MCP350/5 top question

[knopflerbruce]

I'm trying to set up a water cooling setup for one of my folding rigs, but I need some help when it comes to the pump/res combo. I'm thinking the best solution for me is to get a res top for my MCP350/5 pump (not sure if it's a 350 or a 355, but that doesn't matter here, I think).

But... what top to get? Any input?

I got this one the other day: http://www.dangerden.com/store/external-3-x-360-radiator-and_pump-case.html - I want the pump to be placed inside it (note that I plan to run 3 loops, so it has to be placed close to the radiator I think - with the same solution for all 3 loops), to save some space. The case will be placed near the window in the picture below, cooling one of the setups on the table :)

117376

[jumper2high]

I think testing showed EK X-Tops to be one of the best.

There are several variants - for single pumps, dual pumps, I don't know if there's a dual top that allows two separate loops for the 355. The ones for the D5/MCP655 seem to have dual tops for independent loops, but I didn't see one for MCP355.

[knopflerbruce]

I think testing showed EK X-Tops to be one of the best.

There are several variants - for single pumps, dual pumps, I don't know if there's a dual top that allows two separate loops for the 355. The ones for the D5/MCP655 seem to have dual tops for independent loops, but I didn't see one for MCP355.

Here's what I had in mind, and I need a top that can do that :) (unless someone has a better idea). This is a decent route, right? Direction of water etc. (pump/res->rad->CPU block->pump/res). I guess I need a top that will let the water leave the res in the same direction as it enter. LMK if the drawing isn't good enough, I can add more arrows :p

117392

[Church]

knopflerbruce: imho any DDC top will do by size constraints, as all of them have outlet from the side and DDCs have low profile. After all, how high outlet is seems more important in your case (to be more inline with rad inlet) then where/how pump's inlet is (in most 3rd party DDC tops - from top) - as tubing coming to radbox/to pump can be placed higher as well? There is also option on placing pump on side, so that it's side outlet really is almost near the floor of radbox.

[knopflerbruce]

knopflerbruce: imho any DDC top will do by size constraints, as all of them have outlet from the side and DDCs have low profile. After all, how high outlet is seems more important in your case (to be more inline with rad inlet) then where/how pump's inlet is (in most 3rd party DDC tops - from top) - as tubing coming to radbox/to pump can be placed higher as well? There is also option on placing pump on side, so that it's side outlet really is almost near the floor of radbox.

How about this one: http://www.xs-pc.com/manual/ddcres.pdf - with a 90 degree fitting on the inlet? The outlet seems OK, I can always put something under the pump if it's too low... :)

[knopflerbruce]

Buuuump :D

[Martinm210]

Yep, that should work fine. As with any DDC pump under high flow conditions, you should consider some cooling of the pump base either via a home brew method or one of the Swiftech/Koolance heatsink methods.

[knopflerbruce]

Yep, that should work fine. As with any DDC pump under high flow conditions, you should consider some cooling of the pump base either via a home brew method or one of the Swiftech/Koolance heatsink methods.

Had a busy day today, sorry for not replying sooner. :)

I have a couple of really silent 120mm fans (they barely move any air at all, so I dont want to use them for this project), I guess they can create some air flow inside the case - you think that's enough? Without heatsink? They're $35/piece :p

[Martinm210]

I think as long as the pump base is not sitting directly on something and has air flowing over it, you'll be fine.

[knopflerbruce]

Thanks for the help so far :) Just grabbed two of those XSPC tops, now I justt need another one of those and a couple of more radiators. Got an EK coolstream today, not the best one, but I guess enough to cool an i7 970 clocked at like 4.3 ghz plus a bit of heat coming from the sun after noon, when the sunlight hits the window.

[jumper2high]

Hmm, hasn't the XSPC Top been voted the "most likely to kill a DDC" top?

[knopflerbruce]

Hmm, hasn't the XSPC Top been voted the "most likely to kill a DDC" top?

I have no clue, is this the "killer top"? http://www.performance-pcs.com/catalog/index.php?main_page=product_info&products_id=23357

Unfortunately i dont have too many options the way I have planned to set it up. Do you know why it kills the pumps?

[bmaverick]

How about the EK RES top?

[Martinm210]

Hmm, hasn't the XSPC Top been voted the "most likely to kill a DDC" top?

I ran an XSPC top for over two years straight and never had any problem, I still have that same acrylic top. IMHO, I think microscopic leaks and heat are the general problems, neither of which have been pinpointed to any one top.

There were some injection moulded tops a while back that did have cracking problems, but I haven't seen much of those lately. All top do increase stress on the motor to some degree, but not much more than speculation on anything related.

Use a heatsink and take extra care to make sure the motor is installed correctly is about all you can do. Also hook the RPM wire to the MB and shut down via RPM minimum and you'll be good to go..

[Waterlogged]

I ran an XSPC top for over two years straight and never had any problem, I still have that same acrylic top. IMHO, I think microscopic leaks and heat are the general problems, neither of which have been pinpointed to any one top.

There were some injection moulded tops a while back that did have cracking problems, but I haven't seen much of those lately. All top do increase stress on the motor to some degree, but not much more than speculation on anything related.

Use a heatsink and take extra care to make sure the motor is installed correctly is about all you can do. Also hook the RPM wire to the MB and shut down via RPM minimum and you'll be good to go..

Gabe instituted Swiftech's current RMA policy because of XSPC tops, so I think that should be given some credence.

http://www.xtremesystems.org/forums/showthread.php?259609-swiftech-pump-died&p=4562167&highlight=#post4562167

I also tried to do a poll here but a fair number of the ppl that took part in the poll refused to take it seriously (1 person went so far as to say every other top on the poll list but XSPC killed a pump) or give the info I requested so we could come up with some plausible theory as to why they were killing the pumps. At my request, one of the Admins (I think it was MM) removed the topic so the poll and info could not be taken out of context due to the shortage of proper info.

As for cause, IMO...I'd look at low restriction loops for most of the XSPC deaths. Your own tests (http://www.xtremesystems.org/forums/showthread.php?271614-Martinsliquidlab.org-DDC3.2-Pump-Heat-Scoping) show a marked rise in PCB temps with low restriction, combine that with the pump bottom being on (and likely in a low air flow area) and the added head pressure the tops generate , I think that they created such high flow rates that it forced the pumps to draw much more amperage than they were designed to draw (which created even more heat than they were designed to handle). . .I'm pretty sure that's what did most of them in. Putting a heat sink on the pump will likely lower the death toll considerably but I don't think it will end the deaths.

[Wolf132]

I've been using an xspc res top for 3 years now in my gaming rig, no problems. Knock on wood.