Fetty's Cold Touch Review
By: rogerdugans

Summary:
These blocks show potential with a very different
A few short months ago I learned of someone making cpu water blocks using copper tubing instead of the more "traditional" style in use by…..virtually every water block on the market (I learned of this from a thread at the Overclockers.com Forums.) I was a bit leery of this approach but I like to experiment myself so I was willing to consider them. Besides, they looked fantastic!

Nobody had contacted the person making these blocks, so I did. Charlie was very happy to talk to someone about his blocks and in fact joined the thread where they were being discussed. Charlie also was good enough to send me one of his blocks for testing and review.

I have had some problems with my test system unfortunately, which has caused this article to be delayed about a month longer than I had intended, but it also means I have still more data and have tested the blocks with two different water pumps.

The Story:

I received the first block and I was very impressed with its appearance- on pure aesthetics I think it is the best looking block I have ever seen.



Of course when the topic is water cooling, performance comes before appearance as we all know. A block that looks good but performs at the same level as an OEM heatsink is just a fancy paperweight!

The original Fetty's Cold Touch block (pictured above) was made with about 60 inches of ¼" copper tubing soldered to a brass baseplate with an additional cap sandwiching the tube sections that touch the baseplate. When I received the block there was about 10" of straight tube left that could be carefully bent or cut as needed to suit a specific system.
The block was visually very well made and appeared to have a very flat baseplate. It was also shiny enough to make pictures difficult to take with flash! :)

The original test system consisted of:
Epox 8KHA+
Athlon XP 1700+ (Palomino)@1575
Swiftech dual fan radiator
3/8" tubing (reduced to ¼" at the block)
Eheim 1048 pump

The first thing I found when mounting the Revision1 block was that the mounting hardware was a bit lacking- two sets of slightly different sized nylon bolts/receivers for use through motherboard mounting holes- no springs, no way to really ensure adequate pressure evenly across the cpu core. I did try them out but I was not satisfied at all and replaced them with a bolt/spring set from another water block. (That change in mounting hardware has now been made for all of the Fetty's Cold Touch blocks.)

I mounted the block 5 times and tested for 24 hours per mounting to ensure accurate readings:
Net result was a c/w of .218, which is not bad, but not great either.
I also have a Koolance CPU 200g water block which is rated as a marginal block and in the same system had a c/w of .204.

So my conclusion for the first block was: acceptable but needs improvement.

As you may have gathered, Charlie had gotten ideas for some changes at Overclockers.com, and before long sent me a Revision 2 block.

Revision 2 is almost identical to the original except it is all-copper construction: brass does not transfer heat as well as copper.



Pictured above is the well-used Revision 2 water block being tested. (Note that when I received it the block was very shiny- skin oils + copper = dull, stained copper in short order: a clearcoat over the block surfaces other than the cpu contact are would help to minimize this.)

Test results: in the original test system, the all-copper Rev 2 block achieved a c/w of .198, which is slightly better than the Koolance block, but still not great. I removed the block to inspect it and found that the thermal compound was NOT evenly spread over the cpu contact area! Lapping was definitely needed for the block. I lapped with 600, 1200 and finished it off with some 1500 and re-installed the block:
c/w .175 or so. (F.C.T. blocks are now lapped before shipping.)

At that point I was satisfied enough with the performance of the block to install it in my main system and compare it's results with a better block than the CPU 200g, so I installed it in my main system:
Epox 8KHA+ (mosfet, vcore and vdimm volt mods)
XP1600+ (Pally core) @1915mhz
Eheim 1250
Chrysler heater core with 2 120mm fans
½" tubing

The other block in this test: Little River Water Block's Cascade- generally accepted as THE best block currently available.
Results:
Cascade: c/w .08
F.C.T.: c/w .152

No surprise that the Cascade wins but I had not expected the F.C.T. to be as close as it was!

I replaced the Cascade in that system as I do not normally use my main rig for testing, but I am glad I made an exception: the test rig I had set up was crippled by small tubing and a barely adequate radiator. Those results had another benefit for me: I decided that I really needed to have a test pc that was easier to work on when changing blocks, pumps, etc., so I finally got around to finishing (almost) my cube case and third water cooling setup!

My testing rig:

Big old Compaq cube server case
Fortron 350 watt PSU
Epox 8KHA+ mobo (mosfet volt mod)
87 Chevette heatercore
Enermax 120mm fan, shrouded
Maxi Jet pump (more on this soon…..:()
½" tubing

With the new test rig I installed the Rev 2 All-Copper block to begin gathering data- initial results:
c/w .159 average

I ran the system for about a month and did more than 10 block installations to ensure accuracy- 9 out of the 10 were within .002 of the others. One was way off….so I retightened the block. ;)

I then received the OCTC version of the block: Overclockers Twin Coil!
This block was a suggestion made in the thread at Overclockers.com to help improve flow rate through the block: since the cooling coils are made of ¼" tube they are pretty restrictive. Using two coils instead of one should increase the flow rate through the block- at least in theory!
This block needed some extra creativity from the block maker, Charlie, as the two separate coils needed to have the cooling coils joined at both ends so that there would be single inlet and outlets. He was up to the task happily, and sent me a block that was considerably more difficult to fabricate, but looked VERY promising!



The copper 90s are my addition so that my tubing would be able to connect to the block.
(I use copper fittings because they have a larger I.D. than other materials at a given size and therefore are less restrictive.)

This is where I began to have some trouble, sadly.
Charlie had let me know that the mounting holes on this one-off pilot block were very slightly off so extra care would be needed when mounting it- no problem, I thought.
It really would not have been a problem except that one of the nylon washers I use on the backside of the motherboard had been over-stressed by the endless block installations but I did not learn that for a few days!.
It broke and fell off! Since I was using metal bolts that became a problem as I tightened the block down! System locked up and would not boot after that. 9
This happened around 1030 PM on a weeknight and I was not able to check the system out thoroughly for a few days but I feared a cracked CPU core or fried motherboard.
That Friday evening I was able to pull apart the system and found the nylon washer pieces, and the un-insulated bolt. Amazingly the motherboard and chip were both fine and are still both installed in the system now! When I went to start the system back up I found a crack in the impeller housing on the Maxi Jet pump. More downtime while I brought out Old Reliable: JB Weld compound! I JB Welded the entire outside of the impeller housing and let the pump cure for 24 hours. Saturday night I tried once again and found that the Maxi Jet had cracked elsewhere as well, and this time someplace internally where I could not repair it. Scratch one pump. I highly recommend using any pump other than Maxi Jet after my experience with that unit.

I finally decided to be conservative and get a pump that I KNEW I could depend on- another Eheim 1250 was ordered as soon as I was able.

This final "catastrophe" was really a pita because it meant that I had to re-test at least two blocks, and since I insist on at least 12 hours per block mounting and at least 5 mounts per block and an additional minimum of 24 hours for a water change (which is needed when changing blocks) it meant at least 6 days to complete the re-testing!

NOTE: Since I nether have nor am competent to use fancy test equipment, all my testing must be done in a live system over time. My results are not truly comparable to ANY other reviews for that reason, unlike die simulator testing. That does NOT mean that my results can't be of use to others however: don't rely on my numbers, but look at any other blocks I compare to. I will try to test as many blocks as I can get my mitts on to help make these relative comparisons as easy as possible.

Final Test System

Big old Compaq cube server case
Fortron 350 watt PSU
Epox 8KHA+ mobo (mosfet volt mod)
87 Chevette heatercore
Enermax 120mm fan, shrouded
Eheim 1250 water pump
½" tubing

I began this test series with the Twin Coil block, this time mounted with nylon bolts because they are more forgiving of the slightly off mounting holes in the block.
After 3 days I swapped the OCTC for the All-Copper Rev 2 block
Results:
OCTC- c/w .158
Rev. 2- c/w .154

Test results were averaged over 3 days and 5 mountings for each block with 24 hours between blocks for the coolant to vent all air bubbles. Distilled water with no additives was used to remove that as a variable.


Conclusion:


The Revision 2, all-copper block is slightly better than the one-off OCTC prototype and provides performance that is slightly behind the forefront of High Performance water blocks. Slightly. The Fetty's Cold Touch water block does provide acceptable performance levels though, even in heavily overclocked systems: my test rig puts out approximately 90 watts of heat and the FCT Rev. 2 was able to handle it with no difficulty.

The Twin Coil version of the block did not work as well as expected, unfortunately. I am not certain why this is and I may do some more tests to check the two theories that I have, both of which are related to the fact that it is the only prototype made:
Flow reduction somewhere in the tubing soldering defects between the baseplate and copper coils.

I have no problem with recommending the Fetty's Cold Touch water block for use in a water cooled system, in fact there are two situations in which I HIGHLY recommend it:
1) Appearance has a High Priority
2) Water Damage is a MAJOR concern.

I have still seen no other block that I think looks as good as the copper coils of the FCT does, so if you want a system that performs well and is a work of art in appearance, this may be the block for you.

If you want to try water cooling but are terrified of the idea of water leaking inside your computer case the FCT may be the block for you as well: with 10" of copper tube left on the block you can easily route the tube outside the case to connect to and external cooling system. There need be no connections inside the case at all and therefore no chance of leaks!

IF you want the absolute best performing water block around, this ain't it.
But if you like the look or the safety of external water connections and don't mind losing a small amount of performance then the Fetty Cold Touch is a very good alternative.

More info on Charlie and his projects is available at Charlie's Shop.

Thanks to Charlie for his excellent work and allowing me to test his water blocks.

Sponsored Links

Temp Converter

Fahrenheit:
Celsius:

Water Cooling Items
Affiliates

Voided Warranty
Unique Hardware
CSI Gamer
MofoCases.com
The Tech Zone
Futurelooks
JFettig Machines
Extensive Mods
Black Friday Savings
My PC Rig
Hi-Tech Reviews
Root of all evil