Pump Noise Decoupling Comparison Video

[Martinm210]

Another somewhat random pump noise experiment compares direct metal vs. neoprene vs. egg-crate...:D

Zoom to the end if you just want the top speed comparison portion.

http://www.youtube.com/watch?v=nx7-n3FBeDI

A-weighted frequency response differences:
http://martinsliquidlab.files.wordpress.com/2011/03/pumpdecouplerfrequencyrespo.png


And for those who would like to provide cooling and decoupling benefits to the DDC series, here is one idea:
http://martinsliquidlab.org/2011/03/04/pmp-400-or-mcp-35x-ddc-series-cooling-lift/

http://martinsliquidlab.files.wordpress.com/2011/03/pmp-400lift4.jpg?w=614&h=614

[Conumdrum]

Yep, it shows. Good test Martin. The egg crate or just about anything except direct metal mounting makes a world of diff. I'm sure the sound to you is more like a GT fan now, if you get my drift.

Eggcellent chart work too!

[voigts]

Great work and comparison. I've never tried an egg crate on a DDC. I wonder how Petras Gel-Stuff compares.

[Martinm210]

Yeah, I guess I always assumed any little bit of decoupling is good enough, but there is some difference. I've generally been happy with just the neoprene pads myself. Might be interesting to have sort of a larger decoupling shootout. While the egg-crate functions really well, it also gets first place for hideous visual appeal..:D

[Rubycon]

When I think of egg crate I think of those acrylic light diffuser panels. :D
However that was clarified a few years back when I began my career in the cruise biz and found that people ask for (eggcrate) when they feel uncomfortable sleeping on their beds. :ROTF:

Anyhow, as you can see sitting the pump on foam is pretty effective at decoupling. The next step would be your hoses. Yes any rigidity in the hose will amplify the sound.

Want to see how loud a pump can really get? Bolt it firmly to a large aluminum chassis and plumb everything in rigid pipe with no flexible hose at all. I bet it would hum like a really loud tar ballast - easily heard in the next room. :eek:

[avddreamr]

Martin, when I first got my mcp355, I was impressed by how relatively quiet it was, at least in comparison to the pair of ddc-2's that I have.

However placing a load on it, definitely increased the level of noise, and dampening the sound works great if you do it completely. The other option that works not quite as well is to suspend it. In my case, it's hanging via tubing from my res.

I'm still not happy with the noise, but it's tolerable.

[jinsean]

Hi Martin, have you considered the 'gel stuff' from Petras? That gel is amazing. I've been using it for the past 3 years now. PTS still carries the product too.

The only downside is that the gel shrinks in size over time.

[Martinm210]

When I think of egg crate I think of those acrylic light diffuser panels. :D
However that was clarified a few years back when I began my career in the cruise biz and found that people ask for (eggcrate) when they feel uncomfortable sleeping on their beds. :ROTF:

Anyhow, as you can see sitting the pump on foam is pretty effective at decoupling. The next step would be your hoses. Yes any rigidity in the hose will amplify the sound.

Want to see how loud a pump can really get? Bolt it firmly to a large aluminum chassis and plumb everything in rigid pipe with no flexible hose at all. I bet it would hum like a really loud tar ballast - easily heard in the next room. :eek:

Yeah, not sure if this was actually used to crate eggs back in the days, or if it was simply a similar shape. But it's the term most people around here use for the stuff. Most of what you can buy in stores is actally more of a wave than the egg crate. I just happen to have a big chunk in the garage which met up with a pair of scissors...:)


Martin, when I first got my mcp355, I was impressed by how relatively quiet it was, at least in comparison to the pair of ddc-2's that I have.

However placing a load on it, definitely increased the level of noise, and dampening the sound works great if you do it completely. The other option that works not quite as well is to suspend it. In my case, it's hanging via tubing from my res.

I'm still not happy with the noise, but it's tolerable.

Yeah, I've heard of a few people that have done the "Suspend" idea with tubing, depends on the setup, and probably the stiffness of the tubing, but that seems like a good way to decouple plus allow air cooling of the base of the DDC series.


Hi Martin, have you considered the 'gel stuff' from Petras? That gel is amazing. I've been using it for the past 3 years now. PTS still carries the product too.

The only downside is that the gel shrinks in size over time.

I've never had some, but might place an order. I'd also like to try an official shoggy sandwich. :up:

[Serpentarius]

how is it possible for the eggcrate to be that damping? is it because of the height? Martin

[Martinm210]

how is it possible for the eggcrate to be that damping? is it because of the height? Martin

It's soft and more flexible is all. I think in general the thicker and more flexible the interface, the better the decoupling.

Not sure that I would recommend it because it's so hideous, but it seems to work.

I was just tinkering with test conditions and measurement. The egg crate seems like an easy near ideal coupling, and the direct metal contact the worst case.

I haven't quite landing on my methods for testing and comparing all the pumps yet. May round 1&2 seemed to be having microphone troubles. Vibrations from the test bench were transferring into the microphone tripod and I was suspecting the same could be happening to the sound level meter. This was actually intended more as a setup trial than actually testing decoupling methods, but I through in the multiple decoupling scenario's just for fun...:D

I'm going to order some Petra's gel stuff to give that a shot too.:up:

[Waterlogged]

There is a reason we don't put gel under DDC's anymore. ;)

http://www.xtremesystems.org/forums/showthread.php?t=165299

[Martinm210]

There is a reason we don't put gel under DDC's anymore. ;)

http://www.xtremesystems.org/forums/showthread.php?t=165299

Thanks for the memories...:D

I suppose any sort of decoupler is going to have heat issues directly between pump and case with the DDC series....D5's are fine though as their metal casing seems to watercool itself well enough without the heat build up.

The trick to the DDC airflow would be some sort of bracket to lift it, then decouple beween that bracket and case.

A UN bracket (http://www.sidewindercomputers.com/uinbrhobl.html) would work:
http://ep.yimg.com/ca/I/sidewindercomputers_2147_31395714

OR, make your own, Some c-channel steel or aluminum would probably work well enough also. Just need something to lift the pump away from the decoupler and still provide some air flow.:up:

[vhaarr]

Hey martinm, I found a cool way to mount a pump that I'd like to try out :D

Check this out:
http://home.broadpark.no/~pam2/dscf0007thumb.jpg

It's basically some big orings (you can buy them in all kinds of shapes and sizes for a buck in your local plumber store or whatever you have in the US!) fitted to the sides using zip ties :up:

I'd like to try it myself but I don't have room in this cabinet hmm maybe it's time to get a MM case?! :yepp:

[Alexandr0s]

Decided to try it out:

http://i56.tinypic.com/2hod83s.jpg

So far it sounds good. Pretty good noise reduction. The only vibrations I hear are being tranmitted to the reservoir due to the short piece of tubing in between the res and pump.

[voigts]

There is a reason we don't put gel under DDC's anymore. ;)

http://www.xtremesystems.org/forums/showthread.php?t=165299

I wouldn't put a DDC on anything without having the bottom vented. I would cut out a circle in the center of the Gel Stuff under the pump and have it sitting so that the center of the pump is sitting over a case hole or something.

[Martinm210]

I worked up this quick cooling "Lift" blog (http://martinsliquidlab.org/2011/03/04/pmp-400-or-mcp-35x-ddc-series-cooling-lift/)...but a variety of other options exist too.

http://martinsliquidlab.files.wordpress.com/2011/03/pmp-400lift4.jpg?w=614&h=614

http://martinsliquidlab.files.wordpress.com/2011/03/pmp-400lift5.jpg?w=614&h=532

With this you could use gelstuff or any sort of decoupling material without any heat worries.

[BlueAqua]

Nice test Martin. I've used an assortment of things, from Petra's gel pad (which melted under my DDC), the stock foam pad, and sponges. I then decided to try to cook up something of my own and went to Home Depot for ideas. A rubber trowel for spreading tile grout caught my eye. Five dollars later and a bit of handy work and I had myself a seriously good noise isolation device. Note that I didn't attach it with screws, you don't want the vibrations going through from solid surface to solid surface.

http://i996.photobucket.com/albums/af89/BlueAqua2/IMG_2412.png

[Martinm210]

Nice!

I think I'll order some gel stuff and see if I can round up some commercial random materials for a larger decoupling comparison. Not sure anything will beat the really thick soft foams, but maybe there is something that works nearly as well while having some visual appeal as well.

You can also buy the gel stuff from Mcmaster here:
http://www.mcmaster.com/#rubber-sheet-gel/=balrgw

A 12" x 12" piece is $33....could stack a few layers with that much..:)

[meanmoe]

Some traditional isolators would be cool too. I really like the lift. To take care of both problems without a pad, replace the bolts and nylon spacers with some very soft threaded isolation mounts and I bet you'd really have something there.


I would love to get some of these.
http://www.enidine.com/Industrial/CWRMain.html & http://www.enidine.com/Industrial/WireRopeIsolators.html
You get isolation and a thermal path. I worked on a program that intended to use something like these to mount 'electronics'. It was a for a high-g, high-vibration, high-shock environment. If anyone knows where to get miniature wire-rope isolators retail, I'd like to know.

[Martinm210]

Hey martinm, I found a cool way to mount a pump that I'd like to try out :D

Check this out:
http://home.broadpark.no/~pam2/dscf0007thumb.jpg

It's basically some big orings (you can buy them in all kinds of shapes and sizes for a buck in your local plumber store or whatever you have in the US!) fitted to the sides using zip ties :up:

I'd like to try it myself but I don't have room in this cabinet hmm maybe it's time to get a MM case?! :yepp:

I like it. I was actually looking for some large O-rings like that yesterday when I found my nylon spacers, but the store I was looking at didn't quite have any that big. I think any sort of suspension trick like that should work pretty goo.:up:


I wouldn't put a DDC on anything without having the bottom vented. I would cut out a circle in the center of the Gel Stuff under the pump and have it sitting so that the center of the pump is sitting over a case hole or something.

Yeah, they do get hot. Waterlog's link to the melting gel-stuff was enough to convince me. I think I saw some of that a long time ago, but have since forgotten. Even with the base exposed, I was measuring upwards of 55C at the base on the 35X, so insulating that can only make it worse. Lifting or suspending or some other sort of airflow trick should probably go hand in hand with any decoupling.


Some traditional isolators would be cool too. I really like the lift. To take care of both problems without a pad, replace the bolts and nylon spacers with some very soft threaded isolation mounts and I bet you'd really have something there.


I would love to get some of these.
http://www.enidine.com/Industrial/CWRMain.html & http://www.enidine.com/Industrial/WireRopeIsolators.html
You get isolation and a thermal path. I worked on a program that intended to use something like these to mount 'electronics'. It was a for a high-g, high-vibration, high-shock environment. If anyone knows where to get miniature wire-rope isolators retail, I'd like to know.

Oh...those are cool! Curious how well they work. Seems like metal wire would be a bit more rigid, but maybe not in a suspension system like that.:shrug:

[antiacid]

I'm using the foam pads that came with my EK blocks. I put two layers, both folded in half and it works like a charm! Not only that but I also get to re-use packing material so I save the environment with my mighty dampening solution ;)

[Johnny87au]

Not sure what everyone saying dont use the pads, dont really heat up enough to ruin the pcb's.. Good ventilated case is more then enough for mine.. Thanks for the testing martin!

[Alexandr0s]

Johnny, same reason we say: don't use barbs without clamps. 9 out of 10 times it will go perfectly fine, but we'd rather not take responsibility for that one time it does fail just because we said it most likely would be fine. In other words: Best to be safe than sorry. Especially regarding other people's stuff ;).

[meanmoe]

Oh...those are cool! Curious how well they work. Seems like metal wire would be a bit more rigid, but maybe not in a suspension system like that.:shrug:

It wasn't so much a suspension system, as an isolation system. The electronics had to be kept in spec while the world around them vibrated on the way up.

When they were doing the trade on what to use, I got to play with some minis that were bought to go on a satellite. I think they were more rigid axially than laterally and the resistance was nonlinear. In fact, I think that they had very little resistance laterally and I think that's why they worked well as isolators. The vibration was transfered into low frequency motion on one side of the rope dissipating the energy that would've been vibration on the other side of the rope (the attachment point).

I just really want a set 'cause I also think they're cool - it's like the industrial version of the bungee method that we've been using on hard drives for some time. And I personally don't like pads for what Alexandr0s said and because for them to work, they get awkward. The 1" egg crate foam is a perfect example - it works surprisingly well, but I really don't want to put it in my case. And now I have a potential heat issue because I don't ventilate well, I watercool.

I'd also like to see the UN brackets tested in your setup when you decide to go that way of testing ... (hey Armeniandave!)

As always, I really admire what you do.

[armeniandave]

I would too!! :up:

[UrbanSmooth]

Martin, thank you so much for all of the vigorous testing that you do! I'm at the point in my latest build where I will be setting up the pumps, and I just happen to come across this thread on here. I knew that I needed to decouple the pumps to help with vibration, so I cut out some of the soft foam from my AX1200 psu box.

However, the part that peaked my interest here was about the heat from the pump being trapped between the pump and the foam. The "lift" idea really opened my eyes! I actually have a bag of 100 nylon spacers that I ordered from McMaster-Carr that I was going to use in another part of my build, but ended up not doing. It's good to know that these nylon spacers will get some use! Thank you!

[Martinm210]

You bet!..glad it helps.:)

[MagisD]

Ok so this is bad ?

https://lh3.googleusercontent.com/_8NbIoxo4dfc/TYYlyuPi_PI/AAAAAAAAAqw/10xyJ7pHM7o/s800/DSC_0383.jpg

Great now I have to rework my loop and make risers. I'm thinking best bet is a undesign mount on a shoggy sandwich. A fancy version of martins setup.

Also just sorted out super quick and dirty method that most of us would have laying around.
Thumb screws inserted into short pieces of 3/8 tubing.

Can screw into bottom if pump and be mounted on other end to whatever.
I'm thinking mounted to Plated glued to foam.

Anybody see any holes in my idea ?

https://lh3.googleusercontent.com/_8NbIoxo4dfc/TYYlkDGM1bI/AAAAAAAAAqU/EsWRum52QuA/s800/DSC_0381.jpg

https://lh5.googleusercontent.com/_8NbIoxo4dfc/TYYlrvLQD8I/AAAAAAAAAqo/Wr6uVLcJ4gQ/s800/DSC_0382.jpg

Sent from my X10a using Tapatalk

[Martinm210]

That should work. I've also heard of people "suspending" the pumps from tubing, so I'm trying that our right now. It works pretty well for the smaller DDC series pumps, this is how I have one running and also works decoupled fine. I also stuck a few ram sinks to the bottom that I had lying around for extra cooling.
http://martinsliquidlab.files.wordpress.com/2011/03/35xsuspension.jpg