View Full Version : Help! MCW50-T SPRUNG A LEAK
Nohto
12-11-2002, 09:17 PM
I have a MCW50-T (http://www.swiftnets.com/products/mcw50-T.asp) . I have ran it for about a weak with no problems. I removed it to change 9700 cards and decided to test it before I put it on the new card. When I fired up the pump it leaked like a siv. At first I thought it was the fittings leaking, but with a closer inspection it looks like the block itself is leaking. The pelts hot side is incorporated into the waterjacket and the gaskets appear to use adhesive, making it impossible for a closer inspection without damaging them. Anyone have any problems with this item before or have any suggestions?
Nate
AKRedneck
12-11-2002, 10:04 PM
I've got an MCW40-T that leaked in the same manner. It basically leaked between the pelt and the block. My only suggestion would be to take it apart and put some gasket sealant on it. Some of the non-hardening sealant would be best I think. Afterwards I think it's best to coat all around the block with some silicone, especially where the pelt wires come out. This is something I'll be doing in the next few days to get the block back in service. After you've sealed it up, it may be a good idea to give it a small pressure test. Just put the block on the floor and run some tubes from the block up your wall, fill with water and let it sit awhile. 7ft of water column is about 3psi, which should be adequate for a leak test...:D
Nohto
12-11-2002, 10:07 PM
I don't want to take it apart because it is brand new. I definately will get a new one, but if it leaks this easily than Swiftech better come up with a better solution for sealing the 2 and also give me my money back or new gaskets if it can be sealed correctly.
KnightElite
12-11-2002, 10:51 PM
RMA the block. That's probably the best thing for you to do to make sure you get a new one that is in good working order.
Nohto
12-11-2002, 10:53 PM
I wonder if the anti-freeze effects the gasket?
AKRedneck
12-11-2002, 11:16 PM
Ok, I just sealed mine up and we'll see if it holds water. Basically the design of these blocks sucks in my opinion. The tec should be sandwiched between two plates instead of the hot side sealing against the block. All they really need to do is add one more plate and use a flat gasket for sealing against the block. Then were not relying on the tec and it's gasket/neoprene composite seal.
Nohto
12-11-2002, 11:19 PM
Actually it has its advantages, but it also has its disadvantages. What if you want to use just the waterblock? Your screwed in that case. It more than pisses me off that they didn't send the base with the block.
AKRedneck
12-11-2002, 11:26 PM
Originally posted by Nohto
Actually it has its advantages, but it also has its disadvantages. What if you want to use just the waterblock? Your screwed in that case. It more than pisses me off that they didn't send the base with the block.
I'm not sure if it has any advantages the way it is. Like you say, at the moment you can't just take the tec off and use it just as a block (unless they supplied just a flat gasket). The other problem is that your cranking down on those little screws to create clamping pressure for a seal. TEC's need some pressure, but too much and you could damage the thing. Besides, all the pressure you create is on the edges of the block. Maybe that's why I heard some 'cracking' noises as I tightened it down..:eek: Separate plates would certainly alleviate that problem..
Nohto
12-11-2002, 11:29 PM
The advantage is that it is more efficient than adding a plate. (unless the ยงยงยงยง!ng thing leaks all over the place than it kinda defeats the purpose)
Oh great. My MCW50-T is on it's way.........
Now I come across this. Perhaps the block sealing becomes defective if somehow it's overtightened. At any rate I sure hope that it doesn't have any problems. A leak can end up being catastrophic.........
I'll definitely be checking it over first. :stick:
Tedinde
12-13-2002, 10:55 PM
Mines been running fine for 4 weeks now, no problems at all. But now you have me nervous damxit
mdzcpa
12-14-2002, 07:56 AM
Same here...no problems after many weeks of use.
Guys, don't forget, the instructions clearly state that a "finger tight lock" is all that is needed to mount the block. I would assume that any more pressure than that would be a bad thing.
Either way, a finger tightening is all that is need for great performance...so there is no need to over tighten.
mdzcpa:
Either way, a finger tightening is all that is need for great performance...so there is no need to over tighten.
I'll bear that in mind. I do recall that on the site it mentioned that the springs provide up to 20lbs of clamping pressure......
Retention mechanism:
NVidia: 4 spring loaded 2-56 S/S socket screws, providing ~20 lb. of pressure overall
ATI: 2 spring loaded 4-40 S/S philips screws, providing ~17 lb. of pressure overall
Nohto
12-14-2002, 07:15 PM
Overtightning would only effect the plastic bushing or damage the card. I've been in contact with Swift. and we have entertained the possibility of the seal not holding up against extremely low temps (-17 C) and possibly the anti-freeze. I'm going to send it back and will let you know what I find out.
mdzcpa
12-14-2002, 07:33 PM
Thanks for the update nohto. Keep us posted on what develops.
Well I received my MCW50-T. For a pre fab unit it's pretty good. No point in pulling it apart to inspect as the insulation is all stuck together in a way that it's not made to really screw with.
I made certain not to overtighten the screws. Just wish there was a better way to gauge the procedure. As one person's idea of tight may differ to anothers.
Performance is pretty good as well. It actually out performed my self made unit by up to 7mhz. That was with stock voltage. Only minimal testing was performed to see at what point a lock up would occur. First test of 3Dmark was used. Which served well as a before and after exercise. I'll have to experiment with colder coolant temps and see just how far the GPU can be pushed with this setup. My original setup involved an 80 watt pelt as well. However the tec hot side was not directly exposed to the coolant as is the case with the MCW50-T.
I do believe that coolant flow has diminished slightly with the addition of the Swiftech due to being a smaller internal inlet and outlet size. Other factor is that coolant temperature may have increased by up to two degrees. Even though there are two radiators being used.
Let's just hope we don't develope any problems like some have reported...........
Entuscan
12-17-2002, 09:48 AM
Has anyone changed the pelt yet?
Swiftech states on their site that it should be possible to replace it with a different 40x40 tec.
I was thinking of buying such a block, but replacing the tec with my 172watt i have for the moment
Not yet. Will probably do so with a 176 watt I've got. If you end up doing so before I do then update us with your results..... or vice versa.:)
Entuscan
12-17-2002, 10:19 AM
Originally posted by LORD
Not yet. Will probably do so with a 176 watt I've got. If you end up doing so before I do then update us with your results..... or vice versa.:)
Can't seem to get my hands on one of those blocks here in Belgium for now.
Does anyone know of a shop that will ship it to Belgium?
Would be nice to get it the X-mas holliday period.
I can only get it late January at the earliest otherwise :(
Originally posted by Entuscan
Can't seem to get my hands on one of those blocks here in Belgium for now.
Does anyone know of a shop that will ship it to Belgium?
Would be nice to get it the X-mas holliday period.
I can only get it late January at the earliest otherwise :(
Sorry, don't know off hand. Swiftech will more than likely. However the time frame being close to the festive break may create delays. Apart from that ........:confused:
Entuscan
12-17-2002, 11:06 AM
Originally posted by LORD
Sorry, don't know off hand. Swiftech will more than likely. However the time frame being close to the festive break may create delays. Apart from that ........:confused:
Swiftech will direct me to my regular dealer I'm affraid ...