Hi all,
thought that although this probably isn't a normal topic for this forum, there may well be the right sort of experience to help me.
I've got a cupboard in a bedroom where I've installed a server rack for all my networking and computer gear. It's got a couple of network switches and a 1U server with their usual high speed high static pressure fans. I still need to measure everything running at once but it seems to be in the 1k-4k range.
I've got a standard solid core 35mm door to use and I was thinking of beefing it up with a sheet of MRMDF on the back onto which I'll stick some carpet underlay (about to have some spare). I'm then going to use batwing style jamb/head seals from Lorient and one of their drop threshold seals.
I also want to line the walls of the cupboard with absorbing material and perhaps have a diffuser of some type between the 19" cabinet and the door. Any thoughts on the best materials to use for this sort of frequency range? Any other pointers on what I could look for?
Cheers.
thought that although this probably isn't a normal topic for this forum, there may well be the right sort of experience to help me.
I've got a cupboard in a bedroom where I've installed a server rack for all my networking and computer gear. It's got a couple of network switches and a 1U server with their usual high speed high static pressure fans. I still need to measure everything running at once but it seems to be in the 1k-4k range.
I've got a standard solid core 35mm door to use and I was thinking of beefing it up with a sheet of MRMDF on the back onto which I'll stick some carpet underlay (about to have some spare). I'm then going to use batwing style jamb/head seals from Lorient and one of their drop threshold seals.
I also want to line the walls of the cupboard with absorbing material and perhaps have a diffuser of some type between the 19" cabinet and the door. Any thoughts on the best materials to use for this sort of frequency range? Any other pointers on what I could look for?
Cheers.
Hi,
Interesting.
My first question is do you need the server to be 1u?
Second question what is the max temp the gear can accept before failing?
As i see it it should not be too much difficult to soundproof the cupboard but the main issue you'll face is that you need some passive airflow.
The most logical thing should be to organise your rack to let air follow a logical path: coming from beneath with the most heat generating device the lower possible, let some space between units and a passive output in the upper part of the cupboard to let heated air go away.
You could use some sheets of aluminium to direct the air flow ( kind of duct or labyrinth ) and use lead foam/or other absorbent material to lessen the noise along the way.
Could you make a sketchup plan to show what the cupboard look like?
If you could switch from 1u to 3u (for your server) everything could be far easier ( bigger fan running slower= less noise), hybrid psu like some Seasonic's offer and your server could be almost silent.
If your noise spectrum is in the 1k and up range it will be easier to absorb it but you'll face a temperature rise issue, so you'll have to think about the air flow and how to make it the most efficient possible.
Interesting.
My first question is do you need the server to be 1u?
Second question what is the max temp the gear can accept before failing?
As i see it it should not be too much difficult to soundproof the cupboard but the main issue you'll face is that you need some passive airflow.
The most logical thing should be to organise your rack to let air follow a logical path: coming from beneath with the most heat generating device the lower possible, let some space between units and a passive output in the upper part of the cupboard to let heated air go away.
You could use some sheets of aluminium to direct the air flow ( kind of duct or labyrinth ) and use lead foam/or other absorbent material to lessen the noise along the way.
Could you make a sketchup plan to show what the cupboard look like?
If you could switch from 1u to 3u (for your server) everything could be far easier ( bigger fan running slower= less noise), hybrid psu like some Seasonic's offer and your server could be almost silent.
If your noise spectrum is in the 1k and up range it will be easier to absorb it but you'll face a temperature rise issue, so you'll have to think about the air flow and how to make it the most efficient possible.
Last edited:
Ok could you post some pictures of the whole thing?
Do you have some structural born noise issue? I ask because this is the most troublesome problem you could face.
For absorbent material take a look at the link in this post:
https://www.diyaudio.com/forums/room-acoustics-and-mods/327757-sounding-4.html#post5555119
If the noise freq is really 1khz and up you'll see that many materials should be ok ( look for alpha sabine of 1 or up from 1k and up). But keep in mind that this materials work in transforming sound into heat, that said it should be negligable versus the heat produced by gear.
Massive heavy material for the door may work but this is not mandatory what matter is the resonant frequency and the damping of it.
To damp the door you can use bituminen sheet ( same thing used in car's door or for metal roof).
What do you call a diffuser? If this is the acoustical treatment it will be of zero help here.
Do you have some structural born noise issue? I ask because this is the most troublesome problem you could face.
For absorbent material take a look at the link in this post:
https://www.diyaudio.com/forums/room-acoustics-and-mods/327757-sounding-4.html#post5555119
If the noise freq is really 1khz and up you'll see that many materials should be ok ( look for alpha sabine of 1 or up from 1k and up). But keep in mind that this materials work in transforming sound into heat, that said it should be negligable versus the heat produced by gear.
Massive heavy material for the door may work but this is not mandatory what matter is the resonant frequency and the damping of it.
To damp the door you can use bituminen sheet ( same thing used in car's door or for metal roof).
What do you call a diffuser? If this is the acoustical treatment it will be of zero help here.
Last edited:
Heavy door materials will help a little, but more important factor is air tightness.
Yep, that'll be the jamb/head/threshold seals I mentioned. I was hoping for more quantitative results from anyone who's tested different materials.
No, I think it's all air transmission. It's all in a network cabinet that is mounted to the back wall of the cupboard. So I've got the chance to line the cabinet itself which is just glass and thin steel. To line the walls around it. And to sort the door out.
Thanks, will have a read.
I was starting to read up on constrained layer damping but haven't got very far yet. I could use 12mm MDF and sandwich two sheets together before attachign that to the door. And then still add the underlay or any bitumen sheet to the inside face.
Probably my bad terminology. I'm thinking of sound absorbing materials so something hanging in front of the door? Would a heavy curtain help?
Have you found out where the noise originates? If it is 1-4 kHz, I can imagine fans are culprits. If the plastic housing of a fan is screwed directly onto a metal sheet, you might increase the fan noise several times with resonanses in the metal sheeting. A piece of self adhesive bitumen would kill these resonances, a 50 % covering should be enough. Another fix could be rubber grommets between fan housing and metal. Quite cheap and often found in computer stores. Simple o-rings would be OK as vibration isolators / dampening as well. Viton rubber o-rings are often used as vibration isolators in apparatus sensitive to vibrations, -microscopes, astronomical instruments etc. Suitable compression is 10-13% of the o-ring cross section.
If you go the CLD route, MS polymer (like Tec 7 etc), low to medium strength, stays soft and flexible over time and can be good as the constrained layer between the MDFs. A thin layer like 1 mm is better than a thick one. Use a spatula to smear it out quickly, as you have quite little time before the MS polymer starts form a "skin". Moisture is needed for it to cure, so you can dampen the MDFs a bit before you apply the MS polymer.
If you go the CLD route, MS polymer (like Tec 7 etc), low to medium strength, stays soft and flexible over time and can be good as the constrained layer between the MDFs. A thin layer like 1 mm is better than a thick one. Use a spatula to smear it out quickly, as you have quite little time before the MS polymer starts form a "skin". Moisture is needed for it to cure, so you can dampen the MDFs a bit before you apply the MS polymer.
Hi,
About fan noise as Adhoc1 pointed first thing is to try to isolate the fan from the case.
The real problem with 1u rack being they use small fans which have to run very fast and have non optimal fan geometry relative to noise ( they need high static pressure to be effective so noise optimized fan geometry are not adapted in that case).
Overall you are kind of stuck with them anyway as you can't change gear.
So for the cupboard door you could go CLD to lessen noise transmission from the door itself and making the cupboarb airtight.
The products from Lorient seems good but it'll cost (a lot of) money from my previous experience with soundproofed doors.
Maybe you could try something cheaper before: build a wooden frame inside the cupboard spaced some 2/3mm away from the door when it is closed. You seal the 'frame' with silicone to the interior of the cupboard and then you use stick on closed cell foam to make the 2/3mm space sealed/airtight.
I think this is the material jplesset talked about previously ( weatherstripping). It is usually used on ancient window for heat insulation and you can find it easily without breaking the bank.
The foam will serve 2 purpose, sealing and act as a kind of damper which could help if you make a CLD.
Lastly yes using a blanket of absorbent material between the rack and door will work nicely. If i was to do that i would stay away from fiberglass though : it will release particle longterm and this may cause issues with the fans.
Something like a large blanket of 'bonded logic' or 'metisse' (cotton wool) should be less problematic longterm and work as good as a wideband absorber for frequencys of concern. I agree with Rayma about this.
About fan noise as Adhoc1 pointed first thing is to try to isolate the fan from the case.
The real problem with 1u rack being they use small fans which have to run very fast and have non optimal fan geometry relative to noise ( they need high static pressure to be effective so noise optimized fan geometry are not adapted in that case).
Overall you are kind of stuck with them anyway as you can't change gear.
So for the cupboard door you could go CLD to lessen noise transmission from the door itself and making the cupboarb airtight.
The products from Lorient seems good but it'll cost (a lot of) money from my previous experience with soundproofed doors.
Maybe you could try something cheaper before: build a wooden frame inside the cupboard spaced some 2/3mm away from the door when it is closed. You seal the 'frame' with silicone to the interior of the cupboard and then you use stick on closed cell foam to make the 2/3mm space sealed/airtight.
I think this is the material jplesset talked about previously ( weatherstripping). It is usually used on ancient window for heat insulation and you can find it easily without breaking the bank.
The foam will serve 2 purpose, sealing and act as a kind of damper which could help if you make a CLD.
Lastly yes using a blanket of absorbent material between the rack and door will work nicely. If i was to do that i would stay away from fiberglass though : it will release particle longterm and this may cause issues with the fans.
Something like a large blanket of 'bonded logic' or 'metisse' (cotton wool) should be less problematic longterm and work as good as a wideband absorber for frequencys of concern. I agree with Rayma about this.
You can't make an aquarium out of sponges. Soundproof is airtight.
If you have airflow in and out if the cupboard, you will have noise, understand that going into the problem.
That said, if you can seal up the door with weatherstripping it will make a huge difference, and at the frequencies you mention, yes, batting, insulation, carpet, foam, or almost anything soft on the inside of the door and walls will make a substantive difference.
I recently made a soundproof semi-isolated dedicated listening room in my house, and getting it airtight makes more difference in actual sound transmission in and out than all the other steps combined.
If you have airflow in and out if the cupboard, you will have noise, understand that going into the problem.
That said, if you can seal up the door with weatherstripping it will make a huge difference, and at the frequencies you mention, yes, batting, insulation, carpet, foam, or almost anything soft on the inside of the door and walls will make a substantive difference.
I recently made a soundproof semi-isolated dedicated listening room in my house, and getting it airtight makes more difference in actual sound transmission in and out than all the other steps combined.
I think so. Other than very low frequency isolation, airtightness is probably the most important. I used mechanical sweep seal for double layered door (for double wall) and it worked surprisingly well.
http://www.zerointernational.com/images/products/fullcatalog.pdf
or
TMS Automatic Door Bottom/Regular Duty/Surface Mounted/Clear Anodized Aluminum/30" long - Door Closers - Amazon.com
I still have slight low frequency leak, but I think I need serious spring loaded floor or such to isolate LF. My floor is just isolated via 1/2" neoprene, and isolation is limited.
Last edited:
Hi,
Yes Plasnu the real 'ennemy' is the LF structure born vibration. I ve been involved in few (pro and semi pro) studio build and some project had been designed by acousticians.
For this kind they either refused to let someone else than their 'build team' do the works either specified they could not be held responsible of results if the plans wasn't followed exactly as specified by them: a nail ( just one small nail) could ruin LF isolation in cas of 'floating' stucture...
And some build technique are just such a pain in the ... that 'regular' build workers are a cause of concern on this kind of project as they always have 'better, smarter way of doing it than speced on plans' : recipe of failure.
Yes Plasnu the real 'ennemy' is the LF structure born vibration. I ve been involved in few (pro and semi pro) studio build and some project had been designed by acousticians.
For this kind they either refused to let someone else than their 'build team' do the works either specified they could not be held responsible of results if the plans wasn't followed exactly as specified by them: a nail ( just one small nail) could ruin LF isolation in cas of 'floating' stucture...
And some build technique are just such a pain in the ... that 'regular' build workers are a cause of concern on this kind of project as they always have 'better, smarter way of doing it than speced on plans' : recipe of failure.
krivium, I know many semi-pro studio in Brooklyn that have a serious floor resonant issue. I have no structural related vibration issue here fortunately, but it is probably just because the room is rather small. (My LF vibration issue is the furniture in the room when I playback really loud. At certain frequency, certain part of the furniture resonant. )
)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.