Hey guys had a few questions concerning my diy glycol chiller build. I found a few hvac forums, but they either aren’t friendly to diy’ers/home owners and only for pros, or their forum is a clusterfuck with 28374838 sub forums and no ability to upload pics, so I figured I’d try here where there seems to be infinite knowledge. I’m building a diy glycol chiller with an old window ac unit. I have it taken apart and dismantled to where I need it. And I had a few questions about sensor/probes etc. there are 3 wires or sensors that I’m not sure their function or where to put them in my system.
The first is what appears to be a temperature probe, that sat inside a little clasp right on the pipe coming out from cooling condensor. I’m thinking it’s a temp probe to know how cold either the condensor is, or the fluid being discharged/taken in.
The sexond is what seems to be just a thin wire, but the end of his has little bristles, and it was mounted freely right in the path of the air vent where it the cold discharge air would pass it on the way out of the vent. I assume this is to measure the air temp right before coming out of the ac and into the room.
The third is the plastic thermostat that actually attaches to the grill of the condensor. Again, assuming it records temp of the condensor, and have read to keep that separate and a way to “warm it” so the ac continues to run.
I’d like to know exactly what these functions are , and how I would use them to work properly with my chiller setup. I assume if the wrong probe is reading wrong, the ac will assume it’s way hot and overwork? Or is it running at full power what I want anyway? And wondering how these sensors would be affected when submerged in cold glycol. Same applies to the other 2. I don’t know what the ac depends on to work properly. Advice?
The main one that concerns me is the one in the first photo that attaches to the coils. I’m testing the unit right now in a tub of water. I just don’t wanna burn something out because the ac is pumping and reading ambient room temp
I have AC experience. first i need some clarification. Are you sure it is the condenser (Gets hot) and not the evaporator(gets cold)?
1st- If it is the condenser the probe on the pipe is the discharge temp sensor. It monitors temp and will shut down compressor if overheats.
If evaporator it is the suction line temp used to monitor refrigerant charge.
2nd is probably the temp sensor for the system. it displays temp on the LCD screen. If you want to see the gylcol temp put this sensor in glycol.
3rd is the thermostat. this sensor will control the temp. It should be placed in the glycol solution.
You might need a sensor well also for the sensors as they do not appear waterproof.
So, your first question, you’re right. I meant the evaporator. And for sure that probe sits on what I believe is the discharge/sucking pipe. So In that case, should that probe be placed in the liquid? It’s hard to understand how these sensors will work in a different environment. If it’s reading the temp of the discharge pipe, and that pipe is submerged in glycol, makes sense that it will just read the glycol temp?
As for the second sensor I’ll test it and make sure it’s just for the display purpose.
The 3rd sensor I know measures the evaporator cool temp… but isn’t the idea for it to think it’s warmer than it is? If I’m trying to hit -20c but the temp on the ac sits at 60 wouldn’t it try to shut itself off?
I also watched a video that pertained to fixing a broken thermostat. But in the video the guy described how the evaporator coils should ice up initially but should eventually thaw itself. I currently have condensation on those coils, as well as the discharge pip and whatever it is that it runs to (black can). Any advice on it this is ok or not
I tested it in a bin of water with some
Antifreeze. Took it from 20c to 3c in 10 minutes but then everything started to ice over. Not enough antifreeze. But proved the method works which was the point of the test. I’m just wondering about the condensation and frost on those pipes. From the video I watched it may indicate a clog or issue. But I could be wrong. It chilled the hell out of the liquid tho. I’m going to try and inquire with the hvac forums.
I had another inquiry for you smarties, in reference to my little proof of concept test…
The bin i filled with fluid was mostly water, with like a 1/4 - 1/2 gallon of antifreeze. The test was more to make sure the ac even works, and if it will chill the fluid. Naturally as it got lower, a layer of ice began building up around the discharge pipe that was submerged in the water, and the whole evaporating radiator of the a/c was covered in ice, no longer allowing the cold to disperse and chill. I stirred the stuff around a little as it was chilling, but again, was simply testing the concept. So my concern now is… since glycol, antifreeze, etc all contain water, will using these fluids that have very low freezing points bypass this issue of freezing on direct contact with that chilling components? Or will it still freeze and the solution is to make sure fluid is always cycling around via pump and not just sitting idle making contact with cold components.
Why not just say screw it to the internal thermostat and just wire a pid to the flip on the condenser? That’s what I’m doing for my chiller build but it’s 240v and 50k btu for my FFE lol
My original plan was to have 2 small inline pumps. One to pump from my resevoir through my rotovap, which will indeed cycle the fluid on its own.
the other part was having another small pump, which would purely be to keep the fluid cycling in the reservoir.
Now obviously the first pump pushing the fluid to my roto, would also be naturally cycling the fluid. But i figured another small pump would be good to have incase theres a time im not running the roto. The small cycling pump is cheap, so it wouldnt be an issue i dont think. My question to that then is, regardless of how i cycle the fluid, will that solve the issue of ice building up around the discharge pipe and completely freezing over the entire evaporator condensor? As i said, my test was done with mostly water so i get why it happened. but knowing theres water in glycol and such, i want to make sure i wont have freezing problems. I assume it happened both because i was using mostly just water, as well as not really moving the liquid around, leaving stagnant water making contact with the super cold chilling components
can you explain a little more what you mean? What did you do exactly, and to achieve what? My concern is that the i will have my evaporator grill freezing over as it did today… Im fully aware that it happened because it was all water… My question now is to confirm that the water contents of the glycol mix wont end up freezing to the chilling components… The glycol mix has a nearly -50c freeze/pour point, so obviously it can hold up to temps… just confirming that the water molecules wont be freezing to things
cyclopath was able to confirm for me that as long as the fluid is moving and cycling, and not below -30c then i shouldnt have freezing issues…
I realize ive kind of derailed my own thread to avoid making 5 chiller threads.
All i have left to do is figure out what and where to place those 3 sensors. Mainly the one that sits attached to the discharge pipe, and the black thermostat that attaches right to the front of the evaporator grill. Im trying to wrap my head around what they are reading and telling the ac to do. For instance, the 1st sensor, attached to the discharge pipe, if that reads the temp of the discharge pipe, what is it telling the ac to do in response to that data? And what happens when that sensor is now submerged in a uniform temp fluid? It will now be reading that fluid temp, and in response, what will it then tell the ac to do? Not having much experience in this, im just trying to avoid making a deadly mistake. i dont want the ac to be trying to put out freezing air, while the sensor is telling it its 70f degrees
Good info! These guys are hvac guys and they use a reversing valve to give the compressor a soft start and delay timer combined with a standard inkbird pid. I’m copying them for my build
Yeah man I’m going to post a build thread on my chiller once I get the time hoping for the week after next lol I’m combining two 25000btu acs Into a custom case and frame, re doing the condensing coil so I do t have to work with the heat exchanger that’s on them, custom res, all tied into a nice little package should be awesome
