Page 2 of 14 FirstFirst 123456712 ... LastLast
Results 16 to 30 of 201
  1. #16
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    Summary of the Sonoff Dust "switching points" for sanding.
    The PM2.5 data is what the sensor is really measuring and is there for completeness,

    The sensor itself only detects PM2.5 between ~0 and 0.6 ppm - the total dust has been measure by anther device.

    Screen Shot 2017-05-31 at 6.01.34 PM.png


    It looks like the "Good to Moderate" air quality indication would be a fair indicator for Total excessive hardwood dust levels (~1ppm).

    The "Moderate to Unhealthy" switch point of 3ppm would be a fair indicator for excessive softwood dust levels (~5ppm).

    Put another way, if the Sonoff readout stays at "Good" you are OK for hardwood dust, if it stays at "good or moderate" you are OK for Hardwood dust.
    A readout of "Unhealthy" is best avoided for any wood.
    And it looks like its not really possible to use this detector to monitor MDF dust.

    Bear in mind this applies ONLY to sanding and that device is a very basic gizmo and not one to stake your life on, but OTOH is probably better than nothing.

  2. # ADS
    Google Adsense Advertisement
    Join Date
    Always
    Location
    Advertising world
    Posts
    Many





     
  3. #17
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    The new particle PMS 7003 particle detector arrived last week but have not had time to play with it much until today.

    The photo below shows how small the sensor is.
    It's a touch smaller than a match box.
    The cost including the adapter board, cable and deliver was $34.
    The cable plug has a 0.05" connector spacing which is tricky to find plugs for so I just cut it and wired standard breadboard jumpers to it.
    Only 4 of the 8 wires are used.
    DUstSensor.jpg

    Here you can see the teeny tiny fan entrance and the air outlet.
    Air and dust is draw thought the fan opening, past a laser and photo detector and pushed out the hole on the RHS.
    IMG_2314.jpg

    For the technical minded
    - the output of the sensor is in a serial format that has to be decoded to extract the information.
    - to drive and encode the data I used an Arduino Uno clone. A real Uno costs $40 but clones can be had for ~$5.

    Here you can see the hookup to the clone.
    Only 4 wires are use, Power (red), Grnd (orange), and two serial lines blue/green
    Hookedup.jpg

    I adapted the code supplied with the data sheet and just reformatted some of the output so it was easier to read.

    It churns out quite a few numbers because this little gizmo does a lot of things.

    Each line is a set of readings that it spurts out about each second

    The 2nd to 7th columns are concentrations of dust in ug/m^3 - there are two identical sets - still don't know why it does that
    3 concentrations are provided
    - PM1_0c = for particles of 1 microns and smaller
    - PM2_5c = for particles of 2.5 microns and smaller
    - PM10c = for particles of 10 microns and smaller

    Columns 8 - 13 are particle densities (number of particles per decilitre)
    Note: it covers particles of of 0.3, 0.5, 1. 2.5, 5 and 10 microns which is fantastic and if accurate would allow much useful into about WW dust to be extracted from it..
    AT around about count number 12 I started ruffling my hair above the fan and you can see the counts and concentrations went up slightly


    PMS7003data.jpg

    The next BIG question is, how accurate is this thing?
    I will calibrate it agains the particle counter from work.

    There is still a fair bit to do to convert this into a reasonably rugged, stand alone, gizmo with its own display power etc plus I want to incorporate a few other sensors into the device if I can.
    I have a nice little Photon board with a WIFI chip on it it should run and which would be nice to use to log data.
    Attached Images Attached Images

  4. #18
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    And now with basic LCD output.
    Wiring starts to get messy pretty quickly
    LCDout.jpg

  5. #19
    Join Date
    Oct 2013
    Location
    Perth, Australia
    Posts
    1,813

    Default

    Nice work! When I've got controls and displays to use in the prototyping process I usually mount them to some stiff cardboard which helps with the cable mess a little bit. I also got my hands on a Dupont style crimping gun off ebay which made making my own connectors way easier.

  6. #20
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    Cheers Dan. Yeah I have a 5 magnetic interchangeable jaw crimping gun. I've used it mainly on mains gear but can see I will be using it on this stuff. Be good to see any of the stuff you've made. Maybe post in the electronics forum?

  7. #21
    Join Date
    Oct 2013
    Location
    Perth, Australia
    Posts
    1,813

    Default

    I'll have to have a dig around, most of the stuff I finished were guitar pedals and they've all long gone to new owners.

  8. #22
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    More mucking about with the particle counter setup today.

    Added Temperature/Humidity sensor
    A buzzer for when total concentration goes over a set limit.
    An LCD output display switch to switch between Particle numbers and concentrations data.
    PMS7003fullypimped.jpg

    Just about ready to start a package to make it mobile for testing.

  9. #23
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    OK I have completed a working dust particle counter in a box that's portable / mobile.
    It will do for testing in a real situation.

    Like I said in the electronics forum, for the CO gas sensor, its one thing to get these sort of projects working as a prototype on a bench and another to package it into something, potable, rugged, usable etc.

    Depending on the state of the switch on top of the box, the small LCD display shows either; 1) concentration in ppm or; 2) Number of particles for given dust size ranges
    Concentrations are available for 3 particle size ranges: "1:" = 1 micron and smaller, 2.5:= 2.5 micron and smaller and "10:"= 10 micron and smaller.
    Numbers of particles per decilitre are available for 5 particle size ranges: 0.3-0.5 microns, 0.5 to 1.0 microns, 1 to 2.5 microns, 2.5 to 5 microns and 5 to 10 microns. (see second photo)

    The main number for wood workers would be the concentration for the 10 microns and below particles.

    IMG_2330.jpg

    The parts for this cost $72 with the bulk of the cost being the particle detector ($34) and the plastic box $21! , because it is an IP rated box.

    The box could definitely be smaller and less fancy. The reason I went for such a large box is that when they arrive I want to fit a couple of other sensors into the box.
    I have just bought several similar style boxes on eBay for about $8 and I'm sure they could be had cheaper than that.

    For those that are interested, this unit uses an Arduino Mega clone board, and Mega Sensor Shield - The Mega board cost $11 but it would be possible to use an Arduino Uno clone or something even smaller which cost around $4.50. The Shield plugs over the top of the Mega Board and provides loads more Grounds and Power pins which makes it easier to connect up all the components, but is not essential and can be done without.



    IMG_2331.jpg

    Here you can see how the particle counter is attached to the box and holes cut into the box where the air gets in and out of the counter,

    IMG_2332.jpg

    The program as it stands is pretty basic and needs some more work. I have had it running overnight and this showed up some program bugs wth timers etc.

    The counter can operate without a computer in a stand alone manner using a battery or plug pack power, but currently if connected to a computer the program will log time and all sensors data which can be saved and imported into something like a spreadsheet for further analysis.

    Once the program is sorted this prototype will be moved to shed where I will see what it can do.

    The next model/prototype will be smaller and will auto log the data onto an SD card.
    This sort of unit will be used to log data in for example mens sheds over about a week or so.

  10. #24
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    Running tests on the dust detector software to make sure it runs by itself and logs everything properly.
    I started soldering some headers (small sockets) on a gizmo called DigiSpark - a small USB stick size data logger - when I noticed the dust levels had increased significantly.
    The detector was only 1/2 m away from the soldering station.

    All I did was solder 6 small spots and the dust levels in the room tripled.

    Here's the graph.

    Screen Shot 2017-06-27 at 8.59.36 PM.png

    Before soldering the dust level in the room was 7-8 ppb (thats right parts per billion!) and its been pretty much that since 3pm when I started this test..
    While this seems low it is not that far from average winter values in Australian homes

    As soon as I Soldered the levels went to 25 ppb and took around 15 minutes to get down to half the original increase - this is similar to the time it takes for wood dust to settle when sweeping floors. 25 ppb is also not high and typical of what the levels would rise to during an evenings activity in an Australian house.

    What this shows is how incredibly sensitive these detectors are.
    My fear is that they may be too sensitive. The threshold for WRC and MDF is 500 ppb so it should do those OK but I'm not sure after that.

  11. #25
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default


    Not quite there yet but getting close to a commercial product.

    Same sensor I'm using (PMS7003)

    Single number display of PM2.5 concentration. Wood workers really need the full 10 micron range concentration.

    2017 Mini LED LCD Plantower PMS7003 G7 laser PM2.5 detector meter with battery | eBay

    I don't believe the +/-1 ppb accuracy but if you have a spare $90 it might be useful thing to have?

    If you buy one remember you cannot shove this up close to a stream of sawdust or down inside a DC duct to see how much dust there in there or you will kill it.

    It's meant to be placed in a quiet corner or on a shelf , and you monitor things from there.

  12. #26
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    Did a bit more on the program last night.

    The code is getting on for 450 lines or so - it's a point where its starting to be tricky to hold it all in my head and I'm increasing needing to use the "Find" tool.
    It means when I edit . . . . err . break it! - it gets harder to fix so multiple levels/versions of back up are needed.
    It's also grown a bit like old spaghetti code and it needs to be better organised, about 1/4 of the code should be converted into functions.

    When hooked up to a computer running the Arduino program, all data can be recorded.

    In "stand alone" mode the unit does not (yet) log data, the program runs continuously as soon as power is hooked up so have added a couple of external switches to enable better data recording.
    When running continuously the data on the screen changes every few seconds so by the time one has written down the first of a series of displayed data on the screen the others have changed.
    There are also two sets of data (concentration & numbers of particles) displayed by the screen - each are viewable by the state of a switch.

    To get around this a simple on/off switch pauses/restarts the program so that data can be written down. Pausing the program allows one to write down all the numbers on the Concentration data screen and then toggle the screen to show/record numbers of particles, and then resume the program.

    The instantaneous data is highly variable so some form of integration/smoothing over time is needed. The professional level unit I borrow from work has a fixed 30s integration period.
    On my unit I have added a 4 position switch which sets the integration times to either 10, 30, 50 or 100s. These can easily altered in the program.

  13. #27
    Join Date
    Apr 2007
    Location
    Adelaide Hills, South Australia
    Posts
    4,311

    Default

    Quote Originally Posted by BobL View Post
    Did a bit more on the program last night.

    The code is getting on for 450 lines or so...
    I think it was Pascal who wrote at the end of a letter... "I would have written less if I had more time".

    Elegant code becomes a secondary consideration when you are pushing to get something working.

    Anyway, sounds like you are making great progress Bob.
    Stay sharp and stay safe!

    Neil



  14. #28
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    Quote Originally Posted by NeilS View Post
    I think it was Pascal who wrote at the end of a letter... "I would have written less if I had more time".
    Knowing when to stop any project is an art in itself.

    Elegant code becomes a secondary consideration when you are pushing to get something working.
    At uni we used to push the students to write efficient code with lots of comments etc, we even allocated marks to these criteria. I told them exams was where they could be sloppy as long as it worked.
    Some years later a group of ex-students paid us a visit and I asked them about the difference between what they learned during their degree and work environment. One ex student said he had never written a comment in a computer program he had written at work. He wrote JIT (just in time) code that was designed to "break" or defeat a product - like smart rider tickets.

    Anyway, sounds like you are making great progress Bob.
    Well, progress but whether it's great or not will be clearer after proper testing.

  15. #29
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    Started doing some testing this arvo and realised I needed a software reset so that I could get the data more or less from the new unit and the calibration unit at the same time without waiting for the New Unit to get to the end of its current integration period. I could turn the power on and off or do a hard reset via the Arduino board reset but neither are as as quick as doing a software reset.

    So here is the state of the controls so far.
    The black knob controls the LCD screen contrast.
    Pause/Restart (should really be "continue" instead of restart") pauses the program so you can write down any data displayed on the LCD.
    Run/Reset, resets the current count - this is what I was referring to above.
    Integrations sets the number of counts. In low dust conditions you can collect more counts while in high dust you only need the minimum number or the counters in software will over-run.

    IMG_2334p.jpg

    I did a few measurements inside and outside and the first few data sets look reasonable comparable but I will need to collect a lot more data yet.

    The dustiest room in the house is the sewing room - where the PM2.5 micron concentration reached 79 ug/m^3 while my study is around 1/4rd of that. Fabrics are know to be quite dusty things.

  16. #30
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,756

    Default

    I spent most of today testing of the PMS7003 (hereafter called the PMS) dust sensor agains the highly sensitive Arti particle counter from work. I decided to use the Arti but as you will see it may not have been the best choice.

    As I envisaged it's not straightforward, mainly because the two units are different in a number of ways.

    There's lots of gory detail to report but I will only highlight a couple of problems and then give my update so far.

    The first issue is the counters measure slightly different particle size intervals.
    The PMS measures between 0.3-0.5 microns, 0.5-1.0 microns, 1 - 2.5 microns, 2.5 - 5 microns, 5 - 10 microns, >10 microns.
    The Arti measures 0.3-0.5 microns, 0.5-0.7 microns, 0.7 - 1 microns, 1 - 2 microns, 2 - 5 microns, >5 microns.
    Apples for apples comparison ranges are thus reduced to 4 ranges : 0.3-0.5 microns, 0.5-1.0 microns, 1 - 5 microns, >5 microns.

    The next issue is the Arti pumps at a different rate to the PMS so they don't sample the same amount of air/dust over the same time.
    Because the numbers of particle can fluctuate marked often over a few seconds you don end up sampling the same stuff.
    I had to adjust the the pumping periods so they were the same for both counters and then convert the volumes/particle numbers manually.

    With increased emphasis on monitoring smaller particle sizes the Arti is a device made for super clean room monitoring and so has a recommended maximum limit of 70,000 particles per litre. This is not much, so it means measuring higher concentrations of dust making becomes tricky. 70,000 particles per litre (depending on how they are distributed) represents only about 10-20 ug/m^3 of dust. This is a factor of 20 - 50 less that recommended for MDF dust. Above these levels the Arti counters may start to become saturated and will under report. I have to borrow the lower sensitivity counter from work to really compare the higher dust levels accurately.

    As a result the numbers at first appeared to be all over the place - not that different from many other comparisons I have made between particle counters. If you can get them to agree to +/-20% thats considered pretty good!

    So here are the 3 main things I have found so far,

    At low levels < 5 ppb (ug of dust per m^3 of air - this is like fresh clean outdoor air) the PMS7003 seems to be reading concentrations on average about 30% too low.
    Interestingly the total numbers of particles (<6000/L) agree between the two detectors to within about +/- 20% with a slight deficit of the smaller ones and a slight excess of the larger ones.
    I'm not worried about these low concentrations measurements - whether its 2 or 3 or even 4 ug of dust per m^3 of air doesn't matter. The Chinese would love to have air of that quality.

    At levels around 50 ppb the total numbers of particles/L agrees on average between the two counters to better than +/- 10% which is pretty good.
    However there were marked differences in some size ranges and so this did not translate into agreement of the ug of dust per m^3 of air, with the PMS reading too high by a factor of two
    Note: these levels are now above the recommended saturation limit of the Arti counter.
    Even if the PMS reads high by a factor of two its better reporting too high than reading too low.

    The third thing was I checked the consistency of the conversion by the PMS unit itself of its own "particles/L" across the 6 size ranges, into "ug of dust per m^3 of air" and they agree to <2%, and so they should as the sample the same data set.

    There is still some tidying up of the program to do but I need to see what happens at levels approaching the OHS limits for wood dust. The counter is very nice to use - just a bit slower than I'm used to but for ~$70 (The Arti costs about $5000) I'm not complaining too hard.

Page 2 of 14 FirstFirst 123456712 ... LastLast

Similar Threads

  1. Latest Shed Developments
    By RufflyRustic in forum WOODWORK PICS
    Replies: 56
    Last Post: 27th September 2008, 10:26 AM
  2. dust level sensor for cyclones
    By JDarvall in forum DUST EXTRACTION
    Replies: 0
    Last Post: 21st August 2008, 05:36 AM
  3. Table saw blade alignment - Lo-tech / Hi-tech
    By niki in forum TABLE SAWS & COMBINATIONS
    Replies: 6
    Last Post: 25th February 2008, 04:32 AM
  4. Power feed for HM-45 Mill....developments
    By Sterob in forum METALWORK FORUM
    Replies: 10
    Last Post: 13th May 2007, 07:58 PM
  5. Sensor Light
    By Garves in forum GENERAL ODDS N SODS
    Replies: 12
    Last Post: 9th June 2006, 10:42 PM

Posting Permissions

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