Thread: DIY Particle Counter

Somewhat dated but should get you pointed in the right direction. I used to have a how to build one for a grade school kids project but the link doesn't work anymore. Bob has built and modified some and if you do a search or two you might be able to find his posts on them. They have become so cheap to buy that making one might not be worth the effort.

Build an air quality monitor |

The Plantower PMS5003 and PMS7003 Air Quality Sensor experiment

https://uwspace.uwaterloo.ca/bitstre...pdf?sequence=5

Pete

Added. This is the thread I was thinking of. There might be more.
Developments in Dust Sensor tech

My experience has been that it's not possible to buy the parts to make a basic monitor cheaper than they can be bought fully assembled on ebay.

To give you an example my particle counters in 2017-18 cost around $70 worth of parts and a basic PM2.5 monitor in 2018 was available for ~$50

Do a search for "PM2.5 monitor" on ebay and you will see you can get them starting at around $50.

Are they any good?
Good question, - unless you spend at least a couple of grand to purchase a fully calibrated one they are all pretty much the same at measuring PM2.5 dust.

A huge thank you fellas, I have enough information to research.

Cheers Barry

Originally Posted by BobL

unless you spend at least a couple of grand to purchase a fully calibrated one they are all pretty much the same at measuring PM2.5 dust.

So would a monitor in the $1-200 range still be accurate enough to be useful at all? I'm looking at something like these for my bedroom to keep an eye on my allergies
Temtop P600 Air Quality Laser Particle Detector Professional Meter Accurate Testing for PM2.5/PM10 TFT Color LCD Display: Amazon.com.au: Home Improvement
VSON Air Quality Detector Monitor Grey Air Quality Meters Testing HCHO TVOC PM2.5 And PM10 with Accurate Air Quality Pollution Senor for Measuring Indoor Air Quality(Grey): Amazon.com.au: Home Improvement

Originally Posted by elanjacobs

So would a monitor in the $1-200 range still be accurate enough to be useful at all? I'm looking at something like these for my bedroom to keep an eye on my allergies
Temtop P600 Air Quality Laser Particle Detector Professional Meter Accurate Testing for PM2.5/PM10 TFT Color LCD Display: Amazon.com.au: Home Improvement
VSON Air Quality Detector Monitor Grey Air Quality Meters Testing HCHO TVOC PM2.5 And PM10 with Accurate Air Quality Pollution Senor for Measuring Indoor Air Quality(Grey): Amazon.com.au: Home Improvement

Without knowing what sort of sensor they have inside them its really difficult to know what level of absolute accuracy to expect.
However, they should be OK for the application you propose.

Thanks. Just been digging a bit further and found this collection of proper lab tests done for a range of air quality sensors against a calibrated reference. Seems the Temtop sensor accurately reflects changes for P2.5 but reads a fair way over actual levels (precise, but not accurate)

air quality sensor summary reports

Interesting to note how well some of the cheaper sensors performed while some rather expensive ones were rather abysmal...

Originally Posted by elanjacobs

Thanks. Just been digging a bit further and found this collection of proper lab tests done for a range of air quality sensors against a calibrated reference. Seems the Temtop sensor accurately reflects changes for P2.5 but reads a fair way over actual levels (precise, but not accurate)

air quality sensor summary reports

Interesting to note how well some of the cheaper sensors performed while some rather expensive ones were rather abysmal...

Thanks for that link ELan.
A few comments

As to be expected for most detectors the ability to measure PM10 is pretty ordinary to near useless. The problem with most budget sensors is they don't suck in enough air. This is less of an issue for PM2.5 as there are usually many more of these than PM10s. The relatively far fewer PM10 particles passing through the detector is a major reason for the large uncertainties. The uncertainty can be taken as the square root of the number of counts so if only 10 counts in one second are measured in 1 second the square root being about 3 makes it a 30% uncertainty.
This is why my home made counters integrate for 30 seconds before outputting values.
10 cps for about 0 seconds @10 cps = 300 total counts - now the square root of that is ~17 for a 6% uncertainty.

Reading through the list, one of the best performers at a very reasonable price is the "PurpleAir" (PA-II). However its not that useful for a shed as it's constructed for outdoor use and uses 2 x Plantower 5300 sensors in a weather proof container that is designed to be mounted to a wall. It has no display or control mechanisms but transmits data via wifi to a display device - presumably a mobile or tablet.

I see quiet a few devices are use the Plantower5300 sensor - slightly disappointing as these are not the latest sensors from Plantower.

Originally Posted by BobL

I see quiet a few devices are use the Plantower5300 sensor - slightly disappointing as these are not the latest sensors from Plantower.

Bob, can you share what you know about the actual differences (other than size) between the 5003, 7003, and A003?

Originally Posted by dmorse

Bob, can you share what you know about the actual differences (other than size) between the 5003, 7003, and A003?

The A003 is just a physically smaller version of the 7003.

There is some comparisons and links available on this page.
The Plantower PMS5003 and PMS7003 Air Quality Sensor experiment
They all use the same CPU and the main difference is in the air flow with the 7003 supposedly not accumulating dust over time as easily as the 5003

Despite being older the 5003 is still a good sensor especially for PM2.5. Many of these sensors are being used in professional air-quality monitoring networks. They are usually employed in sets of 3 in any one location with average readings taken of all 3 unless one deviates wildly from the other two. I read somewhere that the 7003 was better at measuring PM10 but my experience is the 7003 doesn't do that all that well either.

Thanks Bob. I've seen that report. Too bad they didn't do a tear down of the PMS7003 like they did with the 5003. It would be interesting to see if there were any hardware differences beside the fans. Maybe I'll have to sacrifice one each of mine.

Their data were interesting. I would liked to have seen a longer integration time than 30 seconds. Comparison would have been much easier if consistent scale factors were used on the graphs.

If I see the 5003 on sale maybe I'll get two more and do a similar comparison between them and the A003. I'll have to improve my SQL query skills though. That's would be a lot of work with a spreadsheet!

Originally Posted by dmorse

Thanks Bob. I've seen that report. Too bad they didn't do a tear down of the PMS7003 like they did with the 5003. It would be interesting to see if there were any hardware differences beside the fans. Maybe I'll have to sacrifice one each of mine.

Their data were interesting. I would liked to have seen a longer integration time than 30 seconds. Comparison would have been much easier if consistent scale factors were used on the graphs.

If I see the 5003 on sale maybe I'll get two more and do a similar comparison between them and the A003. I'll have to improve my SQL query skills though. That's would be a lot of work with a spreadsheet!

A teardown comparison would be interesting.
I have 6 7003s, an A003 in DIY units. I also have one of those ebay budget TFT colour screen output sensors with a 5003 in i . I have been running 1-2 of the 7003s side by side with the 5300 but direct comparisons are tricky because mine integrate over 30s and they are not all necessarily sampling the exact same air/dust. To perform the latter they would need to be placed inside a well mixed chamber.

The unit with the 5300 in it is now more or less permanently in the kitchen about 1m from the stove - I want to see how long it lasts in that environment.\

Originally Posted by BobL

A teardown comparison would be interesting.
I have 6 7003s, an A003 in DIY units. I also have one of those ebay budget TFT colour screen output sensors with a 5003 in i . I have been running 1-2 of the 7003s side by side with the 5300 but direct comparisons are tricky because mine integrate over 30s and they are not all necessarily sampling the exact same air/dust. To perform the latter they would need to be placed inside a well mixed chamber.

The unit with the 5300 in it is now more or less permanently in the kitchen about 1m from the stove - I want to see how long it lasts in that environment.\

Well, I was motivated to order two more 5003s. In April I'll have three of those and can compare them to the three A003s that I'm currently using. The big problem will be generating enough dust to test over over a wide range. That's easy to do with smoke but that doesn't really test the PM10 values that are so important with wood dust.

Originally Posted by dmorse

Well, I was motivated to order two more 5003s. In April I'll have three of those and can compare them to the three A003s that I'm currently using. The big problem will be generating enough dust to test over over a wide range. That's easy to do with smoke but that doesn't really test the PM10 values that are so important with wood dust.

The last time I remember seeing HEAPs of PM10 particles in the shed was the time before the last time when I was bluing metal in the shed.
The bluing/blackening process itself involves acids but the last step is to heat all the workpieces in hot (100º+) engine oil.
I do this by filling a 5L SS saucepan and heating it in an electric frypan.
To speed things up I set the fry pan on max for a while before turning it down to "simmer" - heat the pieces in the oil for 10 minutes or so and then turn the fry pan off and let the whole thing cool down.

That time i set the pan on max and forgot about it and went up to the house. SWMBO was up there and after a while said - I can smell something burning. Then I remembered the pan and raced down to the shed - even before I got there I could see a fine blue haze around the shed. I opened the door and there was so much blue haze inside I could barely see across the shed.
I cant remember the PM10s but they were the highest I have ever seen.