Haha, I have 3 of the kitchen cupboards and half my living room full of aquatic stuff, gonna have to get it organised and actually throw some stuff out(or buy a big box for the shed [emoji1]) missus none too happy, today I have moved my rodi from the kitchen worktop to a small cupboard, looks much nicer[emoji41]
I dont know how this work, but I would assume if we can combine multiple ATOs with macro capabilities it should do the job. Again, this is my gut feeling without reading much.
To tell you the truth, i am using these (canakit/sparkfun version with black powder coating) as well as the adafruit version (steel) and none of them has shown any rust whatsoever. Very happy with itI thought I would provide an update about the Sparkfun waterproof DS18B20 sensor. I have been using mine since July of last year (10 months). It is still in excellent condition and I have been pleased with its performance. I am thinking about picking up another one now that reef-pi supports multiple sensors
Have you looked at the UBECs at Adafruit? https://www.adafruit.com/product/1385Today I spend some time to detailed analysis of the AIO circuit. If you are following along, I am seeing some issues with light dimming where the Kessil will flickr, this happen only in the AIO circuit, i.e. when the pwm and all other circuit (ato, temperature, relay) are wired together. This does not happen when I use the Kessil controller independently. So, today I spend some time hooking up everything with a bread board and looking at signals at different hops, in different combinations of circuits. As expected , I noticed that the pwm signal has noise, particularly when used with lm2596 modules. In AIO circuit, if I power the Pi & pca9685 with an independent 5v power supply (such as the micro usb ones) thing run smooth.
Here is a screen shot of the oscilloscope when everything uses a single power supply (i.e. a single 12v supply, then step down to 10v and 5v by two lmm2595 modules)
Generated pwm signal. For refrence, pwm signals are square waves , hence the up and down. The amount of UP is equivalent to percentage dimming. Notice the spikes present in both up as well as down part of the signal
And this is the signal when I use a dedicated 5v power supply for pi/pca9685. Then the signal is much more stable, without the spikes.
I could reproduce this with both Pi based PWM as well as pca9685 based pwm.
I am reading and thinking about what should be my next step. Whether to throw in capacitors (i have to decide the value and the exact location in circuit) or use some other power regulator modules to power the AIO circuits... will keep you all posted. Whatever happens, this is a good data that i can share while discussing the problem with others more knowledgeable than me :)
I am too ashamed to show my parts box after this...mine is a large tote with all sorts of accumulated odds and ends that have accumulated over the years. I know where to go to find the parts, then the hunt begins once I am there.Some parts arrived at last[emoji3]
Let's see them parts boxes you guys must have[emoji116]
I am happy to try it , but I would like to understand whats wrong with the current setup first. One of my concern is that powering pi via GPIO is fundamentally wrong, and if this is true, then i have to rethink.
I always power them via GPIO. I always use a 12v (and in some rare cases where there is no relay or kessil control, then 5v) wall wart as single power source, then multiple lm2596 to stepdown to a 5v rail. Pi GPIO, pca9685 vcc, ATO sensors power, all are connected to this 5v rail. If i am using kessil control, i'll use another lm2596 to convert 12v to 10v.
Thats something i'll test next. i.e. power Pi using micro usb connector instead of GPIO using the same power rail. None the less.. lot more testing going on right now with oscilloscope :-)
Welcome to reef2reef.
Thats an awesome idea. This site has tonnes of good information on how to get started. It does not have to be big and elaborate. You can start a tiny tank, and test the hobby.
