But it wouldn't be ambient. Provided you have a sufficiently large body of water, Like a bathtub for example, would the water not exert significant pressure on the block? It would just be from the outside in instead of from the in side out.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
Production Aquanaut Basic/Extreme - Ultra Low Profile CPU Block & Pump Mount Combo
- Thread starter Nouvolo
- Start date
You are correct in that outside-in vs inside-out would be fundamentally the same, but you'd need a really deep tank to get any appreciable pressure. 30ft of depth would get you ~13psi (~1 bar). This would probably be the absolute minimum pressure I'd be willing to certify to in order to provide some margin as a factor of safety.
My understanding of the test procedure is similar to someone has said, blanked off all ports except for one, then connect to a pneumatic source, pressurise the block to whatever spec the manufacturer is comfortable with, usually the max proof load multiplied by a safety factor, then continuously monitor pressure over some period of time. If there are any appreciable pressure drops, this usually means there's a leak somewhere. This is definitely possible on a manually operated small batch basis, albeit labour intensive. You could follow a similar procedure (blank off all available ports except one, then connect a hand pump with a pressure gauge and pump, then monitor the pressure gauge.
Definitely tedious, but it's an useful sanity check (and this check is often encouraged for people building their own loop to do as well, before they start filling the loop with actual liquid)
Definitely tedious, but it's an useful sanity check (and this check is often encouraged for people building their own loop to do as well, before they start filling the loop with actual liquid)
Development Update
I have ordered about 20 units for the demo (MOQ?). But still need to test for flow rate, leakage, actual mounting, etc. Thinking of offering about a dozen of them out (if they turn out fine). Let me know if anybody is interested. No rush tho, will post more photos when the blocks arrive.
I have ordered about 20 units for the demo (MOQ?). But still need to test for flow rate, leakage, actual mounting, etc. Thinking of offering about a dozen of them out (if they turn out fine). Let me know if anybody is interested. No rush tho, will post more photos when the blocks arrive.
- Aug 2020 - CNC milling of blocks (demo run?) small quantities. DDC first, SPC later
- Mid-end of Aug 2020 - Flow rate, leakage, mounting tests
- cold plate - ready
- AM4/Intel 115x mount - ready
- O-rings - ready
- screws - ready
- packaging - ready (kind of?)
Last edited:
Definitely interested in giving a custom water loop a go with one.
BTW, just want to pick the brains of sff community. Do you think SPC is going to work (before I do the actual tests)?
The max flow rates quoted for the following pumps are:
Note: Based on what I have read in different forums, cooling performance tails off < 1gpm. And closed loop AIO pumps are rated at around 0.3-0.5 gpm, and all of those are for CPU cooling only.
The max flow rates quoted for the following pumps are:
- (EK) DDC - 1000 L/hr , 16.7 lpm, 4.4 gpm
- (EK) SPC - 250 L/hr , 4.2 lpm, 1.1gpm
Note: Based on what I have read in different forums, cooling performance tails off < 1gpm. And closed loop AIO pumps are rated at around 0.3-0.5 gpm, and all of those are for CPU cooling only.
I think you should watch the OT channel https://www.youtube.com/c/OptimumTech/search?query=pump as the author has done quite a lot of tests with different pumps.
And his review of the Alphacool DC-LT 2600 was quite optimistic, but the loop configuration was not very big there, and that pump even weaker than SPC pump with maximum flow rate: approx. 100l/h and pressure head at 12V: 0,6m H2O.
But my personal opinion is that you need to do actual tests. It's always better than read the tea leaves .
You can drop SPC support altogether, no one really cares about it.
It is also discontinued already, so really there is no point.
DDC support is more than enough.
It is also discontinued already, so really there is no point.
DDC support is more than enough.
Tending to think the same thing as @srekal34. I hadn’t heard of SPC until this thread, and given its relative lack of deployment (EK?), DDC just seems like the better platform to start with.
DDC also has a kind of minimum guaranteed performance, in that you don’t have to worry about what else is in your loop. Unless there’s some significant difference in compatibility (in this case, overall block height), I‘d reach for the option that poses the lesser risk.
From a business standpoint, it makes sense to follow in Swiftech’s wake—no one has any questions about the ADII’s performance; only it’s size. Better to address both those problems at once, rather than deal with the skepticism surrounding a weaker pump?
Either way, excited to see what you come up with. ?
DDC also has a kind of minimum guaranteed performance, in that you don’t have to worry about what else is in your loop. Unless there’s some significant difference in compatibility (in this case, overall block height), I‘d reach for the option that poses the lesser risk.
From a business standpoint, it makes sense to follow in Swiftech’s wake—no one has any questions about the ADII’s performance; only it’s size. Better to address both those problems at once, rather than deal with the skepticism surrounding a weaker pump?
Either way, excited to see what you come up with. ?
Last edited:
Similar threads
