We have a lot of wildfires in Oregon and the TV is warning of seriously bad air quality due to smoke.
OK, so I took the Nikken 1394 outside. It registers 2000 particles/liter. The scale is yellow indicating no big deal.
HMMM....
I dug out the Dylos DC1100 Pro Calibrated to a particle size split at 2.5 microns.
It read 5322 total count particles per .01 cubic foot.
96% categorized as below 2.5 micron. The scale on the chart is in the range "kiss your sweet ass goodbye", which is in line with the air quality warnings.
After accounting for the units difference, there's an order of magnitude difference between the readings.
If I take both devices over to the sofa and bang on the cushions, I can drive both off scale.
Symptoms suggest that the problem is relative to particle size.
The Nikken (what passes for a manual) gives no clue as to particle size. The Nikken website has no info. This is all I found on the web
formatting link
I'd rather not take it apart if it's working as designed. Anybody got any relevant info?
I've done some aerosol particle counters down to nm-level. Th Nikken unit looks like it has simple LED-based optics and maybe electrometer-based measurement, which typically measures total 'mass' of particle down to nm-range. There's not a lot of mass in small particles, so electometers are typically used with size selectors when measuring nm-size particles.
Dylos has two detectors, one with a cut-off of 2.5um and other with
500nm. Cut-off is the diameter where 50% of the particles are counted.
Nikken has a cut-off probably in the 1-5um range. And it might not be counting particles, but rather using 'photometric mode', where the total intensity is used to estimate particle count. This is used also in scientific particle counters when going over the capability of counting single particles.
The count of smaller than 2.5um particles can be really high on forest fires. I always remember when we got our condensation particle counter prototype working in the lab, it indicated 5M particles/l. We thought it was an error, but we had just reached way better cut-off of 7nm than we thought and there was a big forest fire about 300km away!
The difference (2000 vs 19000/l) you're seeing is well explained by different cut-offs.
That was my conclusion, but...If I understand correctly, the Nikken measures reflections from particles. The amplitude of the light reflection should be proportional to the square of particle size. You could sort particle size by reflection amplitude.
Dust on the optics could reduce the reflection amplitude and make huge difference in measured amplitude. The detection threshold would move up to larger particle size.
If the cutoff is correct as designed, there's nothing to be done. If it's changed because of optical contamination, taking it apart to clean the optics might be useful. I have no idea of the history of the device. It may have been used in an environment that led to dust coating the optics.
For a device where you can't filter the input, wouldn't you expect optical contamination? Maybe it calibrates itself before it starts the fan???
I've been using the Nikken as a survey tool. In its current condition, it appears to be useless in finding contaminants that are dangerous to health. It does appear to be sensitive to household dust like you get when you pound on the sofa cushions.
If the Dylos will run on 8V, I think I'll make a battry pack for it.
The Dylos has a real-time bar graph. It sits mostly at the zero end of the scale, then flashes periodically to full scale. There's a long averaging time in the numerical readout. That suggests an infrequent hit in a small volume averaged and scaled up by a huge amount.
It's been years, but I thought I'd read that the Dylos does the same dark-field reflection method and differentiates particle size by amplitude of the signal.
I have slightly better knowledge of the Dylos's history. Bought it from someone who'd bought it new and used it in controlled 'inside' conditions.
You're no fun. Learn by destroying which means tear it apart until you understand how it works.
Page 4 of the manual proclaims: Level Color Particles/liter of air 1 blue 500 2 yellow 1000 3 yellow 2000 4 yellow 3000 5 red 5000 6 red 5000+
Link to a previous discusion about a light scattering nephelometer: See: for a list of links to various similar devices by Sharp and Sinyei along with explanations as to how they operate.
The sensor has a big CAM-003 inscribed on its case, but no manufactory and hints as to who made it. I can't find anything useful with Google. General image search for nephelometer: which finds some interesting stuff.
"Sizes of airborne particles as dust, pollen bacteria, virus..." Burning Wood 0.2 - 3.0 microns
25 dust sensors on Digikey: "Wildland fire ash: Production, composition and eco-hydro-geomorphic effects"
Sometimes yes, sometimes no. The reflectance can be different (think of soot and water droplet). For small particles you start getting interesting interference patterns. The refractive index of liqud droplets differ. For larger particle concentrations you get multiple particles in the beam at the same time (coincidence).
Google for aerosol particle sizer for more info.
You can arrange the airflow so that it keeps most of the stuff away from optics, but keeping everything out is almost impossible without additional filtered flows.
Based on your experience, I think I know how it works. The question is not how, but what is this particular model Nikken 1394 actually measuring in terms of particle dimensions. My symptoms suggest that it does a poor job on forest fire smoke from 200 miles away. Is your experience different? Is mine broke? Or just not suited for very small particles?
Thanks for the links. I'm sure that's all good stuff, but does it answer the question, "what can I expect from the Nikken?" Should I take it apart in anticipation that cleaning it will help measure small particles? If I can't, I should just not try to use it for that. The Dylos seems to read closer to what the Air Quality websites publish. I built a battery pack for it, so now it's portable.
Almost all decent particle counters measure scattering cross-section in a darkfield geometry (air flow along X, laser along Y, detector and collection mirror at +-Z). Some use a polarizer on the detector to reject the CW scattered light (and its attendant shot noise), but that's usually only an issue with PMT detection, whereas most use straight photodiodes. It usually isn't worth working that hard, because in an incoherent light scattering system the detected signal goes as the sixth power of the particle radius, so a factor of 10 sensitivity improvement takes you from 0.2 um to 0.14 um, say. (This is in the Rayleigh limit, of course.)
The measured signal depends on position in the beam, composition, density and shape as well as size, so the size readout is calibrated using polystyrene latex (PSL) spheres, sometimes with a histogram correction to account for the effects of impact parameter (i.e. the nearest distance from the beam axis).
Crappy ones just look for average extinction using a LED, just like a smoke detectors.
I know how it is suppose to work. That's in the literature for the various sensors found in the mess of links I provided. What I don't know is what this particular device can do. It is not a scientific instrument, show no sign of calibration or testing, and seems to be made for the paranoia market.
Phil Hobbs explained most of that. However, it looks like all you want is a number. I don't have specs on the specific model light scattering box inside the unit, but I think this one is close: The data sheet does not specify the range of detectable particle sized, but does mumble: GP2Y1010AU0F is a dust sensor by optical sensing system. An infrared emitting diode (IRED) and an phototransistor are diagonally arranged into this device. It detects the reflected light of dust in air. Especially, it is effective to detect very fine particle like the cigarette smoke. In addition it can distinguish smoke from house dust by pulse pattern of output voltage.
Ok, so it can detect cigarette smoke and house dust. Digging through a handy table of dust particle sizes at: I find: microns Burning Wood 0.2 - 3.0 Fly Ash 1 - 1000 Combustion 0.01 - 0.1 Household dust 0.05 - 100 Smoke from Natural Materials 0.01 - 0.1 Nothing found for cancer sticks (cigarettes).
House dust and fly ash have too wide a range of particle sizes to be useful. That leaves burning wood, combustion products, and smoke from natural materials as useful ranges. Therefore, I would guess(tm) that the Nikken can detect particles from 0.01 to something less than 3.0 microns. That fits nicely with most of the smoke produced by the fires. Normally, I would verify this by simply hanging a sheet of sticky paper out the window for a while to collect the local dust and ash. Then, a microscope and reticule to determine the particle size from the fires. Fortunately, the EPA has saved me the effort: Particulate matter in wood smoke has a size range near the wavelength of visible light (0.4 - 0.7 micrometers) That's well within my guess(tm) as to the detection range of the instrument, so it should detect smoke from the fire.
There's quite a bit of information in the above EPA document on what constitutes harmful concentrations of smoke. Methinks that's what you should be interested in calculating, not the number or size of the particles. The document probably explains it better than I can.
Most areas have air quality monitoring stations with associated web pages. This is the one for my area: You should be able to find the approximate smoke concentration on such a site.
I don't know. We had a few bad days were there was quite a bit of haze from the fires. However, I didn't bother checking the concentration with the Nikken for that fire, or any previous fires.
I don't think so. It responded normally (2000 particles/liter) when you went outside. You might want to try my favorite test. Find a cloth sofa or padded seat. Fire up the Nikken meter. It should read low on the graph. Now, beat on the sofa cushions once or twice. The meter should indicate full scale. It's quite sensitive to house dust.
No. The particle sizes calculated by Phil Hobbs is the minimum particle size. There is no maximum expect perhaps limited by the input air filter. I think it's the box to the right of the fan: held in with one obvious screw (but I'm not sure). Blow with clean and dry compressed air. If there's a big pile of dust inside, the IR sensor box is probably also full of dust.
I can't answer that directly. What I've tried to do is use data from a similar IR dust sensor to deduce what the Nikken might do. Without data on the sensor used, I can't do much. I was hoping that you would dig through the mess of links and do your own research. I can only do so much.
Unless you're prepared to go through some kind of calibration procedure, the best I can suggest is to have me run some kind of detection test and compare results on your unit. I could set fire to various common objects and hope that the type and quantity of smoke produces is identical.
Well, if you have two units, and one of them is suspect, take the suspect unit apart as I previously suggested.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.