Comparison of Wood Platforms on Concrete for Tactile Bass

[dgage]

(Posted this on AVS too - http://www.avsforum.com/forum/19-dedicated-theater-design-construction/2443450-comparison-wood-platforms-concrete-tactile-bass.html)

 

The purpose of this series of tests is to try to help answer a question I’ve received quite often from prospective customers, which is what type of platform works best on concrete to transmit bass? The goal is to provide specific recommendations with data for what platform works best on concrete to transmit bass as well as what underlayment works best. These will be scaled tests using 4x8 sheets of plywood.

Equipment utilized:

• VibSensor app on iPhone 6S Plus running iOS 9.3.1
• Windows 10 laptop running REW 5.14
• Focusrite 2i2 microphone preamp
• Earthworks M50 microphone
• Simpson 890-2 Sound Level Calibrator
• Reed SC-05 Sound Level Calibrator
• Deep Sea Sound Mariana 18S 18” Subwoofer with 4,000w amplifier
• Bathroom carpet with rubber backing to go under subwoofer (simulates carpet and pad)
• Bathroom carpet with rubber backing to go under microphone and phone

Platforms built:

 

3 platforms were built with the 2x4 platform also receiving “rim joist” ends for an additional set of tests.

• 2x4 platform
• _____21/32” OSB plywood built with three 2”x4”x8’ “joist” set on 16” centers. The plywood is screwed to the 2x4s using countersunk #8x2.5” deck screws every 11.75”.
• 2x4 platform with “rim joists”
• _____2”x4”x36” “rim joists” were added to each end of the 2x4 platform to simulate a traditional suspended floor.
• 1x4 platform
• _____21/32” OSB plywood built with three 1”x4”x8’ set sideways on 16” centers. The plywood is screwed to the 1x4s using countersunk #8x1.25” deck screws.
• 2 sheets of OSB plywood
• _____Two sheets of 21/32” OSB plywood screwed together with countersunk #8x1.25” deck screws. Plywood has screws 1” in from the perimeter and every 11.75” throughout.

Underlayment used:

As underlayments are often used to separate concrete and wood to keep things quiet, I utilized two materials I had available, 1/4” cork and 1/4” rubber. Each wood platform was tested with both underlayments.

Tests performed:

• Test 1 - On Concrete
• Test 2 - 2x4 platform with 2x4 "rim joist" ends - Cork
• Test 3 - 2x4 platform with 2x4 "rim joist" ends - Rubber
• Test 4 - 2x4 platform - Rubber
• Test 5 - 2x4 platform - Cork
• Test 6 - 1x4 platform - Cork
• Test 7 - 1x4 platform - Rubber
• Test 8 - 2 sheets 3/4" OSB - Rubber
• Test 9 - 2 sheets 3/4" OSB - Cork

Testing Details

• A single DSS Mariana 18S subwoofer was placed in the center at one end of the 8' long plywood.
• The phone was placed in the center at the other end of the 8' long plywood, approximately 6.5' away.
• The subwoofer and phone were laying on top of bathroom carpets with a rubberized bottom to simulate a carpet with pad underneath.
• Volume was calibrated using a pair of SPL calibrators and was approximately 110 dB with the microphone on the carpet 6' away from the subwoofer.
• I used an app called VibSensor on my iPhone 6S Plus and set a 5 sec delay followed by 30 seconds of data capture.
• I turned on REW’s Subwoofer Pink Noise and then started the data capture on the phone.
• Tests were repeated with each underlayment; cork and rubber.
• Tests were also repeated with and without 190 lbs of rubberized weights. This is to see if having a 110 lb sub at one end of the platform was causing more vibration since the other end had no weight. I utilized 190 lbs of weights and since this is a scaled test, this would roughly simulate a couch with a couple people sitting on it if full scale. The 190 lbs of weights were evenly spread on each side of the carpet with phone on it using 2 layers of 1/4" rubber to separate the rubberized weights from the wood.

Results in next post.

[dgage]

I've attached the raw Excel workbook with acceleration data from the VibSensor phone app as well as the charts. I'm also attaching the X, Y, and Z acceleration charts in picture form here.

Quick analysis shows the 2x4 raised platform seems to provide the most "vibration". For those with more limited space, the choice between the 2 sheets of OSB and the 1x4 platform seems to favor the 1x4 platform. Regarding cork vs rubber, rubber appears to be a little more tactile.

Regarding the charts, notice the X-Axis Acceleration charts are at a smaller scale than the Y and Z-Axis charts.

 

I updated the Excel workbook to include a data summary as well as some analysis to determine a winner. Since the data did not paste in well, I have taken a picture of the data from the Excel workbook.

 

 

Wood on Concrete Test Results.zip

[dgage]

Reserved 1

[dgage]

Reserved 2

[maxmercy]

Interesting experiment!  

 

JSS

[MemX]

Excellent work!!

 

Just to ask the question, the sub is FF, isn't it?  Would be interesting to see the results of a FF vs DF test, although I imagine that's quite a lot more hard work to rotate the sub and double the number of tests...  lol

[Ricci]

Excellent stuff Dave...Someone has got some new toys.

 

I too would be very interested to see the results of the sub rotated 90deg and 180deg away with perhaps just one riser config.

[Kvalsvoll]

Great experiment, thanks for sharing.

 

From the pictures and the description it seems the platform is not sealed around the edge?

Sealing the platform will make a huge difference, especially at the lowest frequencies.

 

If it is open, it will be close to an acoustic short circuit at low frequency - there will be little mechanical force acting on the platform because the pressure on both sides is the same, and it will be worse for a smaller platform. 

[SME]

Very interesting, but how many people can place their subwoofer right on the platform?  I second the idea of sealing the platform because I expect it will respond better when the sub is *not* placed on the platform.

 

Here are some other thoughts.  Have you thought about doing a platform with the rim joists but having rotating the 2x4s so that the (nominal) 2" part is vertically oriented?  Another possibility is to increase the spacing between 2x4s.  My thinking is that this will reduce the stiffness of the platform, which could help shift the response to lower frequencies.  My thinking is that an actual floor is typically longer than your test riser, so your riser will be a lot more stiff than a typical floor if you use the same construction methods.  Naturally, you don't want so much compliance than the riser can't support the weight of furniture and people, but a bit less stiffness would probably help move some of those resonances down in frequency a bit.

[dgage]

That might be a question for Bosso since he knows the span tables.  I was thinking a normal 16 feet long/wide room would have 2x10 joists although I didn't check a span table.  Scaling down to an 8' "floor" I decided to use 2x4 "joists".  I probably could have done some math to literally scale everything but I didn't have much time.  A prospective customer wanted/needed the data as soon as possible and since I wanted the data too, I went for it.  Definitely many ways the testing could be improved, hell, I'd really like to go full scale but don't see that happening any time soon, if ever.

[maxmercy]

You need this:

 

http://www.newnewsstories.com/forum/155-diy-speakers-subs/1507428-follgott-s-build-18-x-peerless-xxls12-3.html#post24226660

 

Less dependence on peaky resonances the structure has, just the overpressure from the subs doing the heavy lifting.

 

But instead of damped springs, use air-springs and have it self-level to adjust for the load (people on the riser), in order to fit a large variety of loads and still work well.

 

If you are on a slab, I can think of no better passive way to get that <10Hz effect that some suspended floors do.

 

JSS

[SME]

That might be a question for Bosso since he knows the span tables.  I was thinking a normal 16 feet long/wide room would have 2x10 joists although I didn't check a span table.  Scaling down to an 8' "floor" I decided to use 2x4 "joists".  I probably could have done some math to literally scale everything but I didn't have much time.  A prospective customer wanted/needed the data as soon as possible and since I wanted the data too, I went for it.  Definitely many ways the testing could be improved, hell, I'd really like to go full scale but don't see that happening any time soon, if ever.

 

What was I thinking?  Yeah, the 2x10s should be a lot more stiff than 2x4s.  IIRC, from my brief review of beam theory (I'm trained as an engineer but not structural engineering), bending stiffness scales with the cube of thickness.  So, going from 2x10 to 2x4 involves a change in stiffness closer to 16X rather than 2.5X.  OTOH, the substantial load on those members in a house may more than make up for that change in stiffness.  I don't know.  There's a huge number of variables involved due to there being many interconnected parts in a house.  The more interesting resonances may have little or nothing to do with bending of the floor supports.  In any case, this is an interesting first step, even if further study is mostly impractical at this point.  I mean, who wants to build a large riser just to test to see how well it responds to bass in the room?

 

You need this:

 

http://www.newnewsstories.com/forum/155-diy-speakers-subs/1507428-follgott-s-build-18-x-peerless-xxls12-3.html#post24226660

 

Less dependence on peaky resonances the structure has, just the overpressure from the subs doing the heavy lifting.

 

But instead of damped springs, use air-springs and have it self-level to adjust for the load (people on the riser), in order to fit a large variety of loads and still work well.

 

If you are on a slab, I can think of no better passive way to get that <10Hz effect that some suspended floors do.

 

I like this idea a lot.  I recall that others have reported that floating floors as are typically used for *sound isolation* also do a great job of transmitting low frequency vibrations that can be felt.  I'm not sure about using air-springs though.  I may be missing something but I expect you would want a fair amount of damping specifically in order to avoid those peaky resonances.  And in fact, if you really want to avoid those resonances, it would seem to me to make even more sense to build an open riser (to avoid transfer of energy from the air) and installing multiple tactile transducers in different locations underneath it.  My reasoning is that you can EQ the signal(s) going to the tactile transducers so that they don't overstimulate those peaky resonances, and by spreading them out, you can reduce the excitement of modal resonances within the structure similar to how you spread out subs to reduce excitement of modal resonances inside the air space of listening rooms.