Banner image

cypho's 36" DIY reef (1 Viewer)

Just finished reading back through this thread. Fun to see how things have progressed over the last few years.

And I realized I left out a significant contoller change. A while back (now probably over a year ago) I totally redid the html portion of my contoller with a lot of emphasis on pretty graphs using
http://chartjs.org I enjoyed the new graphs so much I went live with web 2.0 before building on/off override buttons. Initially I planned to add them back in. But I found that it was almost as easy for me to ssh into the pi and type the override command as it was to push the button. So I still have not gotten around to rebuilding the buttons.
dc00551d00cabe1004fd2f6c1420efae.jpg
 
Very nice, can you give some more detail and pictures of your RO/DI setup with flow meters and TDS meter with output to the Pi?
 
yeah, really sweet graphs. I am in the throws off using chartjs as well. but to graph data coming off the apex not a DIY controller. I have a service written already which is logging the apex data so that I'm not limited to the rolling 7 days of data that the apex keeps. care to share your front-end source code?
 
ikserk said:
yeah, really sweet graphs. I am in the throws off using chartjs as well. but to graph data coming off the apex not a DIY controller. I have a service written already which is logging the apex data so that I'm not limited to the rolling 7 days of data that the apex keeps. care to share your front-end source code?
Click to expand...
Sent you a PM.
 
The whole house smells like dead Acro today


26b91c195e94afa4cb1e502659c3bde7.jpg
de68bb9be61d7aa502bab643c4d65ec9.jpg
7fcbc671111e5196b2bddcbdd184bb6e.jpg
03cb350ede40f1ff0cfdae74c3bb34c1.jpg
4a8b4729f0bca029deda25c9008ece9c.jpg
c91abb7200d4f0ddeac0ccbee6387633.jpg


Sent from my Pixel using Tapatalk
 
Last edited:
What happened?
 
Sorry to see this. What happened?


Sent from my iPhone using Tapatalk
 
A bunch of seemingly unrelated hardware failures over the last few months eventually culminated in a spectacular disaster.

First the tip of the kalkwasser line clogged from precipitation. Pressure built up in the kalkwasser redactor until the maxi-jet mixing pump literally exploded. I eventually got everything pieced back together and working as good as ever. Too good actually (more on that later). But ca/alk were all over the place for about a month. Stressing the corals.

Then the light on the algae turf scrubber died (from rust do to persistent high humidity). I was eventually able to get half of the light to work, but the ATS was out of commision for quite a while. Water quality suffered. Started to get hair algae in the tank - plus more stress on the corals.

Then the sump pump died - no flow/heat for 1 night. More stress on the corals.

To keep things alive while I waited for a new pump to arrive, I borrowed the ATS pump. So that was another week with the ATS offline. Water quality deteriorated further. Algae worse. More stress on the corals.

New pump arrived and installed. But since this was the 4th Jebao pump to last less than a year, I switched things up and went with a different brand. I did not have the parts on hand to control the new pump, so the sump pump was set to constant speed. Controll over the pump shouldn't be critical, but it turns out that yesterday was a really bad day to not have a controllable pump. And for more ironic timing, the parts I needed arrived yesterday.

And then came yesterday. To be continued....
 
Around 10 AM yesterday the controller started freaking out. Even though the outlets are normally closed (if no signal is sent they default in the off position), for some reason the controller was able to turn on outlets, but once it turned them on, it was unable to turn them back off.

First the ATO stuck ON and started dumping RO water into the tank. In addition to the instant drop in in salinity came a drop in temperature and pH. The pH drop caused the kalkwasser reactor to kick into gear and once going, the controller was not able to shut it down, the pH shot up to 9.

The controller noticed the ATO and kalkwasser failures and knew what to do. Shut off the sump pump to keep the bad water out of the display. The pump control system was actually working, but the sump pump was not hooked up to the controller for reasons explained in my last post. So it even though the controller wanted to shut down the sump pump, it could not.

The controller knew exactly what the problem was, it was sending the signal to shut off the outlets, but the outlets unable to implement the command. I foresaw the possibility that the commands could get lost in transit, so after each command is sent, it checks in with the outlet and asks for a status update. If the outlet does not report the correct status, the controller will send the command again - and log the error. Although I planned for this sort of failure, I had not ever had an opportunity to test it. With 16 outlets spamming the log many times per second to complain that the commands were being ignored, the logging system crashed. So even though the controller knew there were problems I was not notified.

The pH jump was limited by the fact that the kalkwasswer reactor only stirs itself in the afternoon (to prepare itself for a night long pH boosting session) so after the first 3 gallons it was just pumping in RO water. And the salinity did not drop too much at first either because the ATO has a safety float valve which shut the ATO down after about 5 gallons. At noon the auto-water change system came online and started pumping saltwater into the tank. And again could not stop. This actually helped things a bit - at first. But also around that time, the Iron and silica pumps came on and could not stop. A little bit of those supplements go a long way - my reservoirs held 10+years worth. I suspect up to this point, the Si and Fe were actually the most harmful thing happening.

But then 4PM came around and the kalkwasser mixing pump kicked on and did not shut off. The tank pH shot up to 10.5
 
Last edited:
Any number of things would have sent me an alert right away, but since there were so many things to alert me about, it crashed and I recieved 0 alerts. I just happened to check on the tank around 2PM and I noticed the control panel was not working right. Web interface was timing out and even ssh access was sluggish.

I troubleshooted a bit during my breaks, and eventually figured out what was going on. But by them, then it had started to snow and rush hour was kicking into full swing. I left work early, but due to the traffic I did not get home until around 6PM. The pH had been over 10 for about 2 hours, but the salinity was still OK (1.021) and at this point the corals did not look too awful.

It is possibly things could have been salvaged at this point if I had done a 100% water change right then.

But I'll never know. Since the water was sooo messed up, anything less than 100% would just be a waste of water. And I had no water left. Because all of my saltwater and all of my freshwater had already been pumped into the tank. I unplugged everything so at least things were not getting worse, made sure the RODI was running, loaded up the car with buckets and headed back out into the traffic to go hunt down water. I did not have the 75 gallons of new saltwater water ready to go until 8PM. So the corals had to sit in that pH 10.5 water for 4 hours.
 
Chances are that yesterdays disaster was enough to kill even a happy tank. But after more than a month of issue after issue stressing everyone out, I was not starting with a happy tank. This morning it was clear that things were not good.

Acropora - all dead
Pavonas and leptos - Not totally dead, but looking really bad (i'd put their odds of surviving at 5% )
Pocilloporidae - three quarters dead, 1 quarter looks ok (so far)
Montipora - half dead, half look OK (so far)
Blue tang - dead
Regal angel - MIA
Flame angels - tattered fins but not dead (yet?)
Snails - All dead


The good news:

All 4 clams seem fine!!!!
crabs and shrimp - fine
mandarin dragonets & starry blenny - seem fine
Mushrooms - look fine
psammocora - does not look too bad
Anemones and LPS - not happy, but look OK (90% odds of surviving)
 
Last edited:
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top