Any Geeks Out There?

[Shirker]

I have an old Raspberry Pi (b) that I am not using. I was thinking about hooking up some sensors for temp, ph, and possibly electrical conductivity (maybe I could correlate conductivity with FC?). I have not done much (any) research yet but wanted to check here first to see if anyone has played around with this and could offer any tips? Thanks in advance!

 

[Shirker]

Looks like PH might be more difficult/expensive than I was expecting. It would still be cool for my pool to tweet me temp and conductivity.

 

[dfahrion]

If I could find the time I really want to do something similar to monitor temp (air, skimmer, main drain, and maybe heater output temp too), pump current, and probably a light sensor in case I want to figure out how to optimize the use of sunlight for temp control. Eventually it could morph to replace my pump timer and include heater control as well. But that is as far as I'm willing to go for my DIY automation dream.

I've looked at Ph and FC and decided they are either expensive, high maintenance, or both. Conductivity usually gives you TDS, but that doesn't equate to FC, you need an ORP sensor for that.

 

[Desterline]

There are a few threads on the forum of people doing their own automation solution. I found this one interesting:
http://www.troublefreepool.com/threads/44567-DIY-with-Liquid-Chlorine-ORP-Probe-Arduino

I doubt you're going to be able to correlate conductivity to chlorine levels. But I'd love to be proved wrong. It would result in ~$5 worth of conductivity probe instead of ~$100s of dollars for an ORP probe.

In general, probes/sensors are going to be the big hurdle. They all require periodic re-calibration with solutions of known chemistry.


At the moment, I have a project in motion with a ~$!0 peristaltic pump from ebay that I'm intending to use to dispense liquid chlorine. My electronics are left over from something unrelated, but it includes a temperature sensor and a real time clock/calendar. I'm hoping I can work out a more sophisticated delivery calculation that compensates for more of the demand variables.

 

[SunnyOptimism]

Equating FC to a voltage output from a probe is very problematic. That is essentially what an ORP (oxidation reduction potential) probe does. The problem is that ORP voltage is affected by multiple variables (pH, TDS, TA, FC, etc) and so it is nearly impossible to deconvolute the FC part of the signal. This has been widely discussed on this forum if you search for the term(s) ORP Sensor.

 

[peterdaly]

I built a custom board based on arduino. http://www.poolduino.com

It runs my pump, controls my SolarAttic heating, and logs temps to a website.

Look at that for some design ideas. I hope to sell them some day, but other things are higher on my priority list right now.

 

[Desterline]

Yeah, I've actually been entertaining the possibility of setting up and automated device that actually performs the FAS/DPD chlorine test.

I'm thinking a valve (tiny) to dispense a bit of pool water into the cylinder. Some type of small metering auger type dispenser for the powder. A servo to tilt and squeeze the reagent bottle. A camera/color sensor to watch for the change (camera could also potentially count drops) and I'm thinking just set the entire sample bottle on a sensitive balance to weigh the process.

Hmmm... more thought required.

 

[SunnyOptimism]

Yeah, I've actually been entertaining the possibility of setting up and automated device that actually performs the FAS/DPD chlorine test.

I'm thinking a valve (tiny) to dispense a bit of pool water into the cylinder. Some type of small metering auger type dispenser for the powder. A servo to tilt and squeeze the reagent bottle. A camera/color sensor to watch for the change (camera could also potentially count drops) and I'm thinking just set the entire sample bottle on a sensitive balance to weigh the process.

Hmmm... more thought required.

Already been done -

http://www.thermoscientific.com/content/tfs/en/product/orion-chlorine-xp-water-quality-analyzer.html

And here -

http://www.hannainst.com/usa/prods2.cfm?id=035010&ProdCode=PCA 310

These guys also do a direct free chlorine probe (no ORP).

http://www.sbcontrol.com/chline.htm

In no way am I vouching for either. Drop based testing is all I'll ever need.


Sent from my iPhone using Tapatalk

 

[Shirker]

OK.... So the FC and PH are way more complicated and expensive than I can do right now. I am More of a software guy than electrical/mechanical. I think I will just have it tweet me the air and water temp. However it is nice to know that you guys are in this community!

 

[russenterprises]

I'm working on a Raspberry Pi based timer for my pool right now. I have a programmable peristaltic pump that I got cheap on eBay (Beta DR-2000) that I'm using for chlorine injection, but obviously need to make sure the pool pump is running when the chlorine is being injected. I was worried about clock drift between the battery backed clock in the DR-2000 and the Intermatic timer on the pool pump, and after pricing 240v digital timers, I decided to do it myself.

After looking at a lot of projects, I used a project called rasptimer as the base for my project. Right now, I'm manually editing the crontab to schedule the pump runtime. The web interface shows the current state of the pumps and allows manual override for a specified duration. I added a weather widget to the main page, and I'm planning to add water temperature as well when the waterproof temperature sensor I ordered gets here. Down the road I may add monitoring/logging of the pump current draw so I could detect a relay failure, for example. Since the interface is web based, I'm planning to mount an old Android tablet on the wall next to the door that leads out to the pool so there's a convenient place to control everything.

On the hardware side of things, I'm using 25A solid state relays (another eBay find) to handle the switching of the pumps, and the interface has logic that will turn on the main pump if someone were to turn on the Polaris booster pump when the main pump was off.

Right now, I've got the interface hardware done, it's currently switching LEDs on and off to verify that the software is working as expected, and I'm hoping to get the SSRs wired up this weekend. I'm sure I'll be tweaking the software side of things for a while.

Russ

 

[RobbieH]

If you have a 110v pump, you could do something as simple as a Belkin Wemo switch. You can program schedules on the switch from your smartphone. Or for $10 more you can get the Wemo Insight switch, and also monitor energy use. You could tell if the pump is running from your cell phone, start it from your cell phone, and even do some wild things such as if it is a stormy day, you could use IFTTT to turn the pump on to help remove debris from the pool.

 

[russenterprises]

If you have a 110v pump, you could do something as simple as a Belkin Wemo switch. You can program schedules on the switch from your smartphone. Or for $10 more you can get the Wemo Insight switch, and also monitor energy use. You could tell if the pump is running from your cell phone, start it from your cell phone, and even do some wild things such as if it is a stormy day, you could use IFTTT to turn the pump on to help remove debris from the pool.

I definitely would have had more options if there was 120v service near my pool equipment. My pumps can be wired for 120v, but there's no neutral available in the subpanel at the pool equipment, so I can't even put in a 120v receptacle without pulling more wire.

Aside from that, I just like to build stuff

 

[RobbieH]

I'm just a nerd, have been since about 1982 - at least when it comes to computers and computer control of stuff. Sometimes my projects drive my wife crazy, sometimes she likes the results. For example, we have Verizon FiOS, including TV. How many Verizon DVRs and tuners do we have? Zero. I have a couple of servers, one being Windows Home Server 2011, for storage of my media. I have a couple of Windows 7 DVRs, and a bunch of XBOX 360s (which we never play games on) as extenders. Tuning is done with a HDHomeRun Prime tuner box with a cable card. I also have an HDHR dual, for antenna programming. I have Wemo switches, and other automation **** all around the house. I was really big in X10 back in the 80s, just love to tinker.

Right now I'm setting up new servers at the house, one on Hyper V and one on ESXi. Have some things I want to do with those. Certainly not my first run at virtualization, but the first time at home. I have my own video surveillance system running on Blue Iris (I've posted some pics here before), and other little tinker stuff I do.

Oh, and I have an IT job...

 

[Desterline]

Ok, I've not actually installed it yet, but I have the $12 ebay pump running on the bench. It seems to be a bit higher volume than the Stenner pumps I've read about, It seems to be moving about 8oz per minute - about 90 gallons per day. Since my pool's been needing about 1/2 gallon per day, it seems I'll need about 8 minutes of run time per day. Obviously I'll need to be able to tune the pumping time.

As it stands, my plan is to tap a small DC power supply off the pump power (mechanical timer, currently about 3 hours morning and 3 hours evening) And write the software such that when my system powers up it waits a few minutes to allow the filter pump to stabilize, then injects chlorine for several minutes and shuts off. Set the injection time to about four minutes to accommodate my twice a day filter schedule. That seems the simplest functional implementation.

I'm trying to avoid "feature creep" and the never ending software design, but the next thoughts are:

Use a combination of non-volatile memory and the real time clock to ensure the injection only happens once a day. Preferably during the evening cycle. (prevents over injection when manually running the filter pump)
Use the clock (calendar) and/or the system temperature sensor to vary the injection times to accommodate seasonal demand changes.
User interface to inform the system when the storage tank is refilled and with what strength of chlorine. System could then calculate tank chlorine strength and adjust pump time to accommodate.
Use the clock (calendar) and/or the system temperature sensor and last fill data to estimate the aging of the chlorine to accommodate for strength loss in storage.


Thoughts?

I'm especially interested in if anyone has some insights into the seasonal/temperature demand calculations. My initial thought is below 60 degrees F there's a near zero demand, so I could just linearly reduce. Something like V*n*(T-60), where V is the highest summer demand, T is the average daily temperature and n is a normalizing term so the T-60 term becomes 1 at peak summer temperatures (probably 0.02 given peak summer temperatures of 110 F) Result should be something like this:

 

[chem geek]

The reaction rates of chlorine roughly double in speed every 13ºF so you can use that as a rough guide for the chlorine demand as a function of temperature. So rate@T = rate@85 * 2^((T-85)/13). It's an exponential curve, not a linear one.

On top of this is the amount of chlorine lost from the UV of sunlight and that does not vary with temperature but it does vary with the time of year since the sun is lower in the sky in the winter and usually there are more clouds.