Temp Sensor Placement

[Jon123]

I'm about to do this myself - I'm curious most about the placement of the probes. I'm planning to use the tee with the 1-1/2" slip on each end and the 1/2" NPT with one of the 50mm thermowells with the DS18B20.

The question is - does placement matter?
Let's say you have a horizontal pipe - should the T be facing such that the NPT part is on top or bottom?
Should I avoid a vertical placement?

I don't have much room to work given how the equipment is laid out. In the photo below I'm considering placing sensors in either 1 (suction side) or 2 (right after the pump, pressure side). I know I probably won't see much/any temperature variance, but I'm still tempted to put one on at position 3 to get readings on the solar return.



Currently, my valve for switching the solar on/off is manual - that's fine for now, but I'd like to get an automatic valve.

For my fellow nerds I'm planning to hook into a Sonoff TH16 (running multiple sensors in parallel) running ESPHome.

I have the GE Z-Wave switch set up via Home Assistant already and it works great - timers are set up, it auto kills the pump if you try to turn it on when it's too cold out (based on just weather API temperatures right now). I can set it to run for X hours at any time, etc.

I should clean this up a bit more, but I'll do that once I have the new sensors installed.

 

[Dirk]

Mine is in position #2. It sticks straight up, but I kinda wish it didn't. I didn't place it, a plumber did. I would have placed it such that the pipes cast a shadow on it and its wire, just to keep them out of the sun's UV rays (which doesn't seem to apply in your case), and to keep me from accidentally kicking it while stepping around or over the pipes. Other than that, you're right, it doesn't really matter temperature-wise. Water will be flowing through those pipes around 40 gallons per minute, so there will be virtually no temperature difference between a few feet of pipe.

There are a few here that have written about monitoring water temperature coming off the panels. I suppose that would be interesting, but I question the value. A properly tuned solar system output temperature is just barely warmer than input. You don't want to try to get the water as hot as possible, you want to get as much water flowing through the panels as they, and the pump, can handle. It's a matter of heat exchange physics that I won't get into, but it's kinda counterintuitive.

Anywho, the temp at output should only be a few degrees above input, so what is that other gauge going to monitor? A flow meter would offer more valuable information. Panel manufacturers generally spec the optimum flow for their panels, which balances how much heat they can deliver against the energy cost of pumping the water through them. So once you have the optimum flow rate, achieving that is more important than measuring the temperature differential.

That said, if that extra sensor and your system makes sure the output isn't zero degrees warmer than the input, and not, say, 10 degrees warmer, that might be handy to track. I can't talk, I have an entire thread here dedicated to my crazy temperature sensor adventure. I found an excuse to have eight of 'em!

 

[Dirk]

Oops, I just noticed how old this thread is. We hijacked it. Not a big deal, but sometimes others ignore new posts to old threads. You'll get more eyes on your question if you start a new one. Maybe @Newdude can move your post to a new one.

 

[Jon123]

Looks like we got moved to a new thread - sorry about that mods

Thanks, Dirk, I appreciate the feedback. I'm thinking 2 is where I'll likely go as well. I think I'll probably have it face up, just because otherwise getting to it could be a pain in that mess of pipes.

I agree, I don't think the sensor on the return is needed either, but it can act as a sanity check. Adding a flow sensor is an interesting idea - I just spent way too long looking into it. I found a flow sensor that may work, but I'm concerned about calibrating it (if I have data, I want it to be correct). Also, it maxes out at 150L (about 40 gal/sec), so if the pump is moving the water that quickly all the time I'm wondering if I'll get some bad data off the sensor.

Right now the valve is manual anyway, but the pump is automatic, so for now I plan to mostly take the roof temp and use that (along with time of day) to decide if we run the pump or not. Other data I'll collect for now and then decide if I should do anything with it.

 

[Dirk]

A FlowVis flowmeter is the way to go. It's mechanical and goes up to some number that is way more than you'd need. It also performs the function of a check valve, so you just put it in place of a check valve. If you already have a Jandy check valve in your system, you can buy just the guts of a FlowVis separately as they fit into a Jandy check valve body. If your solar is on the roof, or will be, you'll need a check valve in at least one spot, maybe two, so you could just pay the extra for a FlowVis instead of a check valve.

I've gone round'n'round with another here about how to get a solar system working well without a dedicated solar controller in the mix. I don't think it's possible. You can get the system to heat a pool without a controller, no problem, but not as efficiently. There is more to it than just turning on the system at a certain time or air temperature. A lot more. Here's the thread:

DIY solar using Sonoff

I'm trying to build my own pool automation using sonoff devices. for solar control I have the device/probe for water temp but I'm trying to implement the solar sensor part of the system (see image). Is there a device/sensor available to determine readiness of the panels to heat the water? How...

www.troublefreepool.com

I'm not a solar installer. My "expertise" comes from installing my own system, and only that one system, and from the research I did to figure out how to do it. I share what I think I know about it, and some folks make use of my mistakes, others prefer to make their own! Happy to help you any way I can, or butt out.

I saved about $4K installing my own system, so I'm all for DIY. It's not all that hard to do. The only thing you can do wrong that sometimes cannot be undone is falling off the roof!

 

[Dirk]

This is what I am referring to. They come in different sizes for different pipes. It can read over 100 gallons per minute (GPM). I don't think you actually meant "about 40 gal/sec" unless you have the Tim Allen pump!

Amazon.com : H2 Flow Controls FV-C Control FlowVis 2 x 2.5in. Complete Pool Flow Meter and Check Valve : Garden & Outdoor

Amazon.com : H2 Flow Controls FV-C Control FlowVis 2 x 2.5in. Complete Pool Flow Meter and Check Valve : Garden & Outdoor

www.amazon.com

And as I mentioned, if you already have a Jandy check valve, you can buy just the innards. The FlowVIs is a check valve, and can be used instead of one.

 

[Jon123]

Thanks Dirk. For some reason I didn't see your replies earlier. I was here posting and reading today (and funny enough, I was reading the other thread you linked to!) and came back to check this thread and saw your replies.

When I first moved in, there was no check valve on the solar, but I installed one(It's a Campbell valve, not Jandy) after reading on TFP. I had seen that flow sensor, I was hoping for one I could wire into my software though I could probably rig something up that uses vision or something with one of these and a cheap digital controller. I don't think it's that necessary though haha.

I have the solar working fairly well with Home Assistant right now. I have some simple rules that do what I need for now.
Again, I don't have an electronic valve, and I don't need to run my pump as often as an outdoor pool would - so right now I just run the pump, or I don't based on the solar.

My current rules:
1. Only run between the hours of 8am - 6pm, but don't kick on until the roof temp is over 80°F (15 minute average) - I may push that higher as the pool temp increases, or update it to calculate the on/off temp based on the latest pool water temp reading. We'll see.
2. If the temperature on the roof drops below 80 (15 minute average), turn off the pump and notify me. It checks every 15 minutes, and the pump doesn't turn back on automatically after the first start of the day - this is to prevent boomeranging. And honestly, the only times it's kicked on so far this year, I wouldn't have wanted it to turn back on, I'm sure there will be a day or two during the summer where it happens though.
3. If we get to 6pm and the roof temp is still over 80, it still turns off the pump.
4. If the pump turns on, and the weather API says it's under 53°F or the roof is under 75°F, turn the pump off and notify me (this happens fast, probably less than 1/2 a second). I lowered the weather cut off from 60 as I did have a few times at around 60°F that the roof temp was was over 80 and I consider that good enough to run.

I had started to extend the end time to 7pm, but monitoring the pool temp was showing diminishing returns later in the day, so I pared it back to the 6pm cutoff, I may go back further to 5pm, we'll see.

Honestly, this part of the project is the fun part for me, so I don't mind playing with it a bit.

The biggest issue right now is that I have a SWG and running the pump these extra hours has increased my chlorine level beyond where I want it - had to turn it off completely for a few days.

 

[Dirk]

I, too, have found that I don't make much progress heating the pool after about 6pm, even mid summer.

I also had the same SWG issue, when I wanted to run the solar heater longer occasionally, as the SWG would also be churning away and over-chlorinating my pool. I found a unique work around, but I'm not sure how you could use this trick. My pump and heater normally turn off around 4:00pm. My pool's automation controller is for pool and spa, and has two different SWG settings, one for each. But I don't have a spa. So I set my "pool mode" SWG setting to the desired output level, and the "spa mode" SWG setting to zero. When I want to heat my pool outside of the normal pump runtime schedule, say for a few extra hours after 4:00pm, I manually initiate spa mode. This doesn't change anything plumbing-wise, or change anything about how the solar heats the pool, it just continues to do so, but with the SWG set to zero output. I also set an egg timer schedule, so that this "extra heat mode" shuts itself down four hours after I initiate it.

I'll use this mode when entertaining in the evening, so that the pool cools off a little later, but without effecting the FC. Because I have a full time solar controller built in to the pool automation controller, whenever the system determines it can't get the pool any warmer (or if it's already at the thermostat setting), it'll turn off the solar heater and just idle the pump at 400 RPM. So this extra heating mode doesn't cost me anything if the pool can't get any warmer.

I guess the point is, even though you might have determined that a full-on pool automation controller is not necessary, they can be more useful than you might have imagined.

And/or maybe you can imagineer some way to simulate a similar strategy using your HA.

 

[Jon123]

I'll be honest, I've looked into the controllers a few times, but it kept not being clear what I'd really need. I can automate most things myself (and, I can't stress this part enough - I actually like this part of things, if I didn't, I wouldn't recommend it to anyone haha) - if we ever built a new pool I'd go with a controller from the start, but I'd be really picky about it haha

I'm so glad you mentioned your Spa "hack". It didn't occur to me to just automate turning the SWG on/off.
I've got a new project for the weekend (already found a switch I can use to set it up) Thanks again Dirk!

 

[Dirk]

Coolio if you got a little inspiration from my automation antics!

Regarding using HA to run your SWG on/off. Just be aware that you absolutely do not want an SWG generating chlorine gas if the water is not moving. Does your SWG have a dedicated flow switch? It most likely does. But even so, we often advise to think twice about only having one safeguard. For example, my SWG is wired to the automation circuit that tells my pump to run, and the SWG itself has a built in flow switch. So both things must be "green" before my SWG will generate. But my automation won't know if the pump actually started pumping water or not, and will happily direct the SWG to produce. In that scenario, the SWG wouldn't generate if it failed to detect flow. And even if the SWG's flow switch got stuck in "flow mode," the SWG wouldn't generate because if wouldn't get power unless the pump was also directed to run. So I've got the two safeguards, backing each other up. They could both fail, of course, but two is safer than one.

My pool automation is 20-year-old tech. They worked out the bugs and, knock on wood, it has never failed to issue commands it was supposed to. I can't say the same for my HA switches and relays. They're maybe 98% reliable, maybe even better, now that I worked out most of the bugs, but a 2% failure rate for a light bulb is one thing, for a device that can produce enough gas to blow up your pool plumbing, 2% (two days out of a hundred days) is not safe enough (IMO). Not even a fraction of that failure rate.

Just think it through, is all I'm saying. If a false command from your HA could turn on your SWG without the pump running, then the SWG's flow switch is the only safeguard. Others have worked around this by using a current sensing relay. It straps (somehow) to the wiring going to your pump, and only closes when the pump is using amps. The relay is in the loop of power going to the SWG. So whether the SWG's flow switch is functioning or not, the SWG won't get power if the current sensing relay doesn't sense the pump getting power. It's actually safer than my setup, because a current sensing relay is reacting to the actual power flowing to the pump. My automation controller only knows when my pump is supposed to be running, not when it actually is or not.

 

[Jon123]

I appreciate the concern here! My first thought was to run it on it's own leg, but I was leaning towards, and you've put me over the tipping point, of instead keeping the SWG wired to the same timer as the pump, then I'll have the HA relay on that - so the relay itself will get shut down when the timer is off.

My SWG actually doesn't have a flow sensor... So in theory, if power on the timer goes on, but the pump fails for some reason, the SWG would still turn on. Hmm. I'll have to look into this.

 

[Dirk]

My SWG actually doesn't have a flow sensor... So in theory, if power on the timer goes on, but the pump fails for some reason, the SWG would still turn on. Hmm. I'll have to look into this.

Yah, that's exactly what you need to guard against.

Shortish version: I have both an SWG and an acid injection system. The SWG has a built-in flow sensor, and the acid injection system (Pentair's IntellipH or IpH for short) can monitor the SWG's flow sensor, so neither will inject without the water flowing. They are also both connected to the same circuit that power the pump, so they both have the two safeguards I was describing. But in the winter, when the SWG won't work (due to cold water) the IpH also stops working (Pentair's safety protocol, I presume). I circumvented that with a circuit I devised so that my IpH would still dispense acid in the winter months.

But because my circuit orphans the IpH from the SWG's flow sensor, I had to recreate that safeguard another way (sometimes I follow my own advice!). So I installed a flow switch in my plumbing that operates a relay that only allows power to the IpH when there is actual flow. Dual safeguards restored! You could do something similar for your SWG.

The details of what I did are unfortunately buried in a rather long thread about the entire mod, but I could point you in the right direction if you want to pursue this particular "fix." Here's the thread, the bit about the flow switch is paragraph 7 of post #1:

Hacking IntellipH to work in cold water

As Pentair IntellipH (IpH) automatic acid dosing system owners know, the IpH depends on a Pentair IntellliChlor (IC) saltwater chlorine generator. The IpH uses the IC's power supply, and monitors its flow switch, so that it won't dispense acid if the IC is reporting low flow. Unfortunately, it...

www.troublefreepool.com

There's a wiring diagram in post #2 that illustrates how to wire the flow switch with a relay, but you'd have to extrapolate just the bit that might apply to your setup, as most of that diagram wouldn't apply to what you're doing.