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

[AbjectMaelstroM]

Hi All, just wanted to throw out a reef-pi idea for you. I have figured out a way to use a non-contact optical water level sensor with my Skimmer to detect overflow. I use the ATO feature and invert the skimmer output. so when the sensor sees water it will turn the skimmer off. Here is the sensor I used - it is much smaller in person - just need to add a pullup for the NPN signal to work. I only wish I could tell reefpi to modify the on and off color bars since it is inverted in this case. I cant imagine I'm the only one who fears the occasional skimmer overflow lol

It doesn't get tripped by foam? Just liquid?

I've tried a different "shape" contactless sensor for the same thing before as I cant fit a float in the cup, but it was getting tripped up by tick foam so yanked it. Still works great on my ATO container as a refill reminder.

 

[ReeferLou]

It doesn't get tripped by foam? Just liquid?

I've tried a different "shape" contactless sensor for the same thing before as I cant fit a float in the cup, but it was getting tripped up by tick foam so yanked it. Still works great on my ATO container as a refill reminder.

There is a sensitivity adjustment on the sensor that I am using. I will say I have not testing it extensively yet, so it may end up with issues. But with initial testing it seems to work. Here is a pic of the setup:

 

[brian.badge]

Tonight I noticed my fuge light was stuck on despite being on a timer. The outlet no longer responds to reef pi behaving like the relay is stuck on. Testing other outlets I see 4 total that have issues and the rest are ok. I assume either the relay has failed or something has happened in the wiring. Any bright ideas? Now that the controller is online I don't want to spend a lot of time troubleshooting .

 

[stefanm]

Tonight I noticed my fuge light was stuck on despite being on a timer. The outlet no longer responds to reef pi behaving like the relay is stuck on. Testing other outlets I see 4 total that have issues and the rest are ok. I assume either the relay has failed or something has happened in the wiring. Any bright ideas? Now that the controller is online I don't want to spend a lot of time troubleshooting .

Seems like an issue with the PWM connections, are you using plug in jumper wires by any chance?

 

[brian.badge]

Seems like an issue with the PWM connections, are you using plug in jumper wires by any chance?

I do have jumpers that slide over pins. Are suggesting that they may have come off?

 

[Des Westcott]

Seems like an issue with the PWM connections, are you using plug in jumper wires by any chance?

I have had quite a few of the "Cheap Chinese" relays get stuck - particularly when the relay controls equipment that requires a transformer - like DC wavemakers

 

[Des Westcott]

Does anyone have an idea why the temp graphs do this on occasion?

 

[Mikeneedsahobby]

Does anyone have an idea why the temp graphs do this on occasion?

I haven’t seen that response before.

 

[Mikeneedsahobby]

I do have jumpers that slide over pins. Are suggesting that they may have come off?

Use a multimeter to follow the signal and find where the problem is. This is most likely a wiring problem.

you should get a switchable 3.3v from the pi to the uln2803 chip. After the uln2803 you should get 5 or 12v on the signal wire to the relay. Check the return path as well.

 

[brian.badge]

I have had quite a few of the "Cheap Chinese" relays get stuck - particularly when the relay controls equipment that requires a transformer - like DC wavemakers

Any suggestions for better quality relays assuming that turns out to be the issue?

 

[Des Westcott]

Any suggestions for better quality relays assuming that turns out to be the issue?

I'm now using the Power bars from Robo-Tank and @robsworld78 uses Omron relays in them.

You can also build a simple snubber circuit to protect the relays.

 

[AbjectMaelstroM]

There is a sensitivity adjustment on the sensor that I am using. I will say I have not testing it extensively yet, so it may end up with issues. But with initial testing it seems to work. Here is a pic of the setup:

Where did you buy it from? I'm game to try it. The one I had is from DFRobot and it also had a 3-level adjustment via push-button on a board, unfortunately, even on the lowest setting it would trip if the foam got thick.

 

[AbjectMaelstroM]

Any suggestions for better quality relays assuming that turns out to be the issue?

Legit Omron (buy from Digikey, Arrow, etc) relays are probably the best you'll get without going to a solid state. Adding a snubber helps with eating the spikes when relay cycles (especially when DC equipment is turned off).

 

[brian.badge]

Legit Omron (buy from Digikey, Arrow, etc) relays are probably the best you'll get without going to a solid state. Adding a snubber helps with eating the spikes when relay cycles (especially when DC equipment is

Legit Omron (buy from Digikey, Arrow, etc) relays are probably the best you'll get without going to a solid state. Adding a snubber helps with eating the spikes when relay cycles (especially when DC equipment is turned off).

My current controller setup is leviathan board and enclosure. Looking at the relays I have now vs those at grainger I may be a bit over my head.

 

[ReeferLou]

Any suggestions for better quality relays assuming that turns out to be the issue?

Make sure you have snubber diodes on the relays - could extend the life.

 

[stefanm]

Any suggestions for better quality relays assuming that turns out to be the issue?

Try tapping the relays with a screwdriver handle, if they click then they're stuck. I'm going to switch over to solid state relays, I had a relay board fail last year, I bought a made in India board which is nice but two major flaws firstly the holes for standoffs are way too tiny and secondly the PCB thickness is 1mm or less very flimsy.

 

[stefanm]

Does anyone have an idea why the temp graphs do this on occasion?

This happens to me when it's been powered off, it takes about 6 hours to sort itself out, it seems to have new data over old data in the graph but works properly from version 3.4 onwards only.

 

[Des Westcott]

This happens to me when it's been powered off, it takes about 6 hours to sort itself out, it seems to have new data over old data in the graph but works properly from version 3.4 onwards only.

That makes sense to what it's doing then. Although I don't recall it losing power around that time. Might have been a bad dip that caused it.

I'm on version 4.1

 

[Des Westcott]

Anybody?

I run 2 x 500W heaters in my cube (just under 1000l total volume). For a while I just had them connected together and reef-pi treated them as one heater, I noticed that when they had to run, typically, they got the tank up to temp pretty quickly. The tank is in a non-climate controlled house in a sub tropical climate

What I've decided to do now is to separate them onto two separate temp monitors - using the same temp sensor in the sump. I've called them Temp1 and Temp2 (original, I know). Temp1 heater control is set at 25.5degC with a 0.2 hysterisis and Temp2 is set at 25.2 with a 0.2 hysterisis.

he thinking behind this is that on days where the ambient brings down the tank temp slowly, it heats up slowly and the days when it is colder, the extra horsepower of the 2 x heaters is available. I relaise in terms of the basic energy equation, that it takes the same amount of energy input to get a given volume of liquid from one temperature to another, so this isn't an energy saving measure, it's more about smoothing the curve of the temperature fluctuations.

As far as your issue with the boot-up time on the heaters you use...... What does that actually matter? Does the boot-up process require some human input, or does it run it's course automatically? To me, it wouldn't matter if there was a 1, 2 or even 3 minute delay between the controller turning the heater on and the heating actually starting.

I also like the Reef-Pi controlling the temperature and the built-it thermostat in the heater acting as the failsafe. Biggest benefit to me is the heater / chiller usage stats. It's helped me catch a faulty relay before.

 

[Ranjib]

For my pico tanks i use one two heaters both controlled by reef-pi. the hydor 50w one is set at 78 with reef-pi set at 79.5, hence its always on. The second one, (hygger??) is the one actually controlled (on/off) by reef-pi, at 79.5, and its own thermostat set at 81. This is due to my learning that the hydor by itself is not sufficient and tank was dropping down to 76 when the ambient temperature goes below 54 (tank is in my workshop). And the workshop does not have hvac, it gets 54 when ambient temperature dips below 32/31. From late December, i am using the ambient temperature to control a 800w dish heater via a kasa hs100 to keep workshow above 55. Its required only at late night to early morning, right before the sun comes out.


For the tank, all equipment are connected to hs300 power strip, and i can use the current monitoring data for the corresponding outlet (proxied as a ph probe) as an indicator of when /how frequently its turning on/off. During winter time its mostly on. In summer its opposite, the extra heater is replaced with fan and reef-pi starts to control both hydor heater as well as the fan. My tanks still hit 84/86F couple of times early in the season, later i built a evaporative cooler (controlled by reef-pi) that addressed it. Im still thinking of a better strategy for this year (to deal with the heat waves)