from this thread ?
That should set up I2C correctly. Personally I’m using ReefPi version 6.0, and my PCA9685 worked. Another tried and tested version is 5.3

I’ll try setting up the code

I did not, I did it from the guide. Thank you for sharing I will try to reinstall that way!

 

[Hugo_Fish]

! it works! It was a software issue... so for right now I'm finished with the project. my main concern now is ensuring my apartment complex will allow me to have a small tank. after which, I plan on creating the PH probe mod and then the dosing mod.

 

[Hugo_Fish]

Found my answer...

 

[Sral]

I would be more then happy ! For this version, I am playing it safe and fallowing the guide though I have made small changes. But I do want to make improvements and have an Idea of what I want to do for the next version. I would like to see what you would do to conserve space.

Okay, I was thinking to replace the 10V LM2596 board and switch it out with a 10V linear regulator like this one.

You can then replace the big 10V board for this small regulator with a 10µF cap on the board:

Here is the schematic of this:

In case somebody doesn't have a kessil and needs a smooth DC voltage instead of 10V PWM (like one needs for Nicrew lamps), you can add a few parts to the transistors to smooth the PWM out like in this thread by @bishoptf .

 

[Hugo_Fish]

! nice thank you for the write-up!! I will be setting up the tank here very soon. as such I will be reordering the light mod .. just have to wait for my stand to come in -.-..

 

[Hugo_Fish]

so I have run into a problem, it seems the intensity channel is in reverse and seems to only go to 93%.. I did have the flashing issue, which I fixed by changing the driver's feq to 1500. It also seems the light schedule that I create does not work meaning I set it, but it doesn't do it. unsure what do to. anyone have a clue?

 

[Sral]

so I have run into a problem, it seems the intensity channel is in reverse and seems to only go to 93%.. I did have the flashing issue, which I fixed by changing the driver's feq to 1500. It also seems the light schedule that I create does not work meaning I set it, but it doesn't do it. unsure what do to. anyone have a clue?

Not quite sure what you mean by the channel being in reverse and going only to 93%.

If you mean that increasing in light produces a dimming of your lamp, you can reverse that in the config by reversing the connector jacks associated.

Maybe that also fixes the scheduling ?
Try that and post a screenshot of your light schedule settings.

 

[Hugo_Fish]

like when I have channel one set to 93% (Intensity) the light is dimmed down to the lowest setting and will be fully off at 94%. I believe it should be at the highest. I did try to switch the pins from the PCA to the breadboard .. still having the issue.. It could be a PWM feq issue.. but I have the driver set to 1500.. I can not post my schedule currently as I broke one of the cables to the converter ... going to fix it when I wake up I haven't gone to sleep yet. what I can say about the light schedule is this, I have the light on a fixed 12-hour timer from 10:00 to 22:00. When I saved the schedule it was 7am. the lights should not be on.

 

[Sral]

like when I have channel one set to 93% (Intensity) the light is dimmed down to the lowest setting and will be fully off at 94%. I believe it should be at the highest. I did try to switch the pins from the PCA to the breadboard .. still having the issue.. It could be a PWM feq issue.. but I have the driver set to 1500.. I can not post my schedule currently as I broke one of the cables to the converter ... going to fix it when I wake up I haven't gone to sleep yet. what I can say about the light schedule is this, I have the light on a fixed 12-hour timer from 10:00 to 22:00. When I saved the schedule it was 7am. the lights should not be on.

Great, like mentioned before: go to the configurations tab, open the connectors and check the „reverse“ box on the intensity Jack.

That should reverse that behavior.

 

[Hugo_Fish]

Great, like mentioned before: go to the configurations tab, open the connectors and check the „reverse“ box on the intensity Jack.

That should reverse that behavior.

It should be working now... in one night I redid the whole project... what I learned is to make the Color and the Intensity into Their own jacks. In this guide, it has this

What I did was this and it seems to work.. but I believe there is a UI glitch when I set J2 to the PCA9685 it would switch back to "Raspberry Pi" even if I saved and reloaded. after a couple of tries, it works.

I just do not understand why natively the intensity was reversed. but I guess it does not matter anymore as it seems to be working. Anyway, I'm going to test the light schedule and hopefully get sleep.. God i miss that pillow .. I forget how it feels sometimes lol...

 

[Sral]

It should be working now... in one night I redid the whole project... what I learned is to make the Color and the Intensity into Their own jacks. In this guide, it has this

What I did was this and it seems to work.. but I believe there is a UI glitch when I set J2 to the PCA9685 it would switch back to "Raspberry Pi" even if I saved and reloaded. after a couple of tries, it works.

I just do not understand why natively the intensity was reversed. but I guess it does not matter anymore as it seems to be working. Anyway, I'm going to test the light schedule and hopefully get sleep.. God i miss that pillow .. I forget how it feels sometimes lol...

That sounds strange. Might be useful info for @Ranjib

Did you try that „reverse“ checkbox on the Jack(s) as well ?