Page 1 of 2 12 LastLast
Results 1 to 15 of 24

Thread: Grizzly G0443 Cyclone Dust Collector Testing

  1. #1
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968

    Grizzly G0443 Cyclone Dust Collector Testing

    I thought I'd make another thread to discuss testing my G0443 with a hot wire anemometer.

    First, I'd like to thank John Lanciani for the gracious loan of his equipment. I've treated it better than I treat my own tools, and I'll have it back in the mail ASAP.

    Second, some discussion about the setup. I've got a G0443 that I vent to the outside. Testing with the filter installed did not show much of an improvement with venting outside, so I'm not going to hook up the filter again for these tests. All the ducting is 6' 26 gauge sheet metal ducts, as well as the connections to the machines themselves to ensure maximum performance. All duct seams have been coated in mastic to seal them. In most cases I've stretched out any bends to be ~5' long, rather than two short 90s. The longest run is to the SCMS and jointer at ~45', where it runs across the shop, down a wall along another wall, then up.

    For the purposes of this test I did not test the 4' drops of which there are a few, all very short runs. Previous testing has shown those get a bit squirrely, and don't necessarily reflect the ability of the DC.

    I have not taken any of the piping apart, and I don't intend to do so. Outside of that, I'd be happy to do any re-tests for those who are skeptical of my results.

    Diagram of setup.
    DCSetup.png
    Dust collector in the corner.
    IMG_20190715_201559.jpg
    Ducting to table saw/router table
    IMG_20190715_201604.jpg
    Connection of Table saw to ducting, and split to router table
    IMG_20190715_201611.jpg
    Connection of router table
    IMG_20190715_201618.jpg
    Major intersection, heading towards the other tools.
    IMG_20190715_201624.jpg
    Connection to band saw, and rigid sander
    IMG_20190715_201632.jpg
    Rigid Sander Intake
    IMG_20190715_201644.jpg

  2. #2
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Connection to the SCMS.
    IMG_20190715_202434.jpg

    Testing.

    To test the system, I put a whole into the ducting, right before the ducting starts to branch. With all ducts open I was able to read 5,600 FPM from the Hot Wire anemometer. This corresponds to a speed of 1096 CFM, a little higher than the rating given by Grizzly of 1025 CFM. Previous testing has shown that this is at a SP of 2".

    The numbers at the tools themselves were a little lower, likely due to drops from the ducting.

    Location Testo Hot Wire
    SCMS 3,800 fpm 4,400 fpm
    Sander 3,900 fpm 3,800 fpm
    Table Saw 4,700 fpm 3,900 fpm
    Last edited by Andrew More; 07-15-2019 at 9:33 PM.

  3. #3
    Join Date
    Feb 2013
    Location
    Tippecanoe County, IN
    Posts
    836
    How many points in your traverse?

    Did you only traverse in one direction?

    Why did you put your test port close to a junction instead of midway between the junction and the DC?
    Last edited by David L Morse; 07-16-2019 at 6:50 AM.
    Beranek's Law:

    It has been remarked that if one selects his own components, builds his own enclosure, and is convinced he has made a wise choice of design, then his own loudspeaker sounds better to him than does anyone else's loudspeaker. In this case, the frequency response of the loudspeaker seems to play only a minor part in forming a person's opinion.
    L.L. Beranek, Acoustics (McGraw-Hill, New York, 1954), p.208.

  4. #4
    Join Date
    Nov 2007
    Location
    NW Indiana
    Posts
    3,078
    Good questions about doing a traverse. When I tested using that instrument, I did a traverse multiple times. Also, to get accurate measurements, you must do it in a straight section of pipe. Otherwise, you will have turbulence and inaccurate results.

    My test setup was done to be able to accurately do a traverse with the probe properly aligned and repeatable positions on the traverse.
    Last edited by Larry Frank; 07-16-2019 at 8:11 AM.

  5. #5
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Quote Originally Posted by David L Morse View Post
    How many points in your traverse?
    Sorry, what does this mean?

    Quote Originally Posted by David L Morse View Post
    Did you only traverse in one direction?
    Sorry, don't understand this question either.

    Why did you put your test port close to a junction instead of midway between the junction and the DC?
    Pre-existing hole. If you've got a suggestion for another place I'm open. Is this what you're suTest Hole Setup.jpgggesting?
    Last edited by Andrew More; 07-16-2019 at 8:22 AM.

  6. #6
    Join Date
    Feb 2013
    Location
    Tippecanoe County, IN
    Posts
    836
    Quote Originally Posted by Andrew More View Post
    Sorry, what does this mean?

    Sorry, don't understand this question either.
    These might help:
    https://www.youtube.com/watch?v=NnxTbVrfb-Q
    https://www.tsi.com/getmedia/1a11d34...0Duct?ext=.pdf
    https://www.flowkinetics.com/pitot-s...uct-survey.htm


    Quote Originally Posted by Andrew More View Post
    Is this what you're suTest Hole Setup.jpgggesting?
    Yes. Also, put your static port somewhat close to that, on the DC side of it.

    Repeat the measurements for various combinations of open gates and you will have pressure versus flow data that should line up with Grizzly's.
    Beranek's Law:

    It has been remarked that if one selects his own components, builds his own enclosure, and is convinced he has made a wise choice of design, then his own loudspeaker sounds better to him than does anyone else's loudspeaker. In this case, the frequency response of the loudspeaker seems to play only a minor part in forming a person's opinion.
    L.L. Beranek, Acoustics (McGraw-Hill, New York, 1954), p.208.

  7. #7
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Thanks David.

    I've read through the docs, but I don't see much value in cutting multiple holes as suggested for round duct, so I did a transverse for just the single hole. The new hole is a bit of a compromise, since the duct run is only 4.5' long, it's not possible to find a place that's both 7.5x duct size from the wye, and also 3x duct size from the intake of the DC. As such I decided to do the measurements at 1' from the DC at the top, and in 1" increments. As such the numbers are still coming out pretty close to the same.

    Distance Velocity (fpm)
    1" 5565
    2" 5410
    3" 5220
    4" 5000
    5" 5000

    Average velocity is 5,239, which is ~1050 CFM. Still better than spec of 1025 CFM, but only by a little, and likely because of a drop in static pressure from the removed filter.

    Further the traverse numbers don't appear to be much different from just taking the velocity at the center of the duct, 5220 in the middle vs 5239 traverse average.

  8. #8
    Join Date
    Nov 2007
    Location
    NW Indiana
    Posts
    3,078
    I am glad you are happy with your testing.

    Traverse testing locations are based on equal areas and not equal distance.

    The only good way to do the testing is with a long test duct to get laminar flow. Your testing locations are no doubt quite turbulent.

    It is a shame that doing good testing is so difficult as it would be useful to a lot of people. I monitor the static pressure to look for changes in my system. I have the performance curve I got from my testing and using the static pressure can estimate the flow.

  9. #9
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Quote Originally Posted by Larry Frank View Post
    Traverse testing locations are based on equal areas and not equal distance.
    Okay, since the numbers are pretty close to each other, what does it matter? Serious question, since I don't see how the difference effects the outcome in a meaningful manner. You're essentially quibbling over a range of 1113 cfm to 1000 cfm, or about 10%, and that's at absolute best, assuming that the correct answer is either the high or the low range, rather than closer to the average of the two.

    The scale printed on the side of the anemometer is not done to a level that would allow the level of detail you're looking for, even if such testing was useful.

    Quote Originally Posted by Larry Frank View Post
    The only good way to do the testing is with a long test duct to get laminar flow. Your testing locations are no doubt quite turbulent.
    Really? Since the numbers are pretty close to the manufacturer's specification. We're talking a difference of 25 cfm, or 3% of the total flow rate. And wasn't the original assertion that the numbers provided by Grizzly in their fan curve were not possible? "I do not believe a 1.5 hp Dust collector will deliver 1000 cfm" is what you said. The Grizzly 1.5 HP Dust Collector delivers ~1000 CFM of dust collection, tested with a hot wire anemometer, in a tube that is about 6" short of the requirements for testing.

    Would it be more believable if I posted a movie of the anemometer in operation?

    Further if the flow is turbulent, that's basically what the system is going to be. It's never going to be run in a configuration with a perfect set of pipes that provide laminar flow through the entire ducting system. Rather it's going to have to deal with the compromises I've had to make for it to do it's job, collect chips and fine dust. Far better to know what that means to the air flow, and the system's ability to deliver suction. At all the tools with 6" ducts the system is delivering ~3800 FPM, which is close to Bill Pentz's magic numbers 4000 FPM for dust collection.

    All of this ignores the elephant in the room: the inability to create a decent shroud, particularly for tools like the SCMS, or Table Saw, to deliver that magic 4000 FPM. And those tools feature blades moving at 100 MPH, while even the best dust collection system delivers at best 40 mph.

    And if this testing is useful to people, then they can check the Grizzly Fan Curve.

  10. #10
    Join Date
    Feb 2013
    Location
    Tippecanoe County, IN
    Posts
    836
    Andrew, Larry's comment on testing locations is important. The area of a circle varies with the square of diameter so the test points must me located accordingly. You've sampled the inner 4 inches of the cylindrical air column. That 4" contains 44% of the area of the 6" pipe. You've left the outer 56% untested. Airflow in that outer region changes rapidly with radius. It actually drops to zero at the wall of the pipe. Sampling points in that area are important.

    Did you measure pressure?

    It would be helpful if you would also check the flow and pressure with most blast gates closed.

    I also recommend checking the motor current at your maximum flow condition to be sure it's not overloaded. Also, if you record current at several different flow levels you can use that data to give you a good estimate of flow at any time without needing a pressure gauge or anemometer. If you're interested, I can provide more information.
    Beranek's Law:

    It has been remarked that if one selects his own components, builds his own enclosure, and is convinced he has made a wise choice of design, then his own loudspeaker sounds better to him than does anyone else's loudspeaker. In this case, the frequency response of the loudspeaker seems to play only a minor part in forming a person's opinion.
    L.L. Beranek, Acoustics (McGraw-Hill, New York, 1954), p.208.

  11. #11
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Quote Originally Posted by David L Morse View Post
    Andrew, Larry's comment on testing locations is important. The area of a circle varies with the square of diameter so the test points must me located accordingly. You've sampled the inner 4 inches of the cylindrical air column.
    Doing what you're asking isn't possible with this anemometer. The closest I can get is 3/4" of an inch, due to the way the probe is designed, which is to say the hot wire is 3/4" of an inch from the top of the probe.

    IMG_20190717_142612.jpg

    Also I have not sampled the inner 4", but rather inner 5", starting at 1" away from the wall of the duct, to 1" away from the other wall of the duct.

    So I took two additional measurements.

    3/4" - 5600 fpm
    5 1/4" - 4800 fpm

    These average out to ~5200, which is once again right in line with the numbers already determined. I do not see this adding any additional useful information to the measurement.

    Did you measure pressure?
    No, because I don't currently have a pressure gauge. I have borrowed a pitot tube based one in the past, and the results are consistent with the fan curve provided by Grizzly. This is both with all gates open (2.5" SP) and closed (10.5" SP) Unlike the anemometer, my HVAC buddy needs it on a daily basis.

    I also recommend checking the motor current at your maximum flow condition to be sure it's not overloaded.
    Overloaded based on what please? Do you mean overloaded the circuit, or the motor plate?

    I took readings for amp draw with all the gates closed (10.2 amps) and opened (16.2 amps) at 110V AC. Grizzly specs the motor at 18.2 amps, which I saw as the motor was getting up to speed, with all gates opened, but once it was warmed up it stay at ~16.2 amps.
    Last edited by Andrew More; 07-17-2019 at 3:05 PM.

  12. #12
    Join Date
    Feb 2013
    Location
    Tippecanoe County, IN
    Posts
    836
    Ok, let's try this again. Maybe a picture will help.

    AndrewProfile.JPG

    This is a plot of your data points (plus the boundary conditions) versus area of the duct. You can see the large gaps between 0 and 20% and also 80 to 100%. 40% of the area is unknown territory and this is where nearly all of the velocity gradient is.

    You can try to sketch a smooth curve that fits the data. Be sure it goes through zero velocity at 0 and 100%. Your average velocity would then be the area under that curve.
    Beranek's Law:

    It has been remarked that if one selects his own components, builds his own enclosure, and is convinced he has made a wise choice of design, then his own loudspeaker sounds better to him than does anyone else's loudspeaker. In this case, the frequency response of the loudspeaker seems to play only a minor part in forming a person's opinion.
    L.L. Beranek, Acoustics (McGraw-Hill, New York, 1954), p.208.

  13. #13
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Okay, I understand what you're saying.

    However, you don't explain how a reading from 1/8" of an inch (6" x 0.032 from this) can be taken with a tool that has 1/2" of rod above the measuring device, which itself measures 3/8" including the hot wire, temperature guage, and their orifices. I do not see how the level of precision you're looking to get can come from this tool.

    Also the missing area covered is no longer 40%, it's 25%. The area of 5.25" is 21.6, vs total area of 28.2, or 76%. Further that assumes that we don't actually measure anything from 3/4" to the wall of the pipe, while in fact we're going to get some percentage of that area as well.

    Assuming that it did measure that missing 1/8", is it going to show a difference of 25-50% off the specs that Grizzly generated? To do so it would need to immediately drop to 0 fpm at 3/4", which is physically impossible, and even then it would only be off 25%.

    I understand that marketing is a thing, but the numbers being show are within 3% of the numbers given by Grizzly. After independently validating that they're close to what I'm getting, I'm going to assume that the engineers they have working there are well versed in these techniques, have done the appropriate testing, and have published those results in the fan curve. Further I'm going to assume that Grizzly's engineers are probably better at this than I am, and the numbers they're giving have been tested under better conditions.

    Finally, the results we're looking for have largely be rounded up to the nearest 25% or so anyway. Or do you really think that you need to exactly hit 4000 FPM to remove the fine dust, or that there's a bit of a curve to that as well?
    Last edited by Andrew More; 07-18-2019 at 11:09 AM.

  14. #14
    Join Date
    Feb 2013
    Location
    Tippecanoe County, IN
    Posts
    836
    Quote Originally Posted by Andrew More View Post
    Okay, I understand what you're saying.

    However, you don't explain how a reading from 1/8" of an inch (6" x 0.032 from this) can be taken with a tool that has 1/2" of rod above the measuring device, which itself measures 3/8" including the hot wire, temperature guage, and their orifices. I do not see how the level of precision you're looking to get can come from this tool.
    Exactly! You cannot get the level of precision you are claiming.

    Quote Originally Posted by Andrew More View Post
    Also the missing area covered is no longer 40%, it's 25%. The area of 5.25" is 21.6, vs total area of 28.2, or 76%.
    No, the circle between .75" and 5.25" has a diameter of 4.5". The area of a 4.5" circle is 15.9sqin or 56%.

    Quote Originally Posted by Andrew More View Post
    Further that assumes that we don't actually measure anything from 3/4" to the wall of the pipe, while in fact we're going to get some percentage of that area as well.
    Correct.

    Quote Originally Posted by Andrew More View Post
    Assuming that it did measure that missing 1/8", is it going to show a difference of 25-50% off the specs that Grizzly generated? To do so it would need to immediately drop to 0 fpm at 3/4", which is physically impossible, and even then it would only be off 25%.

    I understand that marketing is a thing, but the numbers being show are within 3% of the numbers given by Grizzly
    There's a big difference between 25% and the 3% you're claiming.


    Quote Originally Posted by Andrew More View Post
    I understand that marketing is a thing, but the numbers being show are within 3% of the numbers given by Grizzly. After independently validating that they're close to what I'm getting, I'm going to assume that the engineers they have working there are well versed in these techniques, have done the appropriate testing, and have published those results in the fan curve. Further I'm going to assume that Grizzly's engineers are probably better at this than I am, and the numbers they're giving have been tested under better conditions.
    I have no issue with Grizzly's claim to those numbers. The point is that you can't use the velocity in the center 60% (or whatever fudged number you think is appropriate) to represent the whole duct. The are adjustment factors that are used when only the center velocity is know. Fully developed flow occurs a long way from the entrance and theory predicts average velocity is about 83% of the center value. For more realistic situations Dwyer recommends a 90% factor.

    Quote Originally Posted by Andrew More View Post
    Or do you really think that you need to exactly hit 4000 FPM to remove the fine dust, or that there's a bit of a curve to that as well?
    No, you've been misinformed by "Internet Wisdom". Fine dust is easily moved by the slightest breeze, it's completely at the mercy of the air. It's the largest particles that require the most velocity to keep them moving. It does indeed take 4000fpm to reliably lift the big stuff up a vertical drop. Typically 2000 to 2500 will keep it moving horizontally. That's one reason drops are often a smaller diameter than the main.

    Finally, it probably seems like you're being attacked but that's really not what's intended. I really appreciate that you have taken the time to make the measurements and share all of your data. It's all valuable. It's a search for knowledge and all contributions are important. Thank you for having the patience to keep plugging away at this.
    Beranek's Law:

    It has been remarked that if one selects his own components, builds his own enclosure, and is convinced he has made a wise choice of design, then his own loudspeaker sounds better to him than does anyone else's loudspeaker. In this case, the frequency response of the loudspeaker seems to play only a minor part in forming a person's opinion.
    L.L. Beranek, Acoustics (McGraw-Hill, New York, 1954), p.208.

  15. #15
    Join Date
    Feb 2019
    Location
    Cincinnati, Ohio
    Posts
    968
    Quote Originally Posted by David L Morse View Post
    I have no issue with Grizzly's claim to those numbers.
    Okay, Larry Frank, who you were agreeing with, does have a problem with Grizzly's claims, which is what started this. My bad for assuming you were making the same claims.

    If you don't have a problem with Grizzly's performance claims, what is the purpose to doing additional testing of the results? There's little I can do at this point to improve the performance of this system without replacing the DC. All the curves are as long as they're going to get, the ducts have been sealed, I have run as large a pipe as possible to all tools. As a hobbiest I find the Grizzly over performs for what I need. In fact it's the most expensive machine in the shop atm.

    Quote Originally Posted by David L Morse View Post
    For more realistic situations Dwyer recommends a 90% factor.
    I'm okay with that result as well. I do not see any gain to arguing over a 10% difference. If there is one, please explain.

    Quote Originally Posted by David L Morse View Post
    No, you've been misinformed by "Internet Wisdom". Fine dust is easily moved by the slightest breeze, it's completely at the mercy of the air.
    Okay, I'm going by what I've read on Bill Pentz's site. He claims this number comes from air quality engineers, and is required to collect all the fine dust from a sphere around 15 inches near the tool. I do note he provides no references to back it up, but his rational seems reasonable. What do you feel is the correct number to be looking for, and what do you base this upon?

    Quote Originally Posted by David L Morse View Post
    It's the largest particles that require the most velocity to keep them moving. It does indeed take 4000fpm to reliably lift the big stuff up a vertical drop.
    How big? I've sucked up small blocks in some of the ducts without issue, and I don't believe I'm hitting quite that number. I routinely throw cut offs into the inlet behind the SCMS, and can hear them bouncing all the way to the collector. I'm also pretty sure there aren't any accumulations of dust in the ducts, through I'd need to run a camera in to be sure.

    Quote Originally Posted by David L Morse View Post
    Finally, it probably seems like you're being attacked but that's really not what's intended. I really appreciate that you have taken the time to make the measurements and share all of your data. It's all valuable. It's a search for knowledge and all contributions are important. Thank you for having the patience to keep plugging away at this.
    If I come across as combative it's likely because I'm frustrated. I do not understand what you're interested in, so it doesn't appear to be possible, or useful to me. Maybe if you could explain more fully exactly what you think is useful to know here I could do a better job of doing the leg work.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •