reef-pi :: An opensource reef tank controller based on Raspberry Pi.

[b4tn]

Mine Does exactly the same thing. Now I Need to put my Dosing pump housing back together until i get a nice one Printed...

@Ranjib Looks Like a Exterminator is need But its mostly a minor issue as long as people know its there ..

Always good to know you are not crazy :)

Mine are acting strange all together since removing and re adding them. I dont have physical access to my reef-pi right now so I cant tell if they are actually running or not but they are not showing in the graph at scheduled times. It shows in the graph if I manually run the calibration but its not showing in the graph at its scheduled run time. Like I said, they may be sitting on my desk running every hour like I have them set but I cant tell because its not showing on the dash board.

 

[Erica-Renee]

Always good to know you are not crazy :)

Mine are acting strange all together since removing and re adding them. I dont have physical access to my reef-pi right now so I cant tell if they are actually running or not but they are not showing in the graph at scheduled times. It shows in the graph if I manually run the calibration but its not showing in the graph at its scheduled run time. Like I said, they may be sitting on my desk running every hour like I have them set but I cant tell because its not showing on the dash board.

I have mine back together (not running the test using multi-meter) Will set a timer up and see if its working properly.. This might be helpful to them to solve the issue. Of course unless the two of us are JUST Mindless ....

 

[Erica-Renee]

I have mine back together (not running the test using multi-meter) Will set a timer up and see if its working properly.. This might be helpful to them to solve the issue. Of course unless the two of us are JUST Mindless ....

I set them .. Day * hour * Minute * second 20 .. Duration 5 and speed 70... They run for Five seconds every one minute and speed sounds about right .
Conclusion they are Functioning Correctly.

My Graph is also updating .. Remember the graph is like the ATO Graph.. you have to watch the bar in relation to the number in seconds at the left. You should see it climb. it will only start a new bar once a hour....

 

[b4tn]

Ahhh, ok, that explains a lot!

 

[Ranjib]

@Erica-Renee @b4tn I'll look into the doser calibration issue. From the description it sounds like a bug, but let me confirm if I can reproduce it.

 

[b4tn]

Can I get some input from any of the electronics gurus in here. @Diamond1 has convinced me to add my lights to reef-pi. I have two Ocean Revive T247 lights and he has confirmed that they run off 0-3.3v PWM at 1khz. Since all my blue and whites are the same I was thinking to just run the 2 ORT247 PWM inputs in parallel and use FETS or a 2N2222A to isolate the output pins. in the crude diagram below I have two resistors forming a voltage divider to step the voltage down to 3.3v. I am not sure why the gate of the FET goes to ground through a 1k resistor but I have seen it in other diagrams. (maybe someone can tell me why its there)? Does anyone see any reason this would not work?

 

[Ryan115]

Can I get some input from any of the electronics gurus in here. @Diamond1 has convinced me to add my lights to reef-pi. I have two Ocean Revive T247 lights and he has confirmed that they run off 0-3.3v PWM at 1khz. Since all my blue and whites are the same I was thinking to just run the 2 ORT247 PWM inputs in parallel and use FETS or a 2N2222A to isolate the output pins. in the crude diagram below I have two resistors forming a voltage divider to step the voltage down to 3.3v. I am not sure why the gate of the FET goes to ground through a 1k resistor but I have seen it in other diagrams. (maybe someone can tell me why its there)? Does anyone see any reason this would not work?

Any reason you would not just run it from the two PWM channels on the pi?
Also, if you do stay with the PCA9685, you can run it off of 3.3V from the pi, then no reason for voltage dividers or the FETs, just wired straight to the pca9685 PWM outputs.

 

[b4tn]

Any reason you would not just run it from the two PWM channels on the pi?
Also, if you do stay with the PCA9685, you can run it off of 3.3V from the pi, then no reason for voltage dividers or the FETs, just wired straight to the pca9685 PWM outputs.

I am already using the built in PWM pins. Thanks for the input, I didnt realize that the PCA9685 could be run off 3.3 volts. Would there not be a need to isolate the outputs? I am not sure what the current draw is from the OR lights.

 

[Erica-Renee]

I am already using the built in PWM pins. Thanks for the input, I didnt realize that the PCA9685 could be run off 3.3 volts. Would there not be a need to isolate the outputs? I am not sure what the current draw is from the OR lights.

THE 3.3V Is the pwm signal not the Power for the lights. you still need to isolate the Higher Voltage of the lights from the lower voltage pwm. this is where the mosfits come in play at..

 

[b4tn]

THE 3.3V Is the pwm signal not the Power for the lights. you still need to isolate the Higher Voltage of the lights from the lower voltage pwm. this is where the mosfits come in play at..

I did 4 years of high school electronics but that was 25 years ago lol. Maybe my terminology is off. The lights have their own 12 volt power source so I just need to make sure the lights and pi have a common ground. My concern is that the outputs on the pi and or the PCS9685 or only capable of suppliying if I read correctly max 50mA. Even the the pwm wire is low voltage I am sure it still has a current requirement but I’m not sure what that is. I know @Diamond1 ran a single light straight of the pin and it was fine but if I add 1, 2, 3, or 4 more? That’s what I was thinking the transistor or FET was there for. So the pwm signal can use a transistor to control a higher load. Kind of like a relay.

 

[LionHeart2017]

I did 4 years of high school electronics but that was 25 years ago lol. Maybe my terminology is off. The lights have their own 12 volt power source so I just need to make sure the lights and pi have a common ground. My concern is that the outputs on the pi and or the PCS9685 or only capable of suppliying if I read correctly max 50mA. Even the the pwm wire is low voltage I am sure it still has a current requirement but I’m not sure what that is. I know @Diamond1 ran a single light straight of the pin and it was fine but if I add 1, 2, 3, or 4 more? That’s what I was thinking the transistor or FET was there for. So the pwm signal can use a transistor to control a higher load. Kind of like a relay.

I think it's there to boost the voltage on the pwm pin so you can use 0-10v pwm, without added components the output is just 3.3v from the pwm so that's why it's boosted to 10v, I'm thinking you will need to get 0-5v from the pwm pin with suitable transistor and maybe resitor.
My understanding is you will need to do each pwm output the same for each channel.

 

[LionHeart2017]

You should be able to tie the the pwm to two channels fine as I've seen a few videos where they will put 4 light panels with 4 pwm into 1 connector and 1 pwm signal in to control them

 

[b4tn]

Thanks everyone!!

 

[Ranjib]

Dosing controller guide is out: https://learn.adafruit.com/reef-pi-guide-5-dosing-controller
btw, there a bug in the dosing controller calibration logic :-/ . I'll cut a release soon with a fix.

 

[Diamond1]

I did 4 years of high school electronics but that was 25 years ago lol. Maybe my terminology is off. The lights have their own 12 volt power source so I just need to make sure the lights and pi have a common ground. My concern is that the outputs on the pi and or the PCS9685 or only capable of suppliying if I read correctly max 50mA. Even the the pwm wire is low voltage I am sure it still has a current requirement but I’m not sure what that is. I know @Diamond1 ran a single light straight of the pin and it was fine but if I add 1, 2, 3, or 4 more? That’s what I was thinking the transistor or FET was there for. So the pwm signal can use a transistor to control a higher load. Kind of like a relay.

I still have the my light fixtures apart and tested it to see if 1 gpio could control both channels at the same time and it worked without any issues. I don't see why you would need anything other than the 2 pwm signals straight from the pi and couple relays to run both fixtures.

 

[b4tn]

I still have the my light fixtures apart and tested it to see if 1 gpio could control both channels at the same time and it worked without any issues. I don't see why you would need anything other than the 2 pwm signals straight from the pi and couple relays to run both fixtures.

. Awesome. I’m going to start piecing together parts.

 

[theatrus]

Any reason you would not just run it from the two PWM channels on the pi?
Also, if you do stay with the PCA9685, you can run it off of 3.3V from the pi, then no reason for voltage dividers or the FETs, just wired straight to the pca9685 PWM outputs.

There is no reason to voltage divide for the FETs in either case - even logic level FETs perform correctly at 5V, and non-logic-level FETs won't even switch on at 3.3V.

If anything, you want a small series resistor (say, 10 ohms) in series with the FET gate.

 

[theatrus]

I did 4 years of high school electronics but that was 25 years ago lol. Maybe my terminology is off. The lights have their own 12 volt power source so I just need to make sure the lights and pi have a common ground. My concern is that the outputs on the pi and or the PCS9685 or only capable of suppliying if I read correctly max 50mA. Even the the pwm wire is low voltage I am sure it still has a current requirement but I’m not sure what that is. I know @Diamond1 ran a single light straight of the pin and it was fine but if I add 1, 2, 3, or 4 more? That’s what I was thinking the transistor or FET was there for. So the pwm signal can use a transistor to control a higher load. Kind of like a relay.

It all depends on the input of the dimmer circuit. Many things will just be logic gates or FET gates, so the overall current requirements are moot (small high current pulses at the transitions, zero in between). If anything, on long wires, you can end up with ringing so adding a series resistor is good form (and you'll not turn your wiring harness into a radio transmitter as badly).

 

[Ranjib]

There is no reason to voltage divide for the FETs in either case - even logic level FETs perform correctly at 5V, and non-logic-level FETs won't even switch on at 3.3V.

If anything, you want a small series resistor (say, 10 ohms) in series with the FET gate.

Exactly . I think this is what I have put in the light guide as well.
You need mosfets only if you are controlling more than 500ma . For less than 150ma a common npn transistor will do. It’s a good idea to put resistor in series for safety .

 

[Ranjib]

It all depends on the input of the dimmer circuit. Many things will just be logic gates or FET gates, so the overall current requirements are moot (small high current pulses at the transitions, zero in between). If anything, on long wires, you can end up with ringing so adding a series resistor is good form (and you'll not turn your wiring harness into a radio transmitter as badly).

On this topic, @theatrus , when one should go for mosfet vs a full blown meanwell driver? Or the builder has to know if the light requires constant current/voltage ?