- Joined
- Jul 20, 2006
- Location
- houston | tx
I decided to post to the memory forum as opposed to the cooling since this is such a niche market item; Mods, feel free to relocate if it's your inclination to do so.
I've posted this to other forums but thought it work replicating here at OCForums:
...
Introduction/Explanations, etc.
---------------------------------------------------------------------
Since I'm big into keeping my hardware for a long time (at least 2 years) and since I pretty much overclock to the maximum while still being able to use my computer professionally, I decided to invest in these swank looking DDR2 heat pipe memory coolers from Thermaltake.
Initially, I wanted to buy the Thermalright HR-07 coolers but since they weren't available local and since I couldn't run them in 4x1gig configuration, I decided to simply purchase the slightly inferior (if only slightly) Thermaltake Spirit RS coolers.
Thermaltake Spirit RS (heat pipe memory coolers)
Thermaltake website: INFO
My Cost: $15.00 US (per chip)
---------------------------------------------------------------------
...
Hardware Details
---------------------------------------------------------------------
Intel E6400 @ 3.6ghz (450x8)
Mushkin Extreme 1x1gig PC2-6400 (D9GMH) @ 900mhz w/ 2.25 vDIMM
Asus P5B-Deluxe
I decided to simply test a single DDR2 chip as opposed to having to mess with both of my dual channel ram sticks; I think the temps running single or dual should be near identical.
Test Method
---------------------------------------------------------------------
To stress the RAM, I ran each of the three below test scenarios under Memtest86 for 30 minutes. I used Memtest86's "Test #5" as others informed me this would generate the most heat.
The three separate test scenarios I carried out are as follows:
1. Tested with bare IC's
2. Tested with stock Mushkin heat spreaders
3. Tested with Thermaltake heat pipe coolers
Also, as stated above, I raised the Mushkin RAM's voltages to 2.25v; the stock maximum volts recommended by Mushkin for this D9 based RAM is 2.1v.
Temps obtained via...
---------------------------------------------------------------------
To obtain temperatures off of the IC as opposed to the heat spreader surface, I used an infrared thermometer aimed directly at the memory IC's visible through the gap at the end of the heat spreaders.
FYI, at each end of the heat spreader, be it Mushkin's stock spreader or the Thermaltake coolers, there is a very clear shot at the D9GMH IC's at the very end of the chip. This worked fairly well.
No temperature data was gathered from the heat spreaders themselves, only from the IC's.
...
...
(A tethered sensor sandwiched between the IC's and the heat spreader would have been a technically more correct reading but after reading similar reviews, I think the results will be within a degree or two regardless.)
Test Results
---------------------------------------------------------------------
Bare IC: Load = 60C / Idle = 48C
Mushkin stock: Load = 51C / Idle = 40C
Thermaltake: Load = 38C / Idle = 31C
...
...
In the end, I honestly couldn't believe how well they worked! A serious drop in load and idle temps were seen once the Thermaltake coolers were installed and tested. The stock Mushkin heat spreaders worked okay, but the addition of the heat pipe and fin assembly really put the Thermaltake cooler over the edge.
Something interesting ...
Before actually putting the chip and cooler in the machine, I took a temp reading of just the heat pipe and found that it was about the same as room ambient, 25C (room ambient was 24C).
THEN, while during the stress testing phase, I took another temp reading from the same area of the heat pipe and it read 31C; meaning, of course, that the heat pipe was actually doing it's job of jetting the dispersed heat up and into the fin assembly.
In Conclusion
---------------------------------------------------------------------
Regarding temperatures, there are some better coolers out there ... among them of course, the Thermalright HR-07, which actually uses a bit different and more efficient *dual* heat pipe design. The Thermalright HR-07 coolers, however, as far as my online research told me, are designed to be used with only 2 sticks in a dual channel system.
For those wanting to use all four slots on their motherboard, the Thermaltake Spirit RS coolers would be the better option as they have a rather thin profile and as well the heat pipe can be rotated to all sorts of angles.
To be honest, I wanted to get the HR-07's first but could not find them local ... so, impulsiveness winning me over, I decided to go with the Thermaltake Spirits as a immediate gratification compromise. After testing, I can say I'm not too concerned now as to their function ... and besides, I sort of dig the black on copper look:
...
...
(this review bROUGHT to yOU by iNSOMNIA!)
I've posted this to other forums but thought it work replicating here at OCForums:
...
Introduction/Explanations, etc.
---------------------------------------------------------------------
Since I'm big into keeping my hardware for a long time (at least 2 years) and since I pretty much overclock to the maximum while still being able to use my computer professionally, I decided to invest in these swank looking DDR2 heat pipe memory coolers from Thermaltake.
Initially, I wanted to buy the Thermalright HR-07 coolers but since they weren't available local and since I couldn't run them in 4x1gig configuration, I decided to simply purchase the slightly inferior (if only slightly) Thermaltake Spirit RS coolers.
Thermaltake Spirit RS (heat pipe memory coolers)
Thermaltake website: INFO
My Cost: $15.00 US (per chip)
---------------------------------------------------------------------
...
Hardware Details
---------------------------------------------------------------------
Intel E6400 @ 3.6ghz (450x8)
Mushkin Extreme 1x1gig PC2-6400 (D9GMH) @ 900mhz w/ 2.25 vDIMM
Asus P5B-Deluxe
I decided to simply test a single DDR2 chip as opposed to having to mess with both of my dual channel ram sticks; I think the temps running single or dual should be near identical.
Test Method
---------------------------------------------------------------------
To stress the RAM, I ran each of the three below test scenarios under Memtest86 for 30 minutes. I used Memtest86's "Test #5" as others informed me this would generate the most heat.
The three separate test scenarios I carried out are as follows:
1. Tested with bare IC's
2. Tested with stock Mushkin heat spreaders
3. Tested with Thermaltake heat pipe coolers
Also, as stated above, I raised the Mushkin RAM's voltages to 2.25v; the stock maximum volts recommended by Mushkin for this D9 based RAM is 2.1v.
Temps obtained via...
---------------------------------------------------------------------
To obtain temperatures off of the IC as opposed to the heat spreader surface, I used an infrared thermometer aimed directly at the memory IC's visible through the gap at the end of the heat spreaders.
FYI, at each end of the heat spreader, be it Mushkin's stock spreader or the Thermaltake coolers, there is a very clear shot at the D9GMH IC's at the very end of the chip. This worked fairly well.
No temperature data was gathered from the heat spreaders themselves, only from the IC's.
...
...
(A tethered sensor sandwiched between the IC's and the heat spreader would have been a technically more correct reading but after reading similar reviews, I think the results will be within a degree or two regardless.)
Test Results
---------------------------------------------------------------------
Bare IC: Load = 60C / Idle = 48C
Mushkin stock: Load = 51C / Idle = 40C
Thermaltake: Load = 38C / Idle = 31C
...
...
In the end, I honestly couldn't believe how well they worked! A serious drop in load and idle temps were seen once the Thermaltake coolers were installed and tested. The stock Mushkin heat spreaders worked okay, but the addition of the heat pipe and fin assembly really put the Thermaltake cooler over the edge.
Something interesting ...
Before actually putting the chip and cooler in the machine, I took a temp reading of just the heat pipe and found that it was about the same as room ambient, 25C (room ambient was 24C).
THEN, while during the stress testing phase, I took another temp reading from the same area of the heat pipe and it read 31C; meaning, of course, that the heat pipe was actually doing it's job of jetting the dispersed heat up and into the fin assembly.
In Conclusion
---------------------------------------------------------------------
Regarding temperatures, there are some better coolers out there ... among them of course, the Thermalright HR-07, which actually uses a bit different and more efficient *dual* heat pipe design. The Thermalright HR-07 coolers, however, as far as my online research told me, are designed to be used with only 2 sticks in a dual channel system.
For those wanting to use all four slots on their motherboard, the Thermaltake Spirit RS coolers would be the better option as they have a rather thin profile and as well the heat pipe can be rotated to all sorts of angles.
To be honest, I wanted to get the HR-07's first but could not find them local ... so, impulsiveness winning me over, I decided to go with the Thermaltake Spirits as a immediate gratification compromise. After testing, I can say I'm not too concerned now as to their function ... and besides, I sort of dig the black on copper look:
...
...
(this review bROUGHT to yOU by iNSOMNIA!)
