Hello everybody,
My subject line pretty much sums it up. I recently got into trying to design and build my own CPU coolers with my own tools and materials. I've been interested in cooling for years now and I finally decided to try to make my own cooling solutions. However, my cooling solutions are nothing at all like what's on the market today and my ideas are very exotic. I really can't tell before testing them if they'll be nearly as effective as what's already on the market or not; I can make some educated estimates, but like I said, it's nothing like what's on the market today and I really feel the need to test my designs before I know if they're actually reliable or not.
I was wanting to make and test a few of these designs, but after making a 3D model of a cooler that I wanted to try to make for a CPU, I realized that in order to test this that I would need to put it on a CPU. However.. The only CPU's that I have around are all costly i7's and I don't really want to test my first exotic cooling solution on my $300+ CPU! I thought about ordering some cheap CPU's like maybe some Pentiums or Celerons... however, I then realize that my tests wouldn't be accurate, because if I use a Pentium for example, it has less cores and is (generally speaking) cooler than an i7 under load, and in the end I want to be putting my cooler on something that produces as much heat as an i7. Also, as all of my computers are in use most of the time I would need to buy other new hardware such as RAM, a PSU, and a motherboard. Overall, I feel like this would be a very costly and ineffective way to test a cooler, and as I would be testing lots of design ideas I have over the next few months/years, it's more or less inevitable that I kill the processor and have to buy a new one.
Now that all of that's been said.. I can get to my actual question and the main point of this entire thread. I was wondering if there was a cheap and easy way to emulate CPU temperatures under load, without actually buying a new CPU and investing in a new test system. I'm not 100% opposed to the idea of buying a new system, but it seems to me like it might be easy to get some ceramic and make it into the shape and size of a processor, and then maybe get a controllable heat element and try to simulate CPU heat temperatures without spending a fortune. Would something like this be possible? I want to be able to easily monitor temperatures, and I would also like the results to be similar to what I would get on an actual processor if possible. And obviously if I overheat a piece of ceramic or don't cool it properly, it won't have fatal effects on the ceramic like it would with a processor, and if it somehow did, it would be much cheaper to replace. Since I'm thinking of testing multiple designs, all of which being extremely different from what's on the market today, something like this would be invaluable to me and I was wondering if anyone knew of a way to do this? Or do I need to simply accept the fact that trying to build my own coolers is going to potentially be very costly?
Sorry for the really long post but I wanted to make everything crystal clear and help everyone understand specifically what it is I'm trying to do. I felt that way I would get the most help with the topic. Thanks to anyone who takes a look and tries to help out. I know it's not really talked about much but that's why I decided to make a thread about it here.
Thanks again!
My subject line pretty much sums it up. I recently got into trying to design and build my own CPU coolers with my own tools and materials. I've been interested in cooling for years now and I finally decided to try to make my own cooling solutions. However, my cooling solutions are nothing at all like what's on the market today and my ideas are very exotic. I really can't tell before testing them if they'll be nearly as effective as what's already on the market or not; I can make some educated estimates, but like I said, it's nothing like what's on the market today and I really feel the need to test my designs before I know if they're actually reliable or not.
I was wanting to make and test a few of these designs, but after making a 3D model of a cooler that I wanted to try to make for a CPU, I realized that in order to test this that I would need to put it on a CPU. However.. The only CPU's that I have around are all costly i7's and I don't really want to test my first exotic cooling solution on my $300+ CPU! I thought about ordering some cheap CPU's like maybe some Pentiums or Celerons... however, I then realize that my tests wouldn't be accurate, because if I use a Pentium for example, it has less cores and is (generally speaking) cooler than an i7 under load, and in the end I want to be putting my cooler on something that produces as much heat as an i7. Also, as all of my computers are in use most of the time I would need to buy other new hardware such as RAM, a PSU, and a motherboard. Overall, I feel like this would be a very costly and ineffective way to test a cooler, and as I would be testing lots of design ideas I have over the next few months/years, it's more or less inevitable that I kill the processor and have to buy a new one.
Now that all of that's been said.. I can get to my actual question and the main point of this entire thread. I was wondering if there was a cheap and easy way to emulate CPU temperatures under load, without actually buying a new CPU and investing in a new test system. I'm not 100% opposed to the idea of buying a new system, but it seems to me like it might be easy to get some ceramic and make it into the shape and size of a processor, and then maybe get a controllable heat element and try to simulate CPU heat temperatures without spending a fortune. Would something like this be possible? I want to be able to easily monitor temperatures, and I would also like the results to be similar to what I would get on an actual processor if possible. And obviously if I overheat a piece of ceramic or don't cool it properly, it won't have fatal effects on the ceramic like it would with a processor, and if it somehow did, it would be much cheaper to replace. Since I'm thinking of testing multiple designs, all of which being extremely different from what's on the market today, something like this would be invaluable to me and I was wondering if anyone knew of a way to do this? Or do I need to simply accept the fact that trying to build my own coolers is going to potentially be very costly?
Sorry for the really long post but I wanted to make everything crystal clear and help everyone understand specifically what it is I'm trying to do. I felt that way I would get the most help with the topic. Thanks to anyone who takes a look and tries to help out. I know it's not really talked about much but that's why I decided to make a thread about it here.
Thanks again!