How to make a laser particle counter?

[followbreathe]

The heavy haze in Peking(China) is getting more regular. I am expecting a big rise on the laser particle counter requirements for inspecting pm 2.5 and air filter performance.

I saw a tear down on dylos laser particle counter: http://woodgears.ca/dust/dylos.html. It seems it could be built with laser diode, photodiode, operational amplifier, microprocessor, and fan.

I am quite confident there are people here have enough knowledge to build it. If it happened that you have the knowledge and also want to share. Would you please shed some light on how to build it?
How can we sense the light reflected by the small particles.  Thank you!

[SeanB]

Normally a simple micron size filter and then a simple baffled chamber with a darkened interior, and a small semiconductor laser shining into the long axis, with a sensitive photodiode coaxial with it to detect the returned light from particles that are reflected by the beam. For smog and haze likely you will just count the number of reflections per unit of time, with the air in the chamber being exchanged at a regular basis via a fan.

[Richard Crowley]

It looks simple enough. All the basic components are readily available. The special shape of the air path isn't trivial, but well within the range of DIU. 

IMHO, the big unknown is the sensor: what are its characteristics:  sensitivity, view "width", etc. etc.  And then the calibration. Basically, the sensor is putting out some kind of "analog" pulse. What is the ampllitude (and duration) of the pulse?  How do they distinguish between "large" and "small" particles?  There must be some sort of decision based on amplitude and/or duration of the pulse. Clearly the microprocessor is sorting out the "messy" analog pulses and counting them as "large or "small" particles, etc.  And how do you calibrate the amount of air going through the sensor?

Of course, if you wanted only a relative indication of "dirty" vs. "cleaner", maybe calibration isn't a big deal.

[followbreathe]

Basically, the sensor is putting out some kind of "analog" pulse. What is the ampllitude (and duration) of the pulse?

Thanks for the replies! But I am slow to catch up, I have some Vishay bpw34 photodiode, and not sure whether it is fit for the job. I hope I can view the analog pulses on the oscilloscope. Can uCurrent do this? In case I don't have a uCurrent, any other references?

[Richard Crowley]

Not clear what role the uCurrent plays here?

I found some interesting things when I did a Google image search for: optical particle detector circuit

[TSL]

Many years ago I used to work in such things for a pollution analysis company. We had many things made by Horiba, Thermo Electron and others.

One device we had, and I can't remember who made it, was a device like you're describing.

A know volume of air was sucked into a chamber ( over time T)  with a laser at one end and a PMT at the other and that chamber had a high density coil wound around it that created a rotating magnetic field. This caused the sample to rotate and thus if a particles reflection went from a dot to a line and back in 360deg rotation it was considered a fibre and so dots and fibers could be counted per rotation and volume within T. Rotation also allowed lumpy particles to be seen easily.

It was portable in a small suitcase and was very popular for use in brown coal mines given its fibrous nature.

We had other particulate counters such as one that had a reel of filter paper about 1inch wide. A radio active source below the tape and a counter above it. First the tape was irradiated for time T and then that portion of the tape was moved into a clamp where the air sample was sucked through the tape for time T, and then the tape moved back to be re irradiated for time T and the difference between before and after gave you your particulate count per cm3.

To do this right you need to control your air flow so a known cm3 or so pass through whatever is doing the measuring so you can extrapolate that to m3 etc.

Also dependent on the particle or fibre some are more reflective than others so amplitude of the reflection is not a good measure of particle size.

And.. I just found this link to a similar device for asbestos monitoring...

http://www.opticsinfobase.org/oe/fulltext.cfm?uri=oe-21-9-11356&id=253224


that illustrates magnetic rotation and particulate discrimination of the sample.

cheers

Tim