[Martinsliquidlab.org] Pump Noise Testing Round 1

[meanmoe]

If they change the mass, and they do, they could at least shift the modal frequencies around.

[Mazlov]

Thx Martin. I have been waiting for this a long time. Nice to finally have data to prove that my direct metal contact mounted mcp355 is noisy.

[Martinm210]

Thanks!

Martin: you don't happen to have xspc or ek ddc tops? Imho they are among most widely used, so interesting how/if they change something noise wise.

I have my old acrylic XSPC top, but sorry not any of the current model acetal tops.

I think it really comes down to acetal having better noise reduction abilities, and the larger the mass/thicker the better. The Koolance tops used both have about a 1" Acetal thickness top or more and the PMP-450 version also has that motor cover which probably helps block some noise coming from the motor shell.

I also have a reservoir for the Koolance tops that I'll try next, and see if there is any benefit to the mass of water/reservoir. I'm guessing that should help with the non-decoupled result a fair amount but we'll see.

[MagisD]

Any chance of sticking the 35x in the 402 ?

Sent from my X10a using Tapatalk

[Martinm210]

Any chance of sticking the 35x in the 402 ?

Sent from my X10a using Tapatalk

Sorry, the 35X is my Sandy Bridge build, I've been trying a long term test with and PWM thermal throttling along with a Sunbeam Rheosmart to control fans.
Also the 402 and one PMP-400 have been getting a daily long term test/use in my son's PC and I don't think he would be too happy if I pull his machine apart again.

So both are in daily long term tests/use right now are not available for other tests.

[voigts]

Any word on the Aquastream XT yet? I'm still keeping my eye on this.

[theseeker]

Any word on the Aquastream XT yet? I'm still keeping my eye on this.

Skinnee is testing the XT.

[Martinm210]

Any word on the Aquastream XT yet? I'm still keeping my eye on this.

Nothing yet in my hands.

I did get a few more pumps tested on this noise piece though. I should have those processed and I'll update here when that's up..

[voigts]

Just wondering as I am currently running an Aquastream XT Ultra, and moved to it from the DDC pumps as I can't stand their high pitched whine. With the advent however of the new MCP35x, I am wondering how this pump with PWM control fares in comparison, and how they both fare compared to the D5 vario with Koolance COV-RP450 top that you tested . In the case I'm currently working on, I could fit the Aquastream or a D5, but the MCP35x would fit easier. I just want to be using the quietest pump I can get my hands on, as to me, the pump is the limiting factor in having a build as quiet as possible.

[Martinm210]

Just wondering as I am currently running an Aquastream XT Ultra, and moved to it from the DDC pumps as I can't stand their high pitched whine. With the advent however of the new MCP35x, I am wondering how this pump with PWM control fares in comparison, and how they both fare compared to the D5 vario with Koolance COV-RP450 top that you tested . In the case I'm currently working on, I could fit the Aquastream or a D5, but the MCP35x would fit easier. I just want to be using the quietest pump I can get my hands on, as to me, the pump is the limiting factor in having a build as quiet as possible.

The key to the 35X is the PWM control and dynamic throttling to get the silence benefits out of it. I have one in my CPU loop at the moment and using speedfan to throttle the pump from 30% to 70% using speedfan and a 42C trigger. This allows me to run it at 30% in pure silence 99% of the time. The only time it even kicks into high speed is gaming or rendering a video. I also have a sunbeam rheosmart dialing fan speeds down to 900RPM and daisy chained with the 35X. My system has never been so quiet prior to getting it setup on temp throttling, I don't know why I waited so long..

[Garrett]

Is that possible with the mcp355 too, Martin?

[meanmoe]

I also have a sunbeam rheosmart dialing fan speeds down to 900RPM and daisy chained with the 35X..

Could you expand on this... How is it set up? I like silence...

[thegcpu]

Pump "temp throttling" I like that. I have a 350 on an open case that’s not loud but you can hear it. I was looking for a reason to get a 35X. On a quest for a 35X now…

[thegcpu]

Could you expand on this... How is it set up? I like silence...

I'm going to us the MOBO's CPU fan temp BIOS profile..

[Martinm210]

Is that possible with the mcp355 too, Martin?

Yes, I think so to some degree using smart fan controller that could either convert PWM signal from the motherboard or ramp up/down on it's own. It's just not going to have as much range as the 35X and depending on the setup..not as safe as using the 35X (Fan controller could drop below the minimum voltage). Haven't tried that myself, but I'm sure there is potential to at least reduce it up and down whatever the pump voltage range is as long as you have a controller to do that with.

There are quite a few options out there that should have really good control on this, I just haven't used much myself.

Could you expand on this... How is it set up? I like silence...

Basically the sunbeam rheosmart plugs into the CPU fan header and has a daisy chain connector that the 35X control plugs into. Since MSI has very poor PWM control software, I've had to resort to using speedfan to do my software control. I've got 42C as my trigger and setup it up so anytime temps are below that it sends 28% PWM out to the rheosmart and 35X. Fans are hooked to the rheosmart and dial up and down like the pump does. It's pretty basic, but works good.

The setup is fairly limited on control using the rheosmart and speedfan that better controllers or motherboards would do better with:

-Because are both daisy chained, there isn't an ability to control them independently. I wouldn't mind dropping my fan speed even lower, but 30% is getting pretty low on pumping power.

-Speedfan control is pretty limited to a switch. It does ramp up and down somewhat smoothly, but there isn't an ability to progressively ramp up. It's basically 30% below 42C and 70% above 42C...so not a smooth as you could probably do with a more advanced controller.

-Water Temp Control - I think it probably also makes more sense to regulate and control fan speed relative to water temperature rather than CPU temps. This would likely be very smooth because of the water's natural heat storage ability. Running fan speeds off CPU temp is ok, but I think water temp would be the ideal.

Anyhow, I had planned to blog something on this...I just haven't yet.. I'll work on that..

[Garrett]

Excellent stuff Martin, thanks for clearing it up for us
I'm going to look into it, I'm curious what the MCP355 will do @ 7v

[Martinm210]

Excellent stuff Martin, thanks for clearing it up for us
I'm going to look into it, I'm curious what the MCP355 will do @ 7v

FYI,
I did some testing on the rheosmart tonight. I think a pump is a little too much for it. I could manually dial a pump down to 8V with a DDC pump, but the heatsink rose quickly to 64C. Then I tried to run it via PWM and it would only run from 11.21V(100%PWM) max to 10.02V(70%PWM) before it failed trying to go lower.

So the rheosmart doesn't seem to cut it for pump control, but there are others out there designed to handle pumps. Koolance & Aqua Computers have some options designed for pumps that could be automated.

[Alexandr0s]

I've already had a short discussion about this with WL, but I'll ask it again here:

According to Swiftech, the MCP350 produced 26dBA in their test. The MCP55 produced 32dBA. Now my question was, what produces more noise: 2 MCP350's or 1 MCP355? According to Swiftech's test, it should be the MCP355, because 2 MCP350's should produce 26+3= 29dBA. I'm curious because leaving issues such as cost and the required extra space aside, 2 MCP350's would offer a better performance/noise ratio, be more quiet than an MCP355 and also provide redundancy.