Lee's Next Project

[lnevo]

So another vacation, 12 day's of leaving my tank completely unattended and automated with my RA.

Time to get ready for the big switchover. If all goes as expected, I will be starting tomorrow afternoon.

I will be mounting the RA*, my MP40 drivers, and the RF adapter to the door of my stand (currently pretty much everything is on that door). I will be mounting the relay expansion, relay bars, dimming, salinity, water level, power strips, and a few other devices to a pegboard. I will be using a combination of zip ties, elastic cording and pegboard hooks to make it all work. I've no idea how it will turn out. I'm still up in the air how the peg board will be situated at the back of the tank, and seriously considering how I could also mount the RA* and the MP40 drivers somewhere in the back as well. I'm hoping the stand that I got to hold the ATO reservoir will be strong enough or can be modified with some minor reinforcement. All the unknowns... always the fun part

Anyway, wish me luck.

[binder]

good luck!


Sent from my iPad using Tapatalk

[lnevo3]

Well, I started typing out a response last night as I was quickly crashing ~5am. Anyway, even though I didn't post it, I did catch my second wind. It was grueling. I kept the tank running until around 9am on just the return pump. Everything is back together now. I've been making code changes through the web wizard all day, getting things tweaked as they should be. I'm pooped though. A few things, I would hope have improved but have not... always had some stray voltage affecting pH. It seems it's a bit worse now. Also, was hoping the salinity probe would have improved, but seeing the same behavior with a new probe and now with everything re-installed. Water level downstairs is being wonky, but I'm sure I'll get that straightened up soon. Still a bit of tweaks to be done. I'll post some pictures soon.

[lnevo]

I had to fabricate that because I didnt have enough room with the mp40 drivers. Thats just a pvc plastic sheet cut to fit the profile of the gap between the wall and stand. i should have mounted it a little lower. Everything tucked neatly behind the stand. Its a pain to get to and if I need to take any modules off it will be a pain putting it back on. But nothing like the pain I had before if I needed to get at an outlet.

[lnevo]

The best part is just the head unit on my door.

[binder]

nice. crazy setup and wire mess initially.
you'll get things figured out once you get some rest.


Sent from my iPad using Tapatalk

[lnevo3]

Some really cool features I have now. D-link smart plug attached to the GL-MT300N. The GL-MT300N is a wireless router running OpenWRT (basically a tiny linux box). I can ssh to this device, run minicom to watch the serial port and see what the RA is spitting out. I can also avrdude a binary if I want to update the firmware without the web wizard. I've got a cron job on the unit that looks in a folder for new firmwares and a shell script on my mac when I'm done compiling with Arduino to upload the file. It's a pretty smooth process

Did I mention how cool it looks with just this on the door

[lnevo]

So, things have been running smoothly for the past few days. I had an issue with my lights because I put the labels on them backward and swapped the plugs without changing the code when I should have just changed the labels. The pH seems to be tied to a bad float switch. It dropped to 1.0 after pulling the ato high out the water leaving atolow in. If I took both out it was fine. i guess I have to change the low one. The other good news is I think the salinity has seemed to be impacted so we will see soon. There does seem to be an issue, maybe with the temp conversion, but as soon as I calibrate, the value is at 33 instead of 35.

Also, the water level seems to have stabilized as well for my channels 1&2, but I still have an issue with channel 3 (RO 5gl bucket). I'm monitoring the serial and the values being read by the module seem good. Something is happening in the conversion so I'll be trying to see what is breaking there.

A lot of the confusion I was having with switches btw was caused by a bug. The star has a 3rd input and the function I use to switch the behavior of the relays in software was applied to the 3rd input (AlarmInput) in addition to LowATO. That drove me insane for half a night.

Anyway close to closing all the gaps. Nothing's dead yet so it's definitely a success.

[lnevo3]

Ok, salinity calibration seems to be skipping setting the temperature. When I checked the memory location it was set to 750 which is 75 degrees. I set it to 80.6 which is what it was at when I calibrated and now the value is 34.7. I'll open an issue for that, but won't be focusing on that at the moment

[lnevo3]

Ok, found the issue with my 3rd water level sensor:

See if you can spot the issue

Code: Select all

  memory_write_int(Mem_I_WaterLevel1Min,387);
  memory_write_int(Mem_I_WaterLevel2Min,387);
  memory_write_int(Mem_I_WaterLevel3Min,621);
  memory_write_int(Mem_I_WaterLevel1Max,2200);
  memory_write_int(Mem_I_WaterLevel2Max,2325);
  memory_write_int(Mem_I_WaterLevel2Max,1923);

For some reason to be determined, the star is not updating the calibration values over to the cloud hub. So in the meantime, I'm setting it manually based on the values I recorded.

[lnevo3]

For the PH, I've tried changing each switch with a fresh one and I have the same issues. I'm leaving the switches out of the tank for now and want to see how we track for a few days. There is some evidence that the ph probe may have started dying, but it's not enough data as if it did, it looks like the ph was dropping from 7.89 on Wednesday afternoon when we got home, never came up as it would during it's normal cycle during the day. It dropped to 7.22 by Thursday morning, and only reached a peak of 7.6 on Thursday off from a usual ~8.4 at the top of the cycle. The other odd thing is that while we were away, ph usually goes up when no one is home expelling CO2. While the PH rose initially when we left, by the time we got home it seems like it had dropped a few points during the week. Anyway, I've hacked the calibration to give me a good number. Let's see how it behaves without the switches in. If it continues to drop, I'll get a new one.

[rimai]

Where was this code?
Is it something you added to the hub firmware or something existing on the libraries?
Are you seeing WLC3:xxx on the cloud out channel?
If you are, are you seeing "WL calibration" on the serial log dump of the cloud hub?

[lnevo]

Oh it was my code added because its not recording on the cloud hub. Will need to do some more tests, its hard now with the controller upstairs and the sensor in the basement

[rimai]

Everything to be monitored can be done on your laptop, can't you?
Start calibration on RA*.
Then on your laptop, launch mosquitto to monitor cloud traffic and minicom for hub serial, right?

[lnevo]

Yes, was focused on getting everything working

[lnevo]

No laptop though actually