Tweet
So the system I'm working on uses a Stiebel Eltron SOM 7 SI controller that unfortunately doesn't include the VBUS option. Even if it did, I think I'd still need another datalogger unit (that is crazy expensive) to be able to get data to the internet. So, I'm looking for advise on the best way to add remote monitoring. Anybody know if you can share the same PT1000 sensors with two devices (like my existing SOM 7 controller and maybe an arduino or something)?
-
-
If you keep the leads reasonably short, you can. Those are just variable resistor sensors.
Thus, the voltage at the leads on the SOM 7 will have a relationship to the temperature, which can be looked up in a table.
A raspberry Pi with an analog sensor would be super easy to program for remote viewing of data and data logging
-
Awesome, I’ll give it a go! Thx!Comment
-
brycenesbitt thx again for the suggestion! I didn't think about measuring the voltage from the controller through the sensor, I just had resistance on the brain. Anyway, turns out measuring the voltage was a perfect job for a particle photon that I had laying around, as the voltages are less than 3.3v (usually never gets over 2v). For those that may read this in the future and are bad at math like me, I used a trendline in an excel scatter chart to get the formula for converting the analog values reported by the photon over to temperature values. In my case the linear trendline seemed to fit the datapoints reasonably well (obviously the sample points were recorded by manual observation).
Correlating Datasets.PNG
So now I'm sending the temperature values to home assistant over MQTT and charting in Grafana. It's working great!
The last thing to do is to try to figure out what the SOM 7 controller is doing to calculate kWh. The manual says that it incorporates the temperature readings from the supply and return lines along with the flow rate (and I know it also asks for the glycol mixture ratio as well) to come up with kWh. Just by comparing to a simple calculation based on the on the temperature rise of the water stored in the tank (while no one is using hot water), it seems that the calc rendered by the SOM 7 is reasonable, so I'd love to be able to reproduce it.Comment
-
I'd love to be able to understand flow as well.
Anyone have a suggestion for a 3/4 copper flowmeter, easy to monitor and count electronically?Comment
-
I've designed and owned several residential solar thermal systems since 1975 or so. For direct systems I've used 4 sensors: One each at the array's inlet and outlet, and one each at the storage inlet and outlet.
If 4 seems like overkill, then there's a lot more to solar thermal or system monitoring than you may know.
For indirect systems, I've placed 2 additional sensors at the HX collector side (hot) inlet and outlet. For either type system, a way to measure flowrates is needed. Can't be done without it.
Direct systems use one rotometer for the flow to/from storage. Indirect systems need two rotometers to measure flowrates in/out of each side of the HX, or just one if it's an in tank HX.
I've always found attaching temp. sensors to a line is a cakewalk. The trick is to know where to attach it, making good thermal contact between sensor and pipe (use thermal grease), and equally or maybe even more importantly, making sure the insulation around the sensor between sensor and the ambient environment is robust and substantial while still enabling sensor checking/servicing.Comment
-
For solar thermal systems I've had good good luck with Pentair. An outfit called "Blue-White" also appears to make a suitable product. Just get a rating of +240 F. and keep the system pressure below the rotameter rated pressure and avoid buying anything with non metallic threads.
If by counting you mean how is the flowrate measured, that's done via the "GEB" system. That is: "Good Eye Ball".
The Rotameter "counting" is done by reading the level of the float inside the rotameter against a scale on the side to the device, which side is transparent so as to allow the float visible.
But I think we're not talking/thinking the same types of monitoring. I like no more complicated than necessary for the task at hand. I'm using 10 k ohm sensors and multimeters to read the output and convert a resistance to a temperature.
Rotameters are a type of flow meter that's of the direct reading type. There are no alarm conditions on my solar thermal systems.
Using simple methods,I know as much or more about my systems' capabilities than most anyone may need to know and probably a good deal more. And - a bonus - , with simplicity comes added reliability. Leaks, breaks and system upsets are minimized with good design and knowledge of what's required to achieve the desired level of safety and reliability. Lots of bells/whistles, the reasons for which are poorly or not understood are a poor substitute for good design.
The sensors are ~ $10 - $20 ea. and the rotometers run between $80 - $150 depending on flowrate range and mostly vendor. I've designed industrial flow control and monitoring systems for refineries and for power plants but a residential or similar size solar thermal system doesn't need that level of complexity. The only alarm necessary on a well designed solar thermal system (and it's not an alarm as such) is to turn on the pump in a direct system to prevent a freeze burst. There are other freeze protection methods, but they are of dubious value. I suppose I could rig a battery powered buzzer if the power failed and a freeze event was likely to occur.Comment
-
Go here for monitoring controls if you feel you can't live without it http://www.solar.imcinstruments.com/index.html.
Graphical software available.
These are excellent quality controls, from basic differential - too all the bells and whistles.Comment
-
A few thoughts. First, I can't say enough good about the Flume water monitor. Got mine for less than $100 used on eBay. It's super simple, just straps to the side of a water meter (guessing it has a hall effect sensor of sorts built in). It is accurate, and a detailed history of usage is accessible from their website and/or mobile app. So the way it's advertised, it would be a solution for monitoring "cold usage of the home" by strapping to the city's meter, but I'd guess that it will work with any meter detectible by a hall effect sensor (like possibly the meters that LucMan linked to). On Flume's website, they have a list of the meters that are compatible, though that list is probably focused on higher end meters that a municipality would install.
Second thought is that you could take the output wires of a hall sensor type meter and count the pulses with a micro-controller (converting to flow rate within your code). I used to use my particle photon for this exact task and it worked great. As I'm sure you are aware, there are a number of ways that you could store and visualize the data from there. Just for kicks, here is a pic of how I'm graphing temps now after your suggestion. Along with the temps, I'm bringing my PV system and flume data into the same InfluxDB database, so it can all be graphed together. I'm a sucker for charts anyway, but this has actually been very helpful in detecting a clogged check valve, as can be seen by the oscillating collector temps in the pic.
Solar_Thermal_Temps.PNG
Third thought is that if you are like me and just trying to track flow through your heat exchanger loop for the purpose of calculating kWh, and if your pump is a fixed speed, and if you know the flow associated with that speed (perhaps just by the eyeball type meter that J.P.M. mentioned), then all you need to do at that point would be to track the "on" time of the pump, which could again be done with a microcontroller (and and relay or current sensor or something). Incidentally, this thread looks like it might have the formulas for this calculation, though I don't yet understand it fully. https://sustainability.stackexchange...t-water-system.Comment
-
My apologies, that pic probably requires further explanation to make sense. The reason you see oscillation in the early AM, but not in the late PM is because I manually closed the ball valve (integrated into the check valve in this case) after the sun went down, in order to stop the thermosiphon. I've actually had this check valve taken apart a couple times within the last few weeks where I found and cleaned out debris that was keeping it from closing. The trouble is that I'm still fighting with those gunked up collectors that you guys were helping me with in the other thread. I did finally get the 3rd collector flow rate up to the same as the other two (and inserted back into the loop now) by cycling vinegar through it by itself with a drill pump. Not knowing any better, I had in my mind that the "gunk" clogging the collector was some sort of sludge. Turns out the material that was clogging it up reminds me more of small pieces of slate. It's rock hard, yet brittle. Anyway, those pieces keep flaking off and making their way to the check valve where they get stuck (this is even after flushing the system for what probably totals to hours altogether). I've ordered a wye strainer that I plan to install before the check valve to help me deal with the issue. Anyway, sorry for the detour, but thought you might be interested.Comment
-
I think I understand. Check valves are, for the most part, junk stops in a system and need regular cleaning, particularly after a system flush as you've done.
I've had several different check valves in my systems and none were worth much.
At one time I settled on a bypass to the plumbing similar to what's often done in industrial applications around the check valve so as to be able to remove/service the valve without shutting the system down.
My current setup uses a motorized ball valve with power passing through the differential controller; pump on = valve open. pump off = valve closed. Fit for purpose.Comment
Comment