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

[bishoptf]

I tested, Its not working. I know we have fixed the code , but due to some glitch it’s not being picked up in 2.3 release build. I was looking into that .., but a different situation unfolded and got busy with that

Thats priceless, you made me laugh, I remember those days, enjoy them while you can they go way to fast :)

 

[Michael Lane]

I have the ADJ-SRP8 for the AC outlets. Does anyone have a clear drawing of pin outs from the chip outputs to the DB9 male. Can not find any documentation on the SRP8. Just want to verify connections before I power up.

Thanks !

I documented the pins here https://www.reef2reef.com/threads/diy-reef-pi-build-log.458088/#post-5156444

 

[Ranjib]

I documented the pins here https://www.reef2reef.com/threads/diy-reef-pi-build-log.458088/#post-5156444

I'll add this in the docs under useful resource section

 

[That Crusso Kid]

I tested, Its not working. I know we have fixed the code , but due to some glitch it’s not being picked up in 2.3 release build. I was looking into that .., but a different situation unfolded and got busy with that

This is great but you know he's going to need more track!

 

[bishoptf]

This is great but you know he's going to need more track!

It's like this in anything you do, aquarium, trains etc, you can never have to much track, train's, railroad cars, aquariums etc....just need more room! :)

 

[Ranjib]

@Bigtrout i just tested out the 2.3 builds, i think its fixed now. Can you test it out(just download the debian file and re-install) when you have time, and let me know if its still not working.

There was two different issues intermingled.. one involving the core pca9685 driver (i2c stuff), and the other around the reef-pi<-> pca9685 integration that was modified during the HAL work,.

 

[Bigtrout]

@Bigtrout i just tested out the 2.3 builds, i think its fixed now. Can you test it out(just download the debian file and re-install) when you have time, and let me know if its still not working.

There was two different issues intermingled.. one involving the core pca9685 driver (i2c stuff), and the other around the reef-pi<-> pca9685 integration that was modified during the HAL work,.

Testing now, woohoo i think that has it...i have pwm, set it to 200hz and able to dim and brighten!! Ill do some meter testing later but it appears I can set the frequency now!!!

Many thanks!

 

[bishoptf]

Testing now, woohoo i think that has it...i have pwm, set it to 200hz and able to dim and brighten!! Ill do some meter testing later but it appears I can set the frequency now!!!

Many thanks!

Sweet! :)

 

[Ranjib]

Some macro shots from my biocube 29,

 

[That Crusso Kid]

Some macro shots from my biocube 29,

Nice!

 

[stefanm]

@Ranjib I just reached home and checked my pi, the status claimed the chiller was on, but it wasn't, it was working as expected ( the pi), though I did reboot yesterday as memory use was constantly increasing.

I rebooted and the chiller came back on, so the relay didn't switch as instructed, I use a separate PSu for the relay board.

Is there a way to analyse this?

 

[Ranjib]

@Ranjib I just reached home and checked my pi, the status claimed the chiller was on, but it wasn't, it was working as expected ( the pi), though I did reboot yesterday as memory use was constantly increasing.

I rebooted and the chiller came back on, so the relay didn't switch as instructed, I use a separate PSu for the relay board.

Is there a way to analyse this?

Next time this happens, you have to check the GPIO state, and if its on, then reef-pi did its job, and something else is wrong, if not then its a bug in reef-pi itself.
You can check a gpio state like this

cat /sys/class/gpio/gpio<N>/value

where N is the pin number of outlet thats controlling the chiller. You expect the value to be 1 (assuming the outlet is not configured as reverse) when reef-pi is expected to turn the chiller on

 

[That Crusso Kid]

@Ranjib I just reached home and checked my pi, the status claimed the chiller was on, but it wasn't, it was working as expected ( the pi), though I did reboot yesterday as memory use was constantly increasing.

I rebooted and the chiller came back on, so the relay didn't switch as instructed, I use a separate PSu for the relay board.

Is there a way to analyse this?

Happy you caught it before anything too bad happened!

 

[Bigtrout]

Next time this happens, you have to check the GPIO state, and if its on, then reef-pi did its job, and something else is wrong, if not then its a bug in reef-pi itself.
You can check a gpio state like this

cat /sys/class/gpio/gpio<N>/value

where N is the pin number of outlet thats controlling the chiller. You expect the value to be 1 (assuming the outlet is not configured as reverse) when reef-pi is expected to turn the chiller on

Adding to this, if the gpio state is on, check the pin for an output, if its on and giving an output voltage, then reef pi, and the raspberry pi are functioning correctly. That leaves your wiring and relay board.
How much power does the chiller draw? Is it motor driven or some type if big inductive load?
The relay boards used with reef pi are good for general equipment, but the relays in both the American dj and the relay boards most of us use..are NOT rated for very much on an inductive type load.

 

[Bigtrout]

I did some verification with my Fluke meter, the PCA9685 is changing frequency as its supposed to, and duty cycle is correct, and looks like the 100 percent bug is gone as well, I can set my lights at 100 percent and they work!
Well done!!!

 

[That Crusso Kid]

I did some verification with my Fluke meter, the PCA9685 is changing frequency as its supposed to, and duty cycle is correct, and looks like the 100 percent bug is gone as well, I can set my lights at 100 percent and they work!
Well done!!!

You must be stoked!

 

[Bigtrout]

You must be stoked!

Been playing around with frequencies and the auto settings...but right now the tank is dimming right down slowly, going to transition to dim moonlighting soon. Whats great is now I can run BOTH lights and dim them vs having to run one or the other on timers, overlapping to get 100 percent light for 2 hours and 50 percent for the rest of the time.
Now i can run all the white on both lights with a slow ramp up, 100 percent for 2 hours and a slow ramp back down!

 

[stefanm]

Adding to this, if the gpio state is on, check the pin for an output, if its on and giving an output voltage, then reef pi, and the raspberry pi are functioning correctly. That leaves your wiring and relay board.
How much power does the chiller draw? Is it motor driven or some type if big inductive load?
The relay boards used with reef pi are good for general equipment, but the relays in both the American dj and the relay boards most of us use..are NOT rated for very much on an inductive type load.

The load is only 6.7 amps (rated by the manufacturer, LG) at start up, the Hitachi compressor is 1100 watts @ 220~240 volts, so that's approx 5 amps, mostly our mains voltage hovers between 190-230 volts, so should be well in range of the relay, rated @10 amps 250 volts.

 

[stefanm]

Next time this happens, you have to check the GPIO state, and if its on, then reef-pi did its job, and something else is wrong, if not then its a bug in reef-pi itself.
You can check a gpio state like this

cat /sys/class/gpio/gpio<N>/value

where N is the pin number of outlet thats controlling the chiller. You expect the value to be 1 (assuming the outlet is not configured as reverse) when reef-pi is expected to turn the chiller on

Last night the timer didn't come on either for the fuge light (it was showing off status), I just tried a different schedule and it worked. It had been working properly for the past week.

 

[stefanm]

Same thing again, the chiller didn't come on, I've just done a shutdown and seems to be working normally, I played with the threshold settings to check and it's coming on.