| Electronics > Projects, Designs, and Technical Stuff |
| How to add a thermocouple to hot water tank |
| << < (2/2) |
| james_s:
That's a good point, I forgot about that. I do wonder what the value of this is, the plain old bimetallic thermostats seem to work pretty well, I never had an issue with them back when I had electric hot water. |
| mikerj:
--- Quote from: Brutte on October 22, 2019, 07:20:08 pm ---The water changes density with temperature so if your tank is slim there should be a significant difference in temperatures along its height. It might be tricky to make that work satisfactory with PID (especially with D). --- End quote --- Yep, typically a bimetallic thermostat for a cylindrical hot water tank will be mounted about 1/4 to 1/3 up from the bottom of the tank. |
| james_s:
Most of the tanks I've seen have two elements each with their own thermostat as I recall, one lower and one upper. It's been a while since I've dealt with one though, my current house has always had gas. |
| Brutte:
--- Quote from: james_s on October 22, 2019, 08:55:15 pm ---the plain old bimetallic thermostats seem to work pretty well, I never had an issue with them back when I had electric hot water. --- End quote --- One can imagine a plethora of control laws based on the current and/or the past state of all the parameters in a heating system. Zillions or more of them out there. The on/off based on temperature at some distinct point in tank is just one way to skin a cat. But you can have a control law also based on the past of the temperature (like I in PID). For D you can make the drive more agile. So, simplifying, a controller with D will give you similar performance with smaller tank than the one without D. Once it will sense the change, it starts heating instead of waiting till the temperature hits the low temperature threshold. I think the observer-based controller could give even better perfomance. If there is a cold water flowing into the tank (based on flow meter) then it is a clear sign we need to pump in some energy and not wait till the temperature drops down. Then even smaller tank would do the job. |
| beanflying:
A fairly common option in Industry is to use a Thermowell. That way you can remove the issue of electrolysis or chemical attack with the sensor. The issue with wanting to PID ( I assume sealed?) a boiler is the thermal gradient and the general lag of that closed system to respond to a heat input. You will still need to run the PID fairly loose if it is a typical tall boiler. So not really a lot to be gained :-// If it is an open vessel at or near boiling like an Urn or Kettle then there is generally move vertical movement in the generally shorter water column so it can be held tighter. I run a small Urn for Brewed Coffee that will hold very close to a set point at the outlet. |
| Navigation |
| Message Index |
| Previous page |