CGameDev

So i am working on a custom unility for a 5" display panel for my case build. I am coding it in C# and so far its going well. Am going top add more stuff and make the font big that it display good on the 5". Going to add graph and temp Guage.

To answer your first paragraph. Putting the peltier on the cpu and have the loop cool it seems logical, but by logic and calculation its not practical. Take for example, your refrigerator works because when you close it, it creates a controled environment, which means there in only a temperature change when you open it or put something in. A cpu have erratice and uncontrollable temperature fluctuation and its difficult to create a control cooling environment that way. Yet cooling the loop that you know your idle and load temperture (mine is idle 24 DegC, full load 34 DegC) it is easier to calculate how much work the peltire have to do and the best possible efficency as you can see in my fiest set of graphs. To answer your second paragraph. The device that i am creating will not be replaceing the custom loop, its its an add-on to the current loop. Its like putting a tubor on your car engine, the turbo do not replace the engine, it just add more performance. When you have time read through the post because these questions where ask in i answered in details.

Thank you, i will add that to my list

I know experiment between thermal paste and thermal pads for the cold side and see which is better.

I will look into it and adjust my calculation to compensate for it base on the thickness

My bad, its Block Base Made Of Hi-Quality Copper

I think its Nickle

the water block material you are reffering to?

Do you know the RMP of you pump and how much litre per minute? and your idle and load temp and i can help you with that calculation.

If you notice my 12730 modle uses under 7V and produce less heat on the hot side and the Coefficient Of Performance in in the 90%+ range.

These are the Peltier models most youtubers use to try cool their loop. At 12 votls the dont cool well and produce too mush heat on the hot side. 180 Watts heat load is equv to 34 Deg C

I understand what you are saying. I think your theory is correct and its based on the type of peltire that is used. most person use the 12705 - 12710 that gives off too much heat to cool stuff and are designed to remove up to 89 wattage of heat. The 12730 version is design to remove 257 wattage which is 57 Deg C, once done properly i can get the hot side to be the same as ameiant temperature. The reason this is not tried is beacuse the 12730 are like $50+ a piece and to improve efficency two is needed.

I am using BitsPower 6 slot ram blocks for the water blocks. Then using spacers to connect them to the reservoir.

Connection review of the thermostat

OK, but i am not the one doing the CNC though, I have other custom reservoir in my system. The work was dont by this guy.

So I have came up with a manifold design for my motherboad and had it made. Knowing the formula VIII, IX, and X use the same dimention layout. it works perfect! can be adjusted for EK water blocks.... I wonder how much person would be interested in this?!

So I have found a way to accuratly calculate the efficiency and how to apply the correct amount of power to get a constant cooling. On full load the water is 34 Deg C constant, even for hours which is equiverlant to 180 watts of heat. In idle it is 24 Deg C which is 140 Watts of heat. with that data i use this TEC Calculater and select the 2 TEC type (I ordered the two TEC 12730) and so i generated a result. I realized now that all the youtube vidoes they use the 12706 or within that range, which can only manage 68 watts of heat max 3 Deg C of constante heat. The TEC 12730 can manage 257 watts of heat which is 58 Deg C, and I will have two, so the weight will be shared.

I will make sure, I did look at several youtube videos and I realized how some people were doing it and why it was not working or inefficient. I think the approach was wrong. I realized that the Pelteir cant be a primary cooling method in a loop, just as a turbo on a car is not a substitute for the engine. But with proper design and calculation, and a way to control the temperature, this method can work as a add-on to a current loop to achieve a desire low temperature, and this is the experiment to show that.

Rough calculation of what to expect in maintaining a constant low temperature based on my loop idle and full load temperatures

The reservoir is done up, but am waiting on the other parts in case I have to make small changes.

rough draft of the thermostat mock up connection.

The entire Chilla is going to be parallel in the loop. I may do it parallel to the GPU.