Moderators: Random Orbital Bob, nev, CHJ, Noel, Charley

 Topic locked
By siggy_7
Questions about dust extraction come up often by new woodworkers. This is my attempt to try and save regular contributors some time spent answering them. Edited 17/11/16 following feedback from other members.

Why have dust extraction?

Most people accept the need for dust extraction when using power tools and machines. However, it’s important I think to understand what our objectives are in dust collection to make informed choices about the right kit. I think there are three main reasons for extraction:
1. Health. Fine dust easily becomes airborne (particularly when generated by machines with high-speed cutters), and then some of this works its way into your respiratory system. This is widely accepted as a cause of respiratory illness. Some materials and some processes are more prone to fine dust creation than others. Also, some materials we work with are chemical irritants, or can provoke allergic reactions. If you care at all about your health, you’ll want to keep airborne dust to a minimum, and the best way to do this is to remove as much as possible from the source.
2. Quality of cut. The work of some tools (mainly thinking the planer and thicknesser) is negatively affected by poor waste removal. Dust and chips get scooped up by the cutterhead and pressed back into the work. The feed rollers also push waste into the newly cut surface, leaving indentations.
3. Mess. Simple really – dust gets everywhere. If you don’t like mess, and if you don’t like having to endlessly clean your workshop, then it’s in your interests to have good extraction. If you’re working in a domestic environment, this is particularly important, since customers/wives have a real aversion to mess in general.

Questions sometimes come up about workshop air cleaners. These are great for reducing the levels of airbourne dust in your workshop. However, they are only removing dust that has already got into the air (and which you are also now breathing). The best approach for dust should always be to remove it at source. Once you have done what you can do to this, by all means consider a workshop air cleaner to reduce particle counts further.

Types of dust extraction

Dust extraction equipment is broadly categorised into two types – “high pressure, low volume” (HPLV) and “low pressure, high volume” (LPHV). A HPLV extractor is akin to a household vacuum cleaner, sometimes referred to as a “shop vac”. A LPHV extractor is also sometimes called a “chip extractor” edit - this term generally applies if the filtration is coarse (see later). The principle of both types – suck air and separate out dust and chips – is the same, but how they perform differs. Broadly speaking, HPLV extractors are good with power tools, while LPHV extractors are good with stationary machines.

HPLV extractors use high-speed brush motors (typically 20,000RPM) to spin a small fan. Some sort of filter is necessary – often a paper or cloth bag like in a traditional vacuum cleaner. Indeed, if you have a traditional bagged vacuum cleaner, such as a Henry, then this makes a reasonable extractor for occasional use. Under no circumstances use a bag-less domestic vacuum cleaner, such as a Dyson. The cyclone separators in these aren’t designed for very fine dust, and you will quickly clog the small filters or wreck the motor with fine dust. Some HPLV extractors also have small filter cartridges which require periodic replacement or cleaning.

Typically, a HPLV extractor sucks between 40-70 litres a second of air. This is enough for power tools, but not enough for a lot of stationary machines. They make a lot of noise (because of the high-speed brush motor), and draw around 1,000-1,300W of power. The stall-pressure of a HPLV extractor is high (typically around 20kPa) – i.e. they have good “suck” – which makes them very well suited to power tools. Normally the hose diameter is 30-40mm.

A few manufacturers (Cam Vac, Record, Numatic) make multi-motor HPLV extractors. These use 2 or 3 motors to increase the air flow, at the cost of increased power draw (2-3kW). These work ok on power tools and smaller machines, so can be a reasonable compromise if you want one extractor to cover all bases. Generally though, using one of these with power tools will restrict the air flow too much through the extractor for prolonged use (they use the air flow to cool the motors), whilst the air flow still isn’t really enough to get really good extraction on table saws and planer thicknessers. They are a neat, low hassle solution for some, but be aware of these compromises. The hose diameter on these machines is typically 63mm or 100mm, although they can be used with smaller diameter hoses via the use of step-down adaptors.

LPHV extractors use induction motors directly connected to a large centrifugal fan (blower), typically about 350-450mm in diameter. Because of the (2 pole) induction motor and direct connection, the fan speed is 3,000rpm. Typically, this is horizontally mounted, with a collection bag beneath the fan and a cloth filter bag or pleated cartridge filter above the fan. Because the extracted air passes through the fan before the dust and chips are separated, the fan is usually of metal construction to afford impact protection. Machines designed for the smaller workshop use single phase induction motors from 500-2,200W. Most use 100mm diameter hoses.

Because of the large fan, LPHV extractors move a lot more air than the vacuum type. They are ideally suited to extraction from stationary machines, which require this large airflow. Because of the low fan speed however, the stall pressure is very low (typically around 1.5-3kPa). If you have ever been around one of these machines, you have probably felt that there is very little “suck” if you cover the hose inlet with your hand. This means that for small diameter hoses and openings (such as those present on power tools), the air flow is much more compromised than when a vacuum type extractor is used. Therefore, they are unsuited to use with hand-held tools.

Confused about which to buy? For most people, a vacuum type extractor is the right place to start as even those of us who work with stationary machines also probably use hand-held routers and other power tools. They are also useful for general cleaning duties. They are generally more flexible and achieving good filtration of fine dust is straightforward. Just don't expect them to do well with stationary machines, if you have them.


Once you have air and dust entrained in the hose, the dust and chips need to be separated from the air stream. The air is then returned to the workshop environment. Some extractors do have the capability to vent the exhaust air outside; however this just makes the outside environment more dusty (not very neighbour friendly) and in the winter time, you are wasting all the energy you may have put into keeping the workshop warm.

Edit - as a simple rule of thumb, because HPLV extractors have loads of "suck" to work with they are generally supplied with some method of fine filtration that is good for separation of fine dust. Because the flow is small and there is loads of "suck", the filters can be small and cheap. For LPHV extractors, because the flow is much higher and the system performance is much more affected by moderate flow restrictions, fine dust filters have to have a very large surface area (i.e. big and expensive, hence not generally supplied as standard). Most LPHV extractors are supplied with cloth filter bags, which doesn't do a good job of capturing the finest dust produced mainly by saws and particularly when working with man-made products like MDF. These are strictly what we refer to as "chip extractors"; they still do a good job on saws capturing a lot of the dust but not the health-damaging fine stuff. You can fit big filter cartridges to most of these extractors to also capture fine dust; however they need to be suitably sized to avoid restricting air flow (the pleated filters sold for this type of machine have several square metres of filter area). Specified large enough, performance should not be adversely affected. For any filter on either type of machine, look for claims about the particle size filtered (in microns) or an industry rating such as class L or class M.

The simplest way to achieve filtration in an HPLV extractor is by a filter bag in the extractor. The bag is permeable (fine fabric or paper); the air passes through the walls of the bag and the dust and chips stay inside. You can then throw the bag away, or in some cases empty it out for re-use. The down-side to just using bags is the quantity of waste you will generate with some tools means you will be changing and buying filter bags very frequently. For fine and hazardous dust, it’s an attractive option though.

Lots of HPLV extractors have the capability to run bagless. In these extractors, the dust hose is connected to a collection bucket that forms the body of the extractor. The motors in the extractor draw air from the collection bucket through cloth and/or paper filters, which remove dust from the air stream. These filters require periodic cleaning or replacing. Some vacuum type extractors have automatic or semi-automatic filter cleaning which is activated during operation.

On LPHV extractors, the air is drawn through the fan before being forced through a cloth filter bag or pleated filter cartridge. A collection bag sits beneath the filter which all the waste falls into.


When working with machines, the large volumes of waste generated means frequent bag changing. Particularly if you’ve ever used a planer or thicknesser, you’ll likely be very familiar with having to empty the collector several times an hour. For this reason, some people use a separator. The machine is connected to the separator inlet, and the dust collector to the separator outlet. All of these separators work using inertial separation – by changing the flow direction in the separator, the denser dust and chips don’t change direction as readily as air and so are essentially flung out of the airstream and then collected in a bin. Three types of separators are commonly used – “bin lid”, cyclones and Thien baffles. Some of these types also capture finer dust quite well, meaning you need to clean filters and empty the dust extractor’s bags less often. They can all be used with any type of dust extractor, although they should be sized appropriately for the airflow.

“Bin lid”
These are very simple devices that are made as the lid of the collection bin. There are two hose ports that penetrate the lid vertically. The incoming air and dust flows down into the bin, and the air then has to do a 180 degree turn to exit again. The heavier dust struggles to turn the corner and falls into the bin. These simple devices work very well for catching bulky chips; if your main objective is to reduce the frequency of emptying the collector when using planers and thicknessers then using one of this with a big collection bin is a great low-cost option. They don’t achieve good separation efficiency for finer dust however, because the inertial separation is very crude. You will still get a lot of dust on the extractor’s filters.

Cyclones work by creating a swirling path of air down the inside surface of an upside-down cone. As the air is constantly turning, denser material is flung out onto the wall of the cone. Because of the “boundary layer” flow effect, the air is very slow moving next to the cone wall, so the separated material is then free to fall downwards and out of the cone at its base (the tip of the cone is cut off which forms the cyclone exit into the drop box). The air enters the cyclone at a tangential angle, and exits via a lid placed on top of the cone.

There are a few manufacturers of cyclones, or you can build your own. Bill Pentz’s website ( is an excellent and very detailed online resource for information on cyclone design and construction. In order to optimise the pressure drop and small particle collection efficiency of the cyclone, the design and sizing needs to be carefully thought out. Essentially, if you make the cyclone too big then the air speed in the cyclone isn’t fast enough so you basically have a “settling chamber” that doesn’t work much better than the bin lid type for extracting fine dust. If you make the cyclone too small then the air speed becomes so fast that a high pressure drop is created, which is particularly a problem for the LPHV type extractors and you get a significant flow reduction as a result.

Thien baffle
Cyclones are quite tall so take up a lot of space. A more compact alternative somewhere between a cyclone and bin lid type separator is the Thien baffle. This is basically a thick lid (75-100mm or so) for your collection bin. The principle is similar to a cyclone – spin the air around the lid to separate dust out – but these are smaller and easier to build. The pressure drops generated are also lower. Some people get on really well with their Thien baffles, and they do a much better job than the bin lid type at capturing smaller dust. Like a cyclone, they should be sized according to air-flow. Their separation efficiency for very small dust isn’t as good as a cyclone however, so you will still get more build-up of fine dust on the extractor filters over time than you would with a suitably designed cyclone. More information can be found on Thien’s site here:
Last edited by siggy_7 on 17 Nov 2016, 12:34, edited 4 times in total.
By siggy_7
In response to the query about having an extractor with a high stall pressure and a high flow rate, MattRoberts is right to state that you would need a very powerful motor to do this. The stall pressure is related to the square of the tip speed of the impeller (i.e. increases with the square of rotational speed or diameter), and the power requirement also increases linearly with the stall pressure if you size the impeller to give the same flow rate. To achieve a stall pressure of 20kPa, you would need about a 1.2m diameter impeller when running at 3,000RPM (clearly not very practical!).

This extractor from Hammer has a maximum stall pressure of 2,150Pa: ... gion/gb-en Its impeller will be around 40cm in diameter to achieve this, and it needs a 2.2kW motor to deliver a rated airflow of 3,000m3/h (which I think is an over-estimate). To increase the stall pressure to 20kPa, you would need to triple the diameter of the impeller. To maintain the same flow rate, you can reduce the thickness of the impeller by a factor of 9, but the increase in power draw is still a factor of 9 - i.e. about 20kW. Clearly a machine of this scale and power draw is impractical and unnecessary - you don't need high stall pressure for big machines. It's more efficient (in terms of power, overall space of equipment and economics) to have different extractors for the different applications.
By siggy_7
Having covered the types of extractor, their uses and waste separation in the first post, this second part looks a bit more at understanding performance. I will try to keep the physics to a minimum, but please point out if things aren’t clear!

Pressure and flow rate

Air only moves in the presence of a pressure gradient. The impeller of the extractor creates a low pressure in the extraction hose, so air at atmospheric pressure is forced into the hose inlet. Extractors are generally characterised by two numbers – their stall pressure and maximum flow rate. It should be recognised however that these are the two extremes of the performance curve of the fan – the stall pressure is generated when there is zero flow, and the maximum flow is generated when there is very low pressure drop through the extractor. A typical performance curve of pressure vs flow rate might look something like this (from


From the perspective of ensuring good dust collection, the parameter you want to get high enough is the flow rate. Why? Well, the more flow rate the faster moving the air is around the dust extraction inlet of the power tool/machine. Faster moving air means more dust is swept into the hose. So, you want to maximise flow rate for efficient dust extraction.

For power tools, a good HPLV extractor with a maximum flow rate of about 60 litres/second will generally be sufficient. The extraction ports on power tools tend to be less than 40mm diameter and you will struggle to get much more flow through than this anyway. Dust extraction effectiveness for power tools is generally governed by how well designed the shrouds, guards and extraction ports are – the pricier manufacturers such as Festool spend a lot of development effort getting this right. If you have a decent HPLV extractor and your power tool is still leaving a lot of dust around, your options are basically limited to attempt to modify the shrouds around the cutting tool or buy a different tool. Don’t waste your time thinking that some magical super-extractor is going to solve your dust problems.

Machines generally have a recommended flow rate specified. For the majority of machines found in the average home workshop (bandsaws, table saws, planers and thicknessers up to about 12”, small spindle moulders), this will be around 1,000m3/h. Unfortunately, simply finding a LPHV extractor that matches the spec doesn’t necessarily guarantee you will have enough flow. This is because, as seen in the figure above, flow rate reduces with increasing pressure drop. To deliver high flow, you need to be a little careful in creating minimal pressure drops to maximise flow rate. The pressure drops in the extractor system arise from two sources: friction of the air against solid surfaces (i.e. hose walls), and inertial losses caused by accelerating the flow which arise from bends, joins, splitters, changes in cross-sectional area, separators such as cyclones if fitted and the entry losses at the woodworking machine. Filters also add to the pressure losses, particularly if they are under-sized. These pressure losses in the flow network are also a function of flow rate. If you were to plot these losses on the graph above, they would increase with flow rate. The point where this line crosses with the fan performance line is the flow rate you will actually get out of the system – i.e. where the driving force from the extractor balances the pressure losses in the network.

So – what to do? How to design your dust extraction system to get adequate flow? Well without detailing how to do a pressure network flow calculation for your system, there are a few pointers I can offer that may be of use:
• Pressure losses due to friction increase proportionally to flow passage length. Therefore, avoid very long runs of pipe and ducting. In a properly sized system, the friction losses for moderate hose runs aren’t a big contributor
• The inertial losses in the system are a function of flow speed squared, and so the dominant losses occur where the flow is fastest (minimum cross-section). Even over a very short length, this means that any constrictions really kill your flow. For this reason, use the same diameter hose as the machine outlet where possible. For the same reason, going bigger than the machine outlet won’t give you an appreciable performance boost, so don’t bother going to 125mm if your machine outlet is 100mm. If one of your machines has an outlet smaller than 100mm, then going to a 100mm hose is probably still a good idea to reduce the friction losses
• I would generally advise a minimum of 2hp for chip extractors. The lower flow rates of the 1hp machines require really careful design of the whole system to minimise losses enough to hit the target flow rates for most machines. Especially if you are using an inertial separator or fine filter cartridge, use a 2hp+ machine. An appropriately sized cyclone with good fine dust separation will create around 1.5kPa of pressure drop on its own, before anything else is considered
• Minimise bends, splits and joins. If you have to have some of these in your network, use transition pieces with smooth blends and bends with large radii – avoid T pieces at all costs
One final point on system performance – if the air moves too slowly in your hoses and ducts, then the dust can fall out of suspension in the air stream and lead to blockages. Aim for a flow speed of at least 20m/s to avoid this. This corresponds to 225m3/h for 63mm pipe, 565m3/h for 100mm pipe and 883m3/h for 125mm pipe. If using a HPLV extractor, then blockages shouldn’t develop because the stall pressure is high enough to clear them, so you can treat this requirement a little more loosely.

Fan design

Interested in building your own dust extraction fan with an old induction motor you have lying around? I have a spreadsheet to provide some guidance on how the performance and power requirements are affected by diameter and width. Unfortunately I can't work out how to attach an Excel file, so PM me if you want it or kindly explain what I'm doing wrong. My general guidance would be firstly to specify your fan diameter to achieve a target stall pressure (I would recommend around 10-12” water column, i.e. about 2.5-3kPa). To end up with a sensibly sized machine, you will want to use a 3,000RPM motor which will result in a fan diameter of 40-45cm. Then, vary the width of the fan until it matches your motor power and this will indicate your likely maximum flow rate (or specify the fan width to provide your target flow rate and see what power motor you need to find). For the specifics of how to build a fan, I like Matthias Wandel’s approach here: The backward curved blades give a more efficient blower than a straight radial design.
By siggy_7
Understanding HVLP fan performance

Seen a few questions recently that suggest a bit more on this topic might be helpful. I've already covered some basics of the relationship between pressure drop and flow rate above. In this post, I'm going to look at what governs the performance of a HVLP fan - the aim being to help people understand the differences between different HVLP machines beyond what power the motor is and what the manufacturers claim for flow rate.

How does a HVLP fan work?

All fans/blowers of this type work using centrifugal fans. In this design, the inlet is the centre of a spinning disk with some blades (or vanes) mounted radially (going from the centre to the outside diameter) - this is the blower/fan assembly. As the fan spins, the air is forced to rotate with the disk by the blades. This flings the air radially outwards. The fan is placed into a housing commonly designed as a scroll, with an outlet approximately tangential to the fan to match the flow direction of the air coming off it. A scroll housing means that just after the outlet there is a small clearance between the fan and the housing. Moving around the circumference, the clearance between the fan and the housing gets wider until you get to the outlet. This gives progressively more room for the air as it is flung off from the housing, but the small gap at the outlet encourages all the air to flow into the outlet rather than spinning around the housing. Wikipedia has more on centrifugal fans here:

Centrifugal fans are universally used in HVLP extractors as they are a very efficient way of moving air at moderate to high pressure drops. It's very difficult to design an axial fan with a high enough stall pressure - just having an axial fan in a duct wouldn't work. If you know what the compressor stages of a jet engine look like (which is a highly efficient axial fan able to generate plenty of pressure) and marvelled at the multi-stage design with stators and rotors built to very high tolerances, then you know how good the engineering needs to be to make axial fans work. It's much cheaper just to use centrifugal fans.

What influences the performance of the fan?

Principally, the performance is down to three things:

1. The diameter of the fan
2. The height/depth of the fan
3. The rotational speed of the fan

Other factors, such as shape of the fan blades and careful design of the housing, have an impact on the performance. Simple radial vane fans are surprisingly efficient around the required performance envelope for a dust extractor. There isn't that much to be gained by working really hard at the design, assuming that a simpler design is reasonably well executed. Obviously the performance is also reliant on having a motor powerful enough to turn the fan at its design speed - we're going to assume this is the case, since an induction motor held down to much below design speed by its load is seriously over-loaded and would not run for very long before overheating.

Rotational speed is governed by the motor. All HVLP extractors I've seen use induction motors, because they are cheap, efficient and good at running for extended periods. They are also much quieter than noisy high speed brush motors used in vacuum extractors. The cheapest induction motors are the simplest two pole designs, which also spin the fastest (3,000RPM on our 50Hz grid) - handily for our dust extraction systems a faster motor makes the system more compact. All HVLP extractors therefore use 3,000 RPM motors, so we can take fan rotational speed as a fixed quantity, leaving us with two parameters to consider.

What influences pressure?

The stall pressure is generated by the speed of the air in the fan - specifically, the tip speed. Why is this? Well, air pressure has two components - static and dynamic. Pressure is the force of air per unit area, and since air has mass then a moving body of air that runs into something will exert a force due to its momentum as it collides with the static object - the force being felt as a pressure. This is what is referred to as dynamic or velocity pressure, and it increases with the square of air speed. Static pressure is the pressure we would measure in a flow if we removed the velocity component. The combination of static and dynamic makes the total pressure. See the Bernoulli principle if you want to know more about this: Back to dust extractors though, the speed of the air in the fan (the relevant speed is that at the fan tip) is governed by two of our three factors - rotational speed and diameter. But we have already fixed the rotational speed from the motor. We can therefore conclude that the stall pressure of the fan is only related to fan diameter. Here's what that looks like for a fan with 2,950RPM rotational speed (allowing for a slight decrease on synchronous speed as the motor is loaded):

Diameter (m) Pressure (Pa) Pressure (mmH2O)
0.25 880 90
0.275 1065 109
0.30 1267 129
0.325 1487 152
0.35 1724 176
0.375 198 202
0.40 2252 230
0.425 2543 259
0.45 2850 291
0.475 3176 324
0.50 3519 359
0.525 3880 396
0.55 4258 435
0.575 4654 475
0.60 5068 517

You will notice that the pressure isn't going up linearly with diameter. In fact, it's going up with diameter squared. That's because the air speed increases linearly with diameter, but the pressure is increasing with air speed squared (see above). So if you want to know the stall pressure of a fan, just find out its diameter. You may find in practice the stall pressure is influenced a little by the efficiency of the fan and housing design, but basically it's all down to fan diameter.

What influences flow rate?

Flow rate is related to air speed and the flow cross-sectional area. If you think about the perimeter of the fan, the air speed is the tip speed of the fan and the area is the circumference multiplied by the height. So flow rate is influenced by all three of our factors. The more air we want to move, the more power we're going to need. It's more efficient to move air at the lowest speed possible, since less energy is lost in friction and turbulence. Therefore, the higher the stall pressure we want, the less air we will move using the same motor (or the more power we need for the same flow rate). The relationship between stall pressure and power per unit flow rate is linear, but remember the relationship between pressure and diameter is quadratic. The following table shows what you can expect from a 1HP extractor. If your motor is bigger or smaller, then just multiply the last column by your motor power and this is roughly what you should get. This is a little more approximate than the pressure table, since some efficiency parameters come into this which are influenced by design, but it's a good indicator of what to expect:

Diameter (m) Pressure (Pa) Flow per HP (m3/hr)
0.25 880 1530
0.275 1065 1264
0.30 1267 1062
0.325 1487 905
0.35 1724 781
0.375 198680
0.40 2252 598
0.425 2543 529
0.45 2850 472
0.475 3176 424
0.50 3519 382
0.525 3880 347
0.55 4258 316
0.575 4654 289
0.60 5068 266
Last edited by siggy_7 on 05 Dec 2016, 12:37, edited 1 time in total.