We recently had the pleasure of working with Tony Baker, who lives in our area.  Tony was interested in adapting a Farr-Style cartridge to his Harbor Freight 2 Hp dust collector, and we though it would be a great opportunity to document the procedure.  The following was accomplished in a few hours, using tools and hardware that Tony had on  hand.

Much of this work is based on designs and instruction from Bill Pentz, who has tirelessly worked to transform the hobbyist's knowledge base over the past few years.  Bill's website is full of valuable information on hobbyist dust collection, and can be found here: http://billpentz.com/woodworking/cyclone/DCConversion.cfm

Bill of Materials:
        (1) Wynn Cartridge #9E300BL
        (10) Plastic Bag #1940PB
        (1) 20" square (or larger) x 3/4" MDF (or Plywood)
        (8) Drywall Screws
        (1) Tube Clear Silicone (or Construction Adhesive, Gutter Sealant, etc...)
        (4) 1/4-20 x 2" Bolts
        (8) 1/4" Flat Washers
        (4) 1/4-20 Nuts
        (4) 1/4-20 Knobs, Wing nuts, etc...
Tools:
        Measuring Tape
        Router with Circle Guide
        Cordless Drill
            -Bits
            -Phillips Driver
        Socket Wrench

(click photos to enlarge)

Above is the "cyclone" section of the Harbor Freight unit.  Note the conical orifice plate.  The filter is Wynn part  # 9E300BL, which is 12 3/4" dia. x 34" tall, with a 14" x 16" flange.  This filter has 300 sq. ft. of 80/20 blend media, and it's a good match for this type of dust collector.  Efficiency is rated 99.99% at 0.5 microns.

 Rather than trying to seal against the funny angle inside the cone, we decided to fabricate a simple donut out of MDF.  Here's Tony mounting the donut using drywall screws.  The nice, round shape of the donut was accomplished using a router.   (We attached a pegboard to the router, and nailed it to the MDF).  As you can see, we drilled matching holes in the filter flange and donut, and installed 1/4-20 studs to mount the filter.  The donut sets down just below the lip, and is sealed inside using silicone.

Here is the system with the filter in place.  Tony had four nice 1/4-20 knobs, which gives the finished unit a clean, professional look.  Although not visible, there is plenty of "crush" on the gasket, assuring a good seal against dust leakage.

Here is the finished unit with the cartridge filter and the lower bag installed.  The whole conversion took about 6 man-hours.  With proper planning, this can be duplicated in half the time.

Results:
        Airflow                     -New Filter- 560 CFM in the 4" inlet pipe (6300 fpm)
                                         -Seasoned Filter- 330 CFM in the 4" inlet pipe (3800 fpm)
        Outlet Emissions    -New Filter- Average 2.1 mg/m³ (3.0 mg/m³ absolute peak)
                                         -Seasoned Filter- Average 0.016 mg/m³ (0.022 mg/m³ absolute peak)
        Efficiency                -New Filter- Approximately 97.4%   
                                         -Seasoned Filter- Approximately 99.7%      
        

Notes:
These measurements were taken after total run time of about 1 hour (less than 1 cubic foot of dust in the lower bag), then again after a few weeks (about 3 cubic feet of dust in the lower bag).  Filter efficiency is always rated with seasoned filters, and the numbers dramatically improve after a few hours of use.  Particle counts were taken during cuts of MDF board in a table saw.  It is interesting to note that on the initial tests, we were expecting particle counts to be much lower, and were surprised to see such high levels at both the inlet and exhaust.  In industry, woodworking operations produce large amounts of course dust, and very little, or no sub micron particles.  It became obvious why, when we realized afterward that all of our cuts were resulting in scorched edges.  Apparently, we had misaligned the fence and the MDF board was burning.  This raises an interesting issue, because common practice does not consider smoke when calculating the efficiency of a wood dust collector.  We've since checked with some experts and, although it's not common in industry, scorched edges are quite common among the hobbyist community.  Food for thought.  Our second set of tests, with the seasoned filters, were performed without burning any wood, which is apparent by looking at the difference in inlet concentrations.

Flow was taken with a Dwyer Pitot Tube and Magnehelic Gauge, and indicates an accuracy of plus/minus 10%.

The Outlet Emissions were taken inside of a temporary exhaust plenum using a DUSTTRAK Aerosol Monitor, and indicate particle emissions in the size range of 0.10 to 10.0 microns (Respirable Fraction).  At an initial peak level of 2.1 mg/m³, these outlet emissions fell safely below the OSHA limit of 5.0 mg/m³, but were above the NIOSH and ACGIH limits of 1.0 mg/m³.  After seasoning, the numbers fell to a peak level of 0.022 mg/m³ (about 200 times lower than the OSHA limit, 50 times lower than NIOSH).  The charts below are the actual runs for inlet and outlet concentrations.  The spike is the wood being cut.

Initial tests with new filter:

Follow up tests with seasoned filter:

Efficiency was calculated using (outlet concentration) divided by (inlet concentration), and does not take into account any particulate above 10 microns.

Conclusion:
    The first thing you will notice upon startup is a huge increase in airflow, as compared to the original bag arrangement.  This is due to the large surface area of the cartridge filter.  The air laws tell us that doubling the surface area of a filter results in a four-fold decrease in pressure.  The stock bags offer about 35 sq. ft. of area, so the 300 sq. ft. cartridge makes a big difference.

    In addition to the increased flow, the filtration efficiency of the system is also improved.  This is partially due to the increase in surface area, but is mainly due to the improved performance of the 80/20 blend paper.  We were able to measure outlet emissions (New Filter) no higher than 3.0 mg/m³, with an overall efficiency of 97.4% at 0.1-10.0 microns.  This is about right for a new filter on a smoke or fume application and, although we did not expect the smoke content in the gas stream, it's important information to have.  The efficiency rose substantially, to about 99.7% after a few hours of heavy use.

We want to know more about burning during a cut, and find out how common this actually is.  Based on the data we were able to collect here, it appears that when burning during a cut, there are about 10 times more particles emitted in the 0.1-10 micron range.  This is the respirable range, which causes the most damage.  Everyone ought to be especially conscious of this, and should take precautions to avoid burning while they cut wood.

Future:
    Further testing will be performed on various filter styles, including the washable spun bond filters. 

Special thanks to Tony Baker, for being kind enough to let us perform these tests, and to Bill Pentz and Brad Olsen for helping us interpret the data.

Home Page
Filter Page
Flex Hose Page