Dust Quiet Extractor System

In this project we turn a bunch of old free stuff, including two old household vacuums into what is arguably the most useful and necessary of workshop tools: the dust extractor. But why stop there? Lets make a really fantastically effective dust extractor, one that is whisper quiet, never stops sucking or plagues you with blocked filters, one that is versatile enough to take dust from a variety of power tools, one that turns on and off on its own so you never forget, and most important of all, one that does a good job of extracting the small - most deadly - particulates from the air you breath... Step forth, 'The Dust Sniper'.+

Just so people know. I am now giving this contraption away, to the first person that can come and collect it.. See details on https://www.facebook.com/floweringelbow/

This project was borne out of my dissatisfaction with commercially available dust extractors. After a fair bit of research I purchased one of the more expensive 'quiet' workshop vacuums, and was not happy with its performance (I sent it back unused after taking a dB reading of it). In exasperation at the dusty noisiness of it all, and wanting to re-use materials and spend as little as possible, I began the Dust Sniper (DS) project.

This DS ended up costing about £20 total. So it is possible to reused a bunch of stuff destined for landfill and end up with an aesthetically pleasing and useful tool-workbench. And of course we can learn loads about sound, cyclones and dust related jazz along the way. Because the DS's parts are mostly recycled, there is no comprehensive list of materials up front, instead I will give tips as we go along suggesting possible reclaimed bits that will do the job and where you might find them (if you don't care why we chose certain materials and just want a 'scavenging list', check out the last step).

My kingdom for some silent clean air

I'll throw it out there to begin with, most dust extractors are bad. Even the expensive ones, like the Festool, extract a continuing fee, needing regular bag and filter changes to keep working properly. The less expensive, well... lets just say they can be seriously bad for your mental and physical health, as you will find out if you follow along with this Instructable.

The Dust Sniper (DS) is effective and very quiet - the two main goals I had when starting this project. It does, however, fulfil these requirements at a cost. Namely, it is very heavy and big (compared to your average canister style vac), so it won't be perfect for everyone. This isn't necessarily the disaster you might think though. In fact it can be darn right useful if we use the DS as a mobile work surface. That way we will end up with nice clean air, a quiet place to create our mad jazz, and a super sturdy, rollable worktop thrown in! Ideal if you are still setting up a workshop, as I am.

Step 1: Noise Loves Dust

We might not often think of noise and dust being co-conspirators, but they do help each other to cause workshop misery. Dust, particularly for those that do much woodwork with power tools, gets everywhere: in the air, in your lungs, and in the belts and bearings of our tools. Power tools, like an orbital sander, a jigsaw, a planer, or a router create a lot of dust, and without good extraction (sometimes even with it) the quantity of dust that gets into our tool's workings is enough to cause big increases in noise levels. 

Lots of noise is bad. As anyone who reads the FE blog will know, I am particularly fastidious about cutting down on noise (see for example, my quest for the quietest bandsaw). I can think of a load of good reasons for my desire for quiet tools, but probably the most important, and one that anyone using power tools should take seriously, may be gleaned from the following: 

"The first handicap due to noise-induced hearing loss to be noticed by the subject is usually some loss of hearing for high-pitched sounds such as squeaks in machinery, bells, musical notes, etc. This is followed by a diminution in the ability to understand speech; voices sound muffled, and this is worse in difficult listening conditions. The person with noise-induced hearing loss complains that everyone mumbles. High frequency consonant sounds of low intensity are missed, whereas vowels of low frequency and higher intensity are still heard. As consonants carry much of the information in speech, there is little reduction in volume but the context is lost. However, by the time the loss is noticed subjectively as a difficulty in understanding speech, the condition is far advanced." (p146 Engineering Noise Control)

Ok, so dust often equals more noise. How ironic that adding a dust extractor can be so noisy then. Lets leave 'noise' at that for now - for more noise related background and nerdy theory, checkout step 3.          

Dust is a serious problem. 

Actually aside from helping along hearing loss, dust can cause bigger problems. At this point I am going to go ahead and assume that everyone is comfortable with the idea that dusty lungs are bad and to be avoided. The problem is most people don't realise just how dangerous dust is, especially to us lone inventors, DIYers, and makers, who do not have the protection of government legislation, which enforces air quality standards* in factory and workplace environments.  

At home, people tend to use cheap and ineffective extraction systems and/or pathetically inadequate masks (or no protection at all). I must admit from time to time I have been guilty of this, not wanting the noise of the vac or being in a rush - very bad! The precautionary principal should definately apply here. Particularly until you have finished your DS, a good dust mask, goggles and ear defenders are your friends! For more info on dust and health check out this post on The Dangers of Wood Dust and this table of wood dust toxicity levels.

*It is interesting to note how these standards are constantly being raised, as more research is done on the effects of wood dust. See, for example, Jette B. Lange, 2008 "Effects of wood dust: Inflammation, genotoxicity and cancer"

Step 2: Cyclonic Filtration - Overview

The details, instructions, and measurements of the actual cyclones have been up on the Flowering Elbow website for some time. Rather than have them repeated here, check the cyclone build guide steps on flowering elbow. 

Step 3: Noise Reduction - Background Information (and Some Theory You Can Skip)

So when I was building this, I spent a while researching about noise, sound and jazz, if you're into that kind of thing too read on... If you are a sound wizz already, look away. If you are bored easily and/or just want to hammer things together you can skip it too.

So what Is sound anyway?

Every school kid will tell you sound is basically the result of things getting excited and vibrating. Everything vibrates anyway, but if you do something like hit the table in front of you (assuming there is one), you change the way it vibrates and it passes on that vibration to the air. In the air we can imagine a series of high and low pressures, which in turn vibrate the internal gubbins of our ear - causing us to register what we call sound.

As with most subjective things in life we humans like to try and measure and quantify these vibrations. The quantity most often used to measure the “strength” of a sound wave is the 'sound pressure level' (SPL or sometimes Lp, not to be confused with 'sound power level') measured with respect to a standard reference pressure of 2 ×10−5Pa.  

SPL is expressed in dB (or Decibels) which are a logarithmic unit, so that for every 6 dB decrease in volume, the sound is perceived as being half as loud. 

Blocking out Nasty Sound (noise)

When an airborne sound wave encounters a solid blocking its path, it effectively bashes into it, the disturbance causes the solid to vibrate. This vibration is transmitted through the solid. Now on the other side, the surface acts as a new emitter by disturbing the air and producing a new sound wave. By this process the sound effectively passes through the barrier. The efficiency of the transmission depends on the physical properties of the solid in particular, its mass.

How much of the sound is blocked out? Well, if L1 dB is the sound pressure level on the noise source side of the partition and L2 dB that on the other side, then the Sound Reduction Index (SRI) or Transmission Loss (TL) is defined as:

TL = SRI = L1 - L2  dB  

The transmission loss, or SRI, varies with mass and frequency. In general the higher the frequency the better the sound is blocked, hence the higher the SRI will be. There are exceptions to this when partitions start vibrating at their resonance frequencies. More on that later, for now all we need to know is that:  

1. For precision work (or for special noises with a particular frequency content), the SRI index is quoted for particular frequencies, normally in octave bands.

2. For many purposes and for convenience, the SRI is quoted as a single number, which is the average SRI between the frequencies 100 – 3,150 Hz. The resultant sound level is then quoted in dB(A). (A) presumably standing for average. 

The Mass Law

The so called 'mass law' simply states that by increasing the mass of a partition, we increase the transmission losses or SRI of the partition proportionally. So mass is generally a good thing when we are trying to reduce sound (think about the useful properties of lead). The mass law however only applies to a given material, over a specific range of frequencies. It could be, for example, that a deep bassy noise (low frequency) travels through a panel with very little reduction in volume even when you increase the mass of a panel. Indeed it is often the case that low frequency noise transmission is more effected by the stiffness of a material.

Again, this all depends upon the material in question. A lead curtain's behaviour, for example, is essentially mass-law controlled over the entire audible frequency range. For a more geeky explanation along these lines, check out "Engineering Noise Control: Theory and practice, Fourth edition, David A. Bies and Colin H. Hansen (2009)" 

For us, the mass law is a good demonstration of the compromise we are going to make between light weight and sound reduction. "[We] should rule out the use of low density fibreglass (such as insulation batts used in house ceilings), as well as typical polyester blankets. In fact polyester blankets are likely to be completely ineffective." (Bies & Hansen, 2009 p 386). Although if we can compress them a lot and have them to hand anyway, it is a different story...

Building less symmetrical and more random please

As with double or triple glazing, it is important not to have all the panes the same thickness, as this accentuates the dip in the TL (transmission loss) curve at critical frequencies. The same goes for our purposes when we construct a double wall box. It is better to use different materials as well as thicknesses for the different layers. That way we will block out a broader range of frequencies.

While preventing resonance by mixing materials and shapes is good, it is also well worth incorporating an air (or foam) gap, which prevents the direct transmission of vibration. Vibration is easily transmitted to other materials by mechanical coupling - avoid if possible. 

"Acoustic isolation is generally accomplished by providing as wide a gap between the panels as possible and by filling the gap with a sound-absorbing material, while ensuring that the material does not form a mechanical bridge between the panels." (Bies & Hansen, 2009, page376)

Absorbing Sound

The nature of the surfaces on which the sound wave falls determines how much will be absorbed. Hard rigid non-porous surfaces like glass, marble or concrete, provide the least absorption and are thus the best reflectors. Soft porous surfaces and those which can vibrate absorb more of the sound. When sound energy is absorbed it is converted into heat energy, but this energy is very small so no need to worry about overheating caused by sound.

The amount of sound absorbed is proportional to the area of the material concerned. So if S is the sound absorbed and A is the area of the exposed material, we can say that S is proportional to A. In general this means that rough surfaces are better at absorbing than finely finished ones. Further,

 S = aA

where:  a is the Absorption Coefficient.

The Absorption Coefficient is a number always less than 1 (because it has no units, it is a ratio) and is small for a material that reflects most sound and large for a material that absorbs most of the sound incident upon it. It is determined by the amount of sound absorbed by a material divided by the sound energy arriving at the surface (so a = absorbed sound energy / incident sound energy). Just for interest the table below (from the Sengpielaudio website) shows a bunch of absorption coefficient values for various materials. As you can see, different materials are better or worse at absorbing different frequencies.

Floor Materials	125 Hz	250 Hz	500 Hz	1000 Hz	2000 Hz	4000 Hz
concrete or tile	0.01	0.01	0.15	0.02	0.02	0.02
linoleum/vinyl tile on concrete	0.02	0.03	0.03	0.03	0.03	0.02
wood on joists	0.15	0.11	0.10	0.07	0.06	0.07
parquet on concrete	0.04	0.04	0.07	0.06	0.06	0.07
carpet on concrete	0.02	0.06	0.14	0.37	0.60	0.65
carpet on foam	0.08	0.24	0.57	0.69	0.71	0.73

Seating Materials	125 Hz	250 Hz	500 Hz	1000 Hz	2000 Hz	4000 Hz
fully occupied - fabric upholstered	0.60	0.74	0.88	0.96	0.93	0.85
occupied wooden pews	0.57	0.61	0.75	0.86	0.91	0.86
empty - fabric upholstered	0.49	0.66	0.80	0.88	0.82	0.70
empty metal/wood seats	0.15	0.19	0.22	0.39	0.38	0.30

Wall Materials	125 Hz	250 Hz	500 Hz	1000 Hz	2000 Hz	4000 Hz
Brick: unglazed	0.03	0.03	0.03	0.04	0.05	0.07
Brick: unglazed & painted	0.01	0.01	0.02	0.02	0.02	0.03
Concrete block - coarse	0.36	0.44	0.31	0.29	0.39	0.25
Concrete block - painted	0.10	0.05	0.06	0.07	0.09	0.08
Curtain: 10 oz/sq yd fabric molleton	0.03	0.04	0.11	0.17	0.24	0.35
Curtain: 14 oz/sq yd fabric molleton	0.07	0.31	0.49	0.75	0.70	0.60
Curtain: 18 oz/sq yd fabric molleton	0.14	0.35	0.55	0.72	0.70	0.65
Fiberglass: 2'' 703 no airspace	0.22	0.82	0.99	0.99	0.99	0.99
Fiberglass: spray 5''	0.05	0.15	0.45	0.70	0.80	0.80
Fiberglass: spray 1''	0.16	0.45	0.70	0.90	0.90	0.85
Fiberglass: 2'' rolls	0.17	0.55	0.80	0.90	0.85	0.80
Foam: Sonex 2''	0.06	0.25	0.56	0.81	0.90	0.91
Foam: SDG 3''	0.24	0.58	0.67	0.91	0.96	0.99
Foam: SDG 4''	0.33	0.90	0.84	0.99	0.98	0.99
Foam: polyur. 1''	0.13	0.22	0.68	1.00	0.92	0.97
Foam: polyur. 1/2''	0.09	0.11	0.22	0.60	0.88	0.94
Glass: 1/4'' plate large	0.18	0.06	0.04	0.03	0.02	0.02
Glass: window	0.35	0.25	0.18	0.12	0.07	0.04
Plaster: smooth on tile/brick	0.013	0.015	0.02	0.03	0.04	0.05
Plaster: rough on lath	0.02	0.03	0.04	0.05	0.04	0.03
Marble/Tile	0.01	0.01	0.01	0.01	0.02	0.02
Sheetrock 1/2" 16" on center	0.29	0.10	0.05	0.04	0.07	0.09
Wood: 3/8'' plywood panel	0.28	0.22	0.17	0.09	0.10	0.11

Ceiling Materials	125 Hz	250 Hz	500 Hz	1000 Hz	2000 Hz	4000 Hz
Acoustic Tiles	0.05	0.22	0.52	0.56	0.45	0.32
Acoustic Ceiling Tiles	0.70	0.66	0.72	0.92	0.88	0.75
Fiberglass: 2'' 703 no airspace	0.22	0.82	0.99	0.99	0.99	0.99
Fiberglass: spray 5"	0.05	0.15	0.45	0.70	0.80	0.80
Fiberglass: spray 1"	0.16	0.45	0.70	0.90	0.90	0.85
Fiberglass: 2'' rolls	0.17	0.55	0.80	0.90	0.85	0.80
wood	0.15	0.11	0.10	0.07	0.06	0.07
Foam: Sonex 2''	0.06	0.25	0.56	0.81	0.90	0.91
Foam: SDG 3''	0.24	0.58	0.67	0.91	0.96	0.99
Foam: SDG 4''	0.33	0.90	0.84	0.99	0.98	0.99
Foam: polyur. 1''	0.13	0.22	0.68	1.00	0.92	0.97
Foam: polyur. 1/2''	0.09	0.11	0.22	0.60	0.88	0.94
Plaster: smooth on tile/brick	0.013	0.015	0.02	0.03	0.04	0.05
Plaster: rough on lath	0.02	0.03	0.04	0.05	0.04	0.03
Sheetrock 1/2'' 16" on center	0.29	0.10	0.05	0.04	0.07	0.09
Wood: 3/8" plywood panel	0.28	0.22	0.17	0.09	0.10	0.11

Miscellaneous Material	125 Hz	250 Hz	500 Hz	1000 Hz	2000 Hz	4000 Hz
Water	0.008	0.008	0.013	0.015	0.020	0.025
People (adults)	0.25	0.35	0.42	0.46	0.5	0.5

So you get the idea.  Armed with all that knowledge you are ready to scrounge up some free materials and get building, right?

Step 4: Enclosure Construction - Material Choice & General Notes

On average, people in the UK trade their kitchen in for a new model every four to five seconds (I may have made that up) - that is a whole lot of kitchen worktops being thrown out and replaced. Indeed, composite wood counter worktop seems to be one of the most commonest things to pop up in skips here there and everywhere. That was of course, until I thought of using it to make part of the enclosure - being MDF'esque, dense, stiff and heavy, it should be a useful material for sound proofing. After looking and incredulously not finding any kitchen worktop for some time ("Credit crunch curtailed peoples propensity towards kitchenocide, discuss."), freecycle came up trumps and delivered an ample bounty of fire door material.

When I went to collect the two freecycle fire doors, they were actually getting rid of four of them (nice big heavy strong composite things), some kitchen worktop, and some useful bits of hardboard too - score! I ended up using some of these bits to make the DS and having plenty to spare besides. 

When it comes to making sound enclosures, those 'audiophiles' and DIY speaker builders are somewhat ahead of the game - by that I mean they are quite happy to try experimenting with unusual materials and techniques and also perfectly willing to share their experience and knowledge. We can learn a fair bit from their build techniques and material preferences.

Here are some things that speaker enclosure makers experiment with that you might be able to scavenge or otherwise get your hands on:

• Plywood (without voids is best), mdf, hardboard, etc. All these laminated sheet materials are rigid and easy to construct into airtight enclosures - look out for them turning up in skips.

• Plasterboard - Used extensively in construction, can be laminated with acrylic latex-silicone caulk to provide very effective damping.   

• Sand. This is well known as a good dampener of sound, I used some of this in the DS and also to damp my bandsaw. Best of all it is free if you know where to look (a beach might be a start, though in the UK it is technically not legal to just take stuff off beaches).  

• Oil based plasticine. This is the stuff that never really dries. I have no personal experience with this, but apparently it can be rolled into flat sheets and adhered to panels to damp sound.

• Scrap steel, can be used to stiffen up panels, and to change their resonant frequency. angle iron makes excellent bracing because it can easily be screwed (and glued with damping glue).

General construction points:

1. Use lots of glue to make joints air tight. The reason for this is twofold: one, so that we can control the flow of air leaving the vacuums and make sure it is filtered and clean, and two, so that no sound escapes. Even little cracks can make a big difference to the sound reduction index on an enclosure - think about a car window - opening it just a little makes a big difference to the noise you can hear outside.    
2. Ensure straight well fitting edges - all gaps must be filled. 
3. MDF and chipboard resonate at averagely 150-400 Hz, with the strongest resonances usually at 250-300 Hz. All materials when they vibrate produce sound waves, so If we don't brace it properly we may have small movements in sides of the DS, but because of the area involved even this could be quite loud. 
4. We need to treat both structural borne sound (so called 'impact sound') and airborne sound. The first involves mechanically isolating any sources of vibration with the main structure of the enclosure. The second, ensuring that we have good rigidity and mass.
5. As I already mentioned, when we add bracing to panels, we want to divide up the various panels into sections of unequal area. If not there is a chance that you will have several panels with a common resonance frequency that will combine (and sound loud).

Step 5: The Inner Enclosure

The inner box will not be bearing much weight, and to make the space usage sensible, it is not a massive construction of fire door or kitchen worktop material. Check out the photos for build ideas. 

A Note on the Enclosure and Heat

"But won't the motors overheat if they are in an enclosure," I hear you cry. Hold on there, vacuum motors are something of a special case when it comes to cooling. They blast all the air that they suck in through the motor windings (after passing it through a filter to remove the dirt). So long as any subsequent filters (post-motor filters) remain unblocked, this system works perfectly, and means that vacuum motors can be much smaller than they would otherwise be, and wrapped in a convenient insulative plastic case. Incidentally, this is why vacuum motors make very poor motors if we try and re-purpose them for anything other than air moving applications.

For the DS this means that we need to keep a reasonable exit path open for the air being pumped out of the motor, and that we can expect warm to hot air to be travelling this path (step 11 & 14 deals with this). But it also means that we don't have to worry about trying to blow in cool air to pass over the motor, the vacuums themselves do a very good job of that already. Almost all vacuums are fitted with a heat sensitive safety switch, that will cut power if the motor is overheating. If yours has not, it is probably worth adding one, or finding a different vacuum to use. 


MDF warning:

MDF is typically about 9% urea-formaldehyde resin, it is the stuff that bonds it all together. When we cut it to size we effectively pump out a load of particles of this stuff.  Dust is a big MDF hazard (read the first few steps for the lowdown on dust badness).  But there is another consideration, particularly if you are sensitive to formaldehyde, and that is the long term 'off gassing' MDF does. Formaldehyde-free MDF does exist, but if we are scavenging our materials one must assume the worst. In this design the 'off gassing' will hopefully be less of a problem as the inner box will be sealed in. In general though, you can control these emissions by finishing the surface with a veneer or a sealing paint, and this is a good practice whenever you make MDF things that will be in living areas.  

Lead warning: 

Lead is great! It can practically be 100% recycled, has fantastical blocking properties, and is comic book style heavy. Lead is not good however, inside the human body! A tiny bit inside, is way more than we want. Luckily it only really gets in there if we are careless. It is best to handle the stuff with thick gloves - you don't want to cut yourself with lead!  Wash hands before you eat after handling the stuff. Do not do anything that creates lead dust, unless you have the ultimate dust extractor (presumably you wouldn't be making this in that case!), are wearing a quality ventilator and goggles, and have a way of properly disposing of the dust.  I would advise against doing anything that might make lead dust, and really you don't have to because it is so soft - it cuts with tin snips. Don't be tempted to melt it, unless you have the correct safety equipment - the vapour is another way it can get inside you.

Step 6: The Inner Enclosure's Tortuous Path

After making a nice sealed box the problem is that we need to allow air to flow in and out (so that the vacuum can suck stuff up and vent its exhaust air!). If we have holes for the air to go in and out, it is a safe bet that noise from the vacuum's air chopping impeller is going to maliciously exploit them and fire sound out at you. That is, unless we create an elaborate maze in which the noise will get lost (bwahahaha), but which our friendly air will have no problem traversing. People (that is, an author of one of the more obscure books I read) sometimes refer to such a system as a 'tortuous path'.         

When we incorporate obstacles into the air stream, we add resistance to its flow. To maintain necessary airflow, most silencers have to increase the cross sectional area, so enough air can run through - making them quite bulky. This tortuous path or baffle system has the same problem.

There are many different designs to reduce sound that is transmitted through airflow passages: reflective, reactive, diffusive, depressive and active. Quick and concise description of different types of silencers can be found here.

For the dust sniper, the back of the inner box is where I made the baffle arrangement. I wanted to keep the two exhaust streams separate so it consisted of two paths, created by fire door off-cuts (produced while making the outer box). It ended up being damped by a sealed off panel of sand (see the pics and descriptions for build info).    

Another consideration is that sudden changes in air pressure can be noisy - the sound of a vacuum usually increases when we put a crevice nozzle on the end for example.  We can extend the changing pressure gradient though, by breaking the exhaust stream into a series of outlets - the same style of thing that you see on a big motorbike exhaust with lots of holes in.  

"Such a device has been shown to accomplish by itself, without any additional muffling, a 10 dB insertion loss in broadband noise in a steam-generating plant blow-out operation." (p434).

So that seems like a good idea, assuming the air coming out is making much noise...

Step 7: Assembling the Inner Enclosure

Lets put this inner box together (see photos).  The main challenge here is to make a very snug fitting front panel, which will have a tight seal, preventing sound from escaping.  

Step 8: The Outer Part 1 - Housing

The outer housing wants to be quite robust, as it will be functioning as a worktop/ multi-use-surface. This is all good as I have a few thick, heavy fire doors to make it from. As a bonus when we make it massive, we are also helping to cut down the noise. I also have a quite delightful bit of scavenged teak to go on top (It was thrown out by my university's science department and matches my current workbench, which has a similar ancestry).    

As it is going to end up on the heavy side of hefty when it is all together, we are going to want some casters to get it mobile (ish - well like a gigantic lumbering titan of a thing at least). As with the inner box, it also needs to be as sealed up and tight as possible, with no weak points for sound to leak out from.  

To begin with I sized it up, based on the Sketchup 'plan' and got cutting. Being way too big to manoeuvre the doors through the bandsaw, and after some 'interesting' jigsaw antics, I borrowed my friend's mighty fine circular saw for the job. This worked very well in combination with a clamped on bit of wood, that I knew was straight, as a guide. 

Similar to the inner box. We can prepare the individual pieces as best we can before sticking them together. It is a good idea to leave the lid off for easy access until after we have finished all the insides.

Step 9: The Outer Part 2 - Air Exit and Filtration

After all that effort, we don't want to just blast out the air into the atmosphere. It is a happy coincidence that a filter, as well as removing very small particles that can kill us, is also a useful sound deadening material for the exhaust passage. This time fortune really did smile upon us, and a perfectly sized, HEPA filter fell from the gods (well, from our friend who works for a big pharmaceutical multinational who thinks nothing of skipping anything that is not made of purest, unmarred angel essence) into our gratefully receiving lap. This is not strictly necessary, as the vacuums have filters (though they are fairly pathetic in comparison) but certainly a welcome boon.

HEPA stands for 'High Efficiency Particulate Absorbing (or Arrestance, or even Air filter, depending on who your talking to)'.  Basically is is pretty much the best commercially available air filtration of the sort that relies on the air passing through a fine mesh which catches the dust particles.  

"The HEPA standard exceeds the MERV specifications because they are the only mechanical air filter with an efficiency of 99.97% at 0.3 microns. This makes them at least 50% more effective than other types of mechanical air filters." http://www.hepafilter-pro.com/Filter_Ratings.php  (For the grades of filter : http://www.filtration-engineering.co.uk/air_filter_testing.htm). 

So we have one of these babies to integrate into the DS. As this will be right by the air outlet, it needs to be heavily soundproofed to make our other efforts worthwhile. Now is the time to deploy the lead!

Step 10: The Outer Part 3 - the Forbidden Cork Forest (or Air Intake Sound Proofing)

Unfortunately, despite the sound having to travel against the airflow created by the vacuums, the gaps needed for the air to come in will still leak a lot of sound.  As MahavishnuMan told me when I foolishly suggested the sound might be reduced by the inrushing air:    

"...in order to "suck up the sound", your vacuum would have to breathe in air at a higher velocity than what sound travels, which is 1,127 ft/sec at sea level and standard temperature and air pressure. Not only am I positive your vacuum doesn't suck at Mach 1, but if it did you would have a sonic boom loud enough to crack the box."

So yeah, we need to 'treat' the air inlets so that we defeat the escaping sound. For this purpose I am experimenting with what I fondly dub the 'Forbidden Cork Forest'.  It is crucial that we don't add much to the air resistance, which is tricky when you are introducing obstacles to block sound.  

The idea with the cork forest is that it will block sound with the sound absorbing 'trees' (the corks), while still presenting a smooth round aerodynamic surface for the air to pass through with minimal turbulence.

Step 11: The Cyclone & Dust Cabinet

As the cyclones are transparent, we want to allow light into their area so we can see what is going on. I used some more of the nice 15mm thick acrylic for the side panels (see photos).

We can continue the theme of oak wood facing and make the control board and corner strut from some lovely scrap oak.

The cyclones themselves are the tallest part of this construction, and because we want the finished work surface to be no higher than our workbench, they need to sit below the level of the DS's floor. This is not a problem because the castors raise it high enough that I can still unscrew and empty the dust containers.

Step 12: The Top and Front

Because we are using the top of this DS to make our cool stuff, it needs to be nice. Some solid reclaimed teak will do nicely. I blogged about the origins of the teak worktop as I was doing it so I don't want to repeat it now. Enough to say it will be sturdy, help damp the sound, and provide years of service. To attach it, we don't really want to have a solid mechanical link, but instead use silicone-acrylic-latex caulk to bond it. This provides much better damping than screwing it.    

The front of the DS gets some handmade catches - which are actually really simple - to hold it in place and compressed against its bubble seal.   

Step 13: The DS Auto Switch

So we want a switch that turns the DS on automatically when we start up our power tool, and then turns it off again when we are done making dust. Ideally it will switch on just after the tool, so that the starting current of the power tool is not exacerbated by the simultaneous starting of the dust extractor. And when the power tool turns of, we want the DS to stay on for a few seconds so that the hoses are cleared of any remaining dust. 

There are a load of different ways to approach the auto switch circuitry. Here, for example, is an auto switch that uses commercial current detector. The cheapest I could find the toroidal sensor component for  this was for $50 so this was out as far as I was concerned. I wanted to make it without buying anything, using the odd bits and bobs I had knocking about, so my design was a bit, hum, unusual. There are a load of alternative ideas for this in the resources section at the end.  

If you want to try out my design, instead of using a current transformer, which seems the standard approach, we use a reed switch, activated by a very small coil in the live power cable that supplies the power tool. The reed switch activates a relay, which in turn energises the heavy duty contactor, which is hefty enough to cope with switching both vacuums on or off at once. If you want to do it this way follow the circuit diagram below, nothing is too complicated, expensive or difficult. 

To make the coil, just wrap some 16 AWG (or fatter) magnet wire round something thin that is a similar size as the reed switch. I used the blank end of a drill bit, but be careful not to scratch the enamel insulation (something plastic or wooden is better). To begin with I didn't even use magnet wire, just standard insulated wire, as you can see in the pics, and it still worked OK. This way is not as sensitive though, because of the insulation gap. So if you want to use it with lower current draw tools as well, magnet wire is better.  About 13 turns is all you are likely to fit on the reed switch - that's fine.  

The capacitor bank in the 6V relay circuit adds a delay to the switch, so that the DS stays on for a few seconds after the power tool is turned off. Having a capacitor bank like this though, means that we need to add a resistor (or around 8KOhm) in series with the reed switch, to protect it from the inrush current when it is switched on (without it the reed switch will weld shut). If you wanted to be a bit more elegant you could put together some kind of 555 time latch circuit, but I didn't have any 555s to hand.  

If you just want a very quick and dirty solution, just having the reed switch activate the AC contactor worked OK when I tested it. You will not get any delay, and the motors will start together with this one, but it is very simple and it works (though how long the reed switch would last I can't say).   

Components (circuit diagram below click the 'i' in top left to get full size):

Reed Switch - just one of a bunch I had laying round. It is about 1" in length, glass body, the coil goes tightly round this.  
D1 - rectifying diode
D2 - rectifying diode
TR1 - a small step down transformer (to 6V) - time to use one of those 'wall warts' you have been saving. 
C2 - 6.3V 4700uF (but just use what you have in your scraps box)
C4 - 6.3V 4700uF
C5 - 6.3V 4700uF
DC Relay - 6V DC relay, a smallish low current thing is what you want.
Contactor - Heavy duty contactor, I found this on a thrown out saw - these are useful for NVR applications.
R2 - 8.2KOhm resistor, important for protecting the reed switch.
C3 - a 0.22uF 275V AC capacitor
R1 - 330 Ohm resistor  
B2 - a suitably beefy bridge rectifier 

Step 14: Controls and Wiring

Ok so here is where we end up with plenty of head scratching, checking, double checking, and re-checking again. Remember that this is mains voltage we are tinkering with so get some qualified help if you need it.

The control board is basically going to consist of:
a dumb plug socket (just a plain socket),
a control socket (what we plug the power tools into, if we want auto dust extraction)
A master on/off switch (this turns everything on or off)
A switch that toggles between on/off/auto one of the vacuums 
A switch that toggles between on/off/auto the other of the vacuums

Begin by making some real size sketches of how it might be on card and work from there. Of course, this control board would be a prime candidate for some laser etching. Anyway, preparing the board can be a classic woodworking task: we need to drill and chisel some holes that will fit our sockets, switches, and air inlets. Make sure to plan it all out carefully based on the switches you have acquired before setting mallet to chisel.    

At this stage I also added in a 16A trip switch (that I was given when a neighbour was replacing their consumer unit). The fuse in the plug should give protection anyway, but a little extra is nice. Once you have everything sorted, and tested carefully rout the cables and secure them so they are all neat and tidy.

Step 15: In Use, Evaluation, Maintenance

Just so people know. I am now giving this contraption away, to the first person that can come and collect it.. See details on https://www.facebook.com/floweringelbow/

Lets evaluate the DS in relation to the design goals which were:

1) Effective removal of dust from hand-held tools and bandsaw
2) Little or no noise
3) Provide strong but wheelable work surface

1) The removal of dust thus far is excellent. The dust is sucked up and separated by the cyclones into the collection barrels. Because the separation efficiency is so good suction remains very high - no regular cleaning filters or changing of bags required. Obviously the collection barrels need emptying occasionally, but being many times bigger than a bag or standard shopvac canister, this is an easy and infrequent chore. Thus far, I have only had to change a vacuum bag after I got carried away and let the barrel become full, which resulted in the dust quickly clogging up the vacuum bag and suction becoming very weak.

So yeah, it might be worth me trying to make some kind of warning sensor that tells me when the collection barrels are approaching fullness to avoid similar problems in future. I already put a viewing window into one of the collection barrels, problem is that the static causes it to be obscured with dust, so that's little to no help. Some of you have already made some good suggestions on how to overcome this little problem in the comments. Of course any other ideas are very welcome...

2) Noise wise I am better pleased that I expected to be. When the DS is all closed, shut up and operational, I can't really hear the noise of the vacuums at all! The noise of the air rushing through the hose is pretty much all that is audible. So jackpot on the sound front. I can't tell you how nice it is to be able to clear up the shop and suck up dust without a loud noise. It makes nice quiet, well balanced power tools more worthwhile ;)

Now bearing in mind there are many differing and complicated techniques of sound measurement, the audiophiles may want to look away now. In a blissfully and probably horrifyingly simplistic manner, I used a mobile phone with an in-built 'sound meter' to do my measuring.

Sound of both vacs out in the open - 85dB
Sound of one vacuum in the open - 83dB
Sound of both in the DS - 61 (but varies a lot depending on where the end of the hose is situated - the air rushing in at the tip is almost the only precipitable noise)
Sound of one in the DS - 55

3) The work surface is nice, functional, and sturdy enough to dance on. I do need to add a breaking mechanism to the wheels, so that I can lock it in place better.

Parting Thoughts

The DS has been a long project for me, with plenty of help, research and tweaks needed along the way. Still, it has come together in the end and with any luck this instructable will help you guys avoid some the mistakes I made. Already a number of you have said you will be making your own DS, so I look forward to feedback, build photos, and areas of development. If it improves the working environment (and health!) of one of you, my fellow makers, hackers and craftspeople, then great!