?? on Dow as water sensor probe

cablesguy

New member
Hi,

Thinking of using the DOW as a water sensor probe with the Danger Den Delrin T-Fitting, was thinking of drilling the EK G 1/4" Thread (1/4" BSPP) Plug, stick the Dow half way thru and seal it with epoxy.

Btw assuming the wires hv no ctc with water:

1) how water proof are the temp sensors?

2) will they corrode or melt or whatever with additives, eg PC ICe

What u guys think? :D Thx

Parts
Looking for additional LCD resources? Check out our LCD blog for the latest developments in LCD technology.
 

Attachments

CF Mark

Administrator
Thats should work fine.

Two things though...

Make sure you use an epoxy that will last.
Some generic 5min epoxy i have used before goes soft after long term contact with water/coolant.

Also make sure that only the sensor chip tip is in the water, and the plastic lead moulding behind it is not.
If water gets into the gap between the the sensor chip and the plastic lead moulding, it may contact the sensors leads and cause problems.
 

Heffo

New member
On my watercooling rig, I bonded a DOW temp sensor to the side of a brass elbow to avoid putting any part of the sensor in the water stream, and possibly creating a leak point if the pressure inside my closed-loop got too high and forced it's way out past the epoxy.

I also run (almost) pure glycol as coolant. I only used water to top the system up once the bubbles settled, so on the whole my coolant should be pretty electrically inert.
 

jc634

Administrator
See https://forum.crystalfontz.com/showthread.php?t=4452 for how I make my water temp sensors.

I like my method simply because I can see exactly what I am doing on both ends of the sensor. I did kill 1 once cause I had a little too much of the sensor sticking out and got water in thru the seam.

So, obviously very important to make sure you have the seam waterproofed.
 

Crystalmancer

New member
On my watercooling rig, I bonded a DOW temp sensor to the side of a brass elbow to avoid putting any part of the sensor in the water stream, and possibly creating a leak point if the pressure inside my closed-loop got too high and forced it's way out past the epoxy.

I also run (almost) pure glycol as coolant. I only used water to top the system up once the bubbles settled, so on the whole my coolant should be pretty electrically inert.

Hmmm...that sounds like a great idea. Just a couple of questions: how accurate do you think your readings are? Does any "compensation" need to be made for the fact that the sensor is not directly contacting the water? Does the brass eventually get as cool as the water flowing through it? OK, that was more than a couple. :D
 

Heffo

New member
I find the accuracy pretty reasonable, I doubt it would be out by much.. if any, so long as you don't have any fans blowing on the brass to bias the temperature, then the metal does heat to the same temp as the coolant.

The thinner the walls of the metal the more responsive the sensor will be to change, it also acts as a filter of sorts, smoothing out any rapid changes in the temperature so your fans won't speed-up and slow-down all of a sudden which can get pretty annoying.

I also found with my rig, even at 100% load in a 3D heavy game, I don't need to adjust the fan speed, granted I have dual 120mm fans on a BlackIce radiator. About the only time my fans speed up any is when I start heating my flat on a cold day.

For reference, my rig is a...
Pentium 4 HT @ 3.15GHz (150MHz OC) - Watercooled
2Gb DDR-400
GeForce FX Ti 5900 Ultra 256Mb - GPU Watercooled
Asus P4C800 Deluxe - Northbridge Watercooled
Eheim Pump 1250lt/hr
Black Ice Extreme Radiator
Automotive Glycol Coolant
No Reservoir

The coolant flows from the pump in the bottom of the case, up to the radiator in the top of the case, down to the CPU, down to the GPU, down and slightly across to the Chipset, down and back to the pump.

My average coolant temp is about 35c at full load.
 
Top