Noise Control Continues

With Summer tour dates heating up, and our favorite festivals appearing to be taking the Summer off, I was lacking an out-of-town overnight show to really kick this effort into high gear. Well... Phil Lesh and Friends have announced Pacific NorthWest tour dates, hitting 2 of my all-time favorite outdoor venues: the Cuthbert in Eugene, OR and the Marymoor in Redmond, WA. Phil played these 2 venues back-to-back with Further in 2013 (See Tale of 2 Trips), so I conclude that these are his favorite NW outdoor venues too. Anyway, June 11 and 12, Phil will be at the Cuthbert and Marymoor respectively, and so will we. With that, we have some serious motivation to have Hapy buttoned up by the end of May. In that spirit, I finally got after the big sections of the floor, so today's post is all about that. 

Tear Down

starting the Jute

I accept that maybe I have been finding ways to delay the full-blown interior tear down. I messed with electrical (See Ceiling Wiring Rough In), built a replacement shelf (See New Cab Shelf), even messed around with the stereo (I will post on that once it reaches completion). At some point, I was going to have to remove the lot couch and the rock-n-roll bed so I could get after the floor underneath them. The lot couch removes easily enough, but the sliders also needed to come out so I could get the complete floor. Back when I first installed these runners, I mounted the runners with 17mm bolts. I added nuts underneath. I initially just loosened the runners and forced the Jute underneath. I later decided to completely remove the runners so I could get a seamless heavy vinyl sheet down.

My implementation of the rock-n-roll bed is held to the floor by a pair of Phillips screws holding a pair of arms to the rear deck, a 17mm bolt underneath, a pair of small nuts on a metal bracket mounted on the passenger side rear wheel well and a single 13mm bolt through to the fridge cabinet. The electrical bits, however, add to the complexity. I have my luxury fuse box under there, so I had to tear down what was left of that. Plus, the original Westfalia electrical includes the cables routing city power into the rectifier (110 to 12V converter) that's buried inside a wood box and running to the outlet on the front face of the bed. The box is held down with 2 Phillips screws and the outlet box is held by 4 Phillips screws. With the electrical separated from the bed, and the bed detached from the floor, I removed the cushions to manage the unit better. Each of the 2 cushions are held to the bed with 4 long Phillips bolts on each end (so, that's 16 bolts).

cabin Jute mostly complete

I would have completely removed the fridge cabinet, but the mounting of the furnace was so difficult, I did not want to revisit that. Just being disconnected from the rock-n-roll bed loosened it enough for me to fit materials an inch or so underneath it.

Constrained Layered Dampener (CLD)

I have probably posted enough on this material. Still, I applied on the now exposed floor areas. Parts of the floor had already gotten a light treatment in the little valleys, so on top of that I applied larger squares of CLD as well. I still have about a third of the box of ResoNix, with a plan to use some inside the sliding door. I have used some of the Noico around as well. For example, the walk-through partitions behind the front seats each got a large splat of Noico. I do not know if I will be keeping those partitions long term, so I did not want to spend the expensive CLD on those panels. Still, the Noico definitely helped reduce the clatter heard when I tap on the partitions with the roller. Of course, the whole floor and the rock-n-roll bed needed a thorough cleaning, starting with a shop-vac and ending with cleansers. 

Heat Wave Jute

MLV over engine

I was so excited to finally get to the Jute. I has felt like forever since I ordered this stuff and started working on noise containment. I started with a rectangle on top of the engine access hatch, with an extra couple of inches all the way around. I pulled the computer cover, and Jute'd the rest of the rear deck, and then moved forward. I fit a segment under the rear edge of the fridge cabinet, and doubled up over the front side of the rear driver wheel well. I pierced the Jute for the seat belts and a grounding screw, and set another piece of Jute up to the lot-couch rear slider and continued my way forward, covering the entire floor with rectangles. My method was to define as large a section as I could easily manage and then measure that out on the material. I marked the line with a sharpie and then traced the line with a box-cutter to start the cut. I made a second pass with a pair of scissors. I tried first with just the scissors, but the thickness of the material and 2 foil-faces were too much even for my best pair. As I set shapes down, the bus got progressively quieter. Without a radio, and just ambient neighborhood noise, there was not a dramatic change, it just felt more and more quiet. Very neat. One last observation about the Heat Wave: the foil top and bottom are electrically conductive. So, I would discourage use anywhere near open circuitry or be sure to insulate. For example, I took a section of excess MLV produced below and have it resting between the foil and the spare tire well where the engine computer and fuse box live. No sparky.

Lot Couch Rails

I had loosened the rails enough to force-fit some of the Heat Wave jute underneath, but decided that getting the MLV under the rails would be worthwhile. This meant removing the rails entirely. I had forgotten that when I first installed these rails, I ran bolts from above and nutted from below. Three of the 4 nuts were hidden by the radiator, so getting in there to free them was an all-morning exercise. Still, the nuts came free, the bolts pulled out and the rails were set aside. I took this opportunity to clean under the rails as well as install 4 riv-nuts into the old holes, so I would not have to go under the bus and reach around the radiator to deal with the rails again. That riv-nut kit is definitely paying dividends.

Mass Loaded Vinyl (MLV)

MLV continues

MLV is heavy; my goodness, it is heavy (1 pound per square foot). I found that the sound inside the bus actually brightened a little bit as I set it on top of the Jute, though. The Jute, in contrast, was simply amazing in how much quieter the bus seemed as I added it to the floor.  While I can't measure the impact of the MLV independently of the rest of what I am doing, I do wonder if the additional weight will translate into a corresponding reduction in noise. The SecondSkin web site has a noise transfer reduction table for the MLV that defines the reduction as anywhere from 14dB to 37dB, depending on frequency, with higher frequency blocked more than lower. We'll see. The actual absorption by the Heat Wave was noticeable. The table on the SecondSkin page for Heat Wave (in coefficient rather than dB) demonstrates very little in terms of dB drop below 500Hz: below 500Hz < 3dB, 1k: 7db, 2k: 14dB, 4k:19dB. So, the very high end will be controlled well and the low end not so much. I arrived at these numbers courtesy of the formulas on this blog post, converting the coefficients published by SecondSkin.

Anyway, the MLV delivers in a roll 54" wide. The bus is around 60-1/2" wide so rather than solve for a strip along one side, I laid a section 54" long front-to-back, covering the engine compartment and on down the rear wall behind the rock-n-roll bed. After cutting contours for the narrowing of the rear of the bus, I pierced the MLV for the seat belts and installed them. This took a considerable amount of time, mostly because of the challenges of getting the seat belt mounts in place. I continued with the next sheet of MLV, covering the mid-floor past the lot-couch rail mounting points (see picture above, right).

Re-Assembly

bed and rails in

I chose at this point to re-install the rock-n-roll bed. Unlike the seat belt mounts, locating and piercing for the bed mounts was much easier for some reason. With the bed in place, I pieced holes for the lot-couch rails and installed them. To make a more firm mount to the floor, I added a nut between the rail and the floor for the bolt to pass through. These nuts act like a 3/8" riser through the Jute material. I confirmed the width of the rails (12") and then torqued the bolts. Then, I put the lot couch back in. The couch slid in more easily than it used to. I think the rails were slightly below the lip of the sliding door before, creating some friction/resistance. With the little risers, the couch slides right in. The interior is pretty much the way I started the weekend. Yes, from the start to end of this post, it was 2 full-for-me (6+ hours) days of work, and some uncovered sections remain.

On a random mid-week afternoon, I added an additional strip inside the rock-n-roll bed where the 2 larger sections meet as well as a small piece inside the fridge cabinet. Last, the area just in front of the fridge cabinet shown in the picture above was addressed. So, from the seat partition rear-wards is pretty much done. That leaves the cab still to do. All of the seams need to be taped as well, but that should be much easier once the rest of the Heat Wave and MLV have been laid out.

In case you were wondering, that reflective material in the wall behind the fridge is foil-backed "RV insulation" that I had put in there years ago. It actually does reduce the temperature inside the bus a little bit, or it did when I first installed it. Since it does not fill the cavity, I will probably just leave it and put the Mega Zorbe or Jute right on top if it. There is the old Q-Quiet vibration absorber on those panels as well, and I will just leave that be. I am running out of material, and this project has taken quite a bit of time already.

That's it for today. Thanks, as always, for following along-

Hapy Noises - Part 2

Today's post continues the thinking-through-my-fingers on noise containment in Hapy, the 1972 camper microbus powered by a turbo diesel engine from a 1998 VW Beetle. This time, we look at the layers of noise control which appear on top of the vibration dampener I mentioned in part 1 (See Hapy Noises - Part 1). Then, I cover some decibel collection I did so we have a starting point for measuring success or failure of the various layers.

Sound Absorb

Second Skin Heat Wave (Jute)

With the Constrained Layered Dampener (CLD) chosen, I wanted to add noise absorption stuff as well. In my research, I arrived at 5 different product types that could help. There are plenty of other choices, but I think by combining these, I will achieve my noise control goal. These act differently and have different strengths, so some are better in different locations and some are best used together.

Jute Thermal - absorbs sound, but is really more for temperature containment. This is similar to old-skool car insulation, but it is lightweight, and actually does absorb some noise. I intend to put this (HeatWave) over the engine compartment, under the rock-n-roll bed (on the floor) through the main cabin, under the front seats and to the front kick panels. It will act as a 3/8" thick carpet pad while keeping road noise and heat down. The Heat Wave product is designed to resist mold, etc, so even here in the rainy Pacific North West, it is safe for use. Of course, getting it soaking wet is not advised, so some thought about where it should be placed is suggested.

Second Skin MLV

MLV (Mass Loaded Vinyl) - blocks the transfer of airborne sound. Some products have integrated foam to absorb sound as well. These sheets are heavy for their size, and folks have expressed difficulty in getting them to hang vertically (like on a wall) or horizontally (on a ceiling, eg) permanently. I intend to use this stuff (Luxury Liner) from tailgate to toe-kick: under the seats, the rock-n-roll bed and the entire floor on top of the jute. By having the jute layer in-between, the MLV is more effective. The audio guys use this stuff behind their door cards too, but I'm not sure I'll have the money nor patience for that. The luxury liner is 1/8" thick, so, anywhere this goes on top of the Jute, the carpet will effectively have 1/2" of pad, which would be a little thick anywhere people are walking around, but the MLV will effectively stiffen the underfoot feel. It won't feel like you're walking on a 1/2" cloud. Using it under the seats will be fine, provided it doesn't interfere with the operation of the seat. This stuff is super important for noise containment; the audio experts / competitors swear by this stuff. I am toying with the idea of creating a "box" of this stuff inside the engine compartment to really cut down on the transfer of the engine-din. I'm not sure how that would all work. Again, thinking thru fingers.

Second Skin Mega Zorbe

Acoustic Foam - There are literally hundreds of different acoustical foam in all kinds of contents, shapes and sizes. Most are not all-weather-safe, and some are rather thick, so the options for smaller spaces where water or humidity might appear reduce that list from hundreds down to tens. That's still a big list. I plan to use this stuff (Mega Zorbe) in the doors, walls and in the ceilings. Of the materials, this is the most expensive, but it may be the most effective in dropping the overall noise. It is 1/2" thick, so anywhere it goes it could make it's presence known, so I can't use it on the door cards, for example, but it can go into every body cavity. There is definitely 1/2" between the wood drop ceiling and the steel, so even after that wood is removed, we would not lose headroom by adding the Mega Zorbe to the ceiling. Omitting the wood drop ceiling removes a sound reflecting surface I would otherwise need to deal with as well.

Open Cell Foam - I will be putting something on the door cards. I may use an open-cell foam product. Second Skin doesn't offer anything, but that's because there really isn't a product that will meaningfully absorb sound that's less than 1/2" thick (aside from the jute) and they are in the sound control business. I want something that will soften-to-touch and maybe reduce some of the sound reflection. For that, any old thin foam would do, because pretty much any soft surface that's, like, 1/8" thick will reduce the sound reflection. Or so they tell me. I will probably order something from McMaster-Carr (like this) or foam by-the-yard (like this) from FoamMart unless I can find something from an automotive fabric place. Regardless, any time we use open-cell foam, it is best to put a moisture barrier material between the foam and fabric (like this) to protect the foam from moisture as well as protect the fabric from rubbing against the foam. I haven't decided if/how I will address the hard cabinet sides. They might get the foam/fabric treatment just like the cards.

Closed-Cell Foam  - Similar to the open-cell foam in terms of flexibility and possible uses, but closed-cell foam is not permeable, so water can't get into it, damaging it. Closed-cell foams are usually stiffer and do not absorb sound as well. I may use closed-cell foam for the door cards instead of open-cell, if I can find something that acts like an open-cell, but is water-resistant (like this Volara or this Ensolite). There are closed-cell products designed for placement between your plastic bits and the steel. I had thought about getting some for between the cards and the body. I still may, but the rough cost of this effort is already a bit nutty and I can always do that later. For now, whatever foam I put on the door cards (under whatever fabric I put on there) may simply wrap around to the backside. Hopefully, that will contain the card-to-body rattle. Of course, I may just omit foam and fabric on the cards if money and time are not available.

 

Noise (dB) Level Tooling

Before we do any of this great stuff, I want to know where we are starting, so we can measure and celebrate the gains. My sitting-in-seat frame of reference for noise: I needed to run my stereo at 25 or above to be able to hear it at all over the noise of driving down the highway on the way back from Hapy's birthday (See Hapy Turns 50). I had the volume at 30 for loud-enough-to-sing-along (but not entirely taking over) volume. That is not at all scientific. So, in the Fall of 2021, I set aside the new driver seat install, took Hapy for a spin with the old seat and recorded these values for future comparison. My "tooling" was a free iPhone application called "Decibel Meter". I chose it because the free version allowed me to save many recordings, complete with graphs, for looking back upon. I was also able to email myself 5 minutes worth of detailed data. None of the other apps which also have a reasonable scale for the dB provide those features for free. I tried a few apps I won't name which had completely inaccurate dB readings. I used "Decibel X" as my control for comparing other apps. Decibel X is great, and very accurate, but you can't record more than 1 session, nor share anything without a hefty fee. I would have used it for my tests, but I would have needed another person in the bus, watching the meter, writing things down. That's silly.

I thought about buying a measuring tool for better sampling. I decided that I was already uneasy about investing a considerable amount into solving the problem, and chose not to spend more on a measuring device. Having better calibrated results, and more accurate frequency levels would have been great. The issue is that a decent point-in-time tool (Extech 407730) is at least $80US. An OSHA-qualified tool (REED Instruments R8050) is closer to $200US without data logging, while a fully functional tool (like Extech 407750, or REED Instruments R8080) can be over $300US. Plus, you need an additional device to periodically calibrate the instrument (for another $100US or more). I'm sure these devices are better than my free iPhone app, but I need that cabbage to cancel the noise first. Having said all that, if this sound control thing is a fun project, I may get a good tool so I can better measure before and after on future car projects. The trick will be finding one that is reasonably priced, accurate, can be easily re-calibrated and has a logging function so I can collect 30+ minutes of data without having to write things down as we go.

Noise (dB) Level Tests

For the test, I idled in the front driveway, and then took our usual route to the highway (OR217) from home, and drove a loop (OR217 - US26 - Murray Blvd) back home. This got multiple samples of idling, around town and byways. I only got one real highway sample, but I think it was long enough. I got up over 65mph for a short stretch, long enough to collect a sample.

Decibels from that long test are below. One important observation throughout all of the tests: the dominant frequency of the noise appeared at the very bottom of the spectrum. This could be a reflection of the tool not reading frequency well, or my noise problem is 95% below 100Hz. Some comparisons on my couch between the 2 apps indicated that Decibel Meter reads Hz low when compared to Decibel X. Neat. I chose to ignore the Hz, and accept that I can only really measure overall dB. On to the dB readings! In all cases, the value moved around within a range as well as had an overall peak:

dB test as a graphic

Sitting at idle (900 RPM): 60dB

Around town (less than 40mph): 65dB +/- 5dB. 75dB peak

On the byways (above 40 but under 55): 65dB +/- 10dB

Highway speeds: 70dB +/- 10db, 83dB peak

Over the course of the 30 minute test, noise was consistently between the mid 60's to mid-70's with spikes up into the mid-80's.

For fun, I compared against Boo's 2009 Audi A4 (GoRo) on my old standard test-drive loop, and then ran Hapy on that same test loop. This loop does not include a highway, but it does include idling, a residential street and a 4-lane "by-way". The Audi ran from the upper 40's to the lower 60's. Hapy on the same loop was in the low 60's up into the lower 70's, topping at 75dB.

Noise (dB) Level Test Thoughts

In simple comparison, GoRo was in the lower 50's most of the time and Hapy was 10dB louder or more. "10" seems like a small number, but I consider these 2 things: First, the range for both vehicles remained within about a 10dB span with Hapy's quietest/lowest readings starting at the loudest/highest for GoRo. Second, a modern refrigerator or a moderate rain is 50dB and hearing damage starts at 85dB, so a 10dB difference in this range is actually huge. I think it is kind of like 10* difference in Celsius: 23*C (73*F for US folks reference), is totally comfortable, like a conversational 60dB. But 33*C (91*F) is intrusive-hot, like a running vacuum cleaner 70dB. Another 10*C (43C or almost 110*F) is quite uncomfortable in practically all climates, like listening to a whistling kettle (80dB) for an extended period of time would be. For some real-life decibel level references, I found this link to be especially fruitful. Last, I think it is worth noting that the highway was much louder. My usual shift point is around 3200 RPM, and I cruised the freeway around there as well. So, maybe the noise gets much louder above 3k RPM -OR- the increased speed leads to more noise (wind or tire). I can't know for sure without more tests, but I think resolving the highway noise will be my biggest challenge. It would be the greatest reward for longer trips, if I can solve it.

EDIT: I added the image just above  on the right on 2022-Mar-29. I sourced it from a research paper about the dominant frequencies produced by a 3-cylinder diesel engine. While this may not be an exact match for a 4-cylinder TDI, I think we can draw sufficient parallels to lead us to focus on frequencies above 250Hz, but especially around 1k. Based on the material sheets for the items described above, all of these should make a meaningful impact on containing the noise.

That's it for today. I have more planning to do, and things to purchase. If all goes as I hope, the noise control efforts will start soon. My goal is for Hapy's entire noise range to drop by at least 5dB on the test loop and for the mid-80's highway peaks to drop by at least 10dB.

Thanks, as always, for following along-