Does anybody know how they actually do it without a fan and a defined volume around it? The marketing fluff says "It applies sophisticated algorithms to measure the PM2.5 concentration directly in free space, without requiring a fan", so I assume the main difference to traditional PM sensors is the software?
TimByte · 8m ago
My guess is they're relying on natural air convection and using a more sensitive optical setup to compensate for the lack of forced airflow
trailbits · 1h ago
Also wonder how the sensor can stay clean without a fan. I suppose mounting upside down would help. Other fanless designs require periodic cleaning.
dist-epoch · 1h ago
The integration picture shows an "optical cover" transparent surface. I guess it's not meant to be used in highly contaminated areas.
> Laser light is emitted from the sensor and focused by the sensor lens at approximately 5 mm from the top of the
sensor’s lens surface.
> Particles traveling in free space due to the natural ambient airflow are detected when passing through the laser focal (sensitive) region.
> Due to the interaction between particles and light, the light scatters in different directions; a fraction is back-scattered towards the sensor, where the integrated photo-detectors detect it.
> The back-scattered signal is processed by unique algorithms (based on particle counts, particle relative velocity, probed air volume during measurement) to derive the particulate matter mass concentration.
trailbits · 57m ago
That's pretty much how all laser particle counters work... except the good ones use a fan and a chamber. Guess we'll have to wait and see how this compares to the reference sensors.
ImaCake · 52m ago
Yep, I suspect this is all marketing fluff and no substance. I see a lot of superlatives but no substantial technical breakthrough here.
TimByte · 11m ago
The fact that it's fanless and still works reliably makes me curious how well it handles airflow variability in real-world settings though
9dev · 3h ago
Waiting for Achim Haug of AirGradient! What’s your thoughts on this?
pppone · 1h ago
I've been running multiple Sensirion SPS30 PM sensors for years, and I'm honestly amazed how well they've held up. Particularly with respect to, within their spec'd error, how new/unused SPS30s report similar values to my heavily used ones.
Curious if there are any maintenance requirements for the bosch sensor.
Catbert59 · 1h ago
Many particle sensors are useless in foggy/hazy conditions, which ruins many citizen science projects in terms of data quality. Currently, the best solution is to calculate the dew point and then switching them off once you hit a specified limit.
How does this model deal with this?
TimByte · 5m ago
I didn't see anything in the article about humidity compensation
pppone · 1h ago
I don't think it would. Data in these instances would have to be ignored.
You can detrend for high humidities, but once water condenses, the only way round this would be to add a drying instrument.
HPsquared · 21m ago
I wonder if a radiant heat source beside the sensor, maybe also somehow focused in the same area, could get rid of the condensed water.
Catbert59 · 19m ago
The problem is not condensed water. This can be solved with heating.
It's that fog is being detected as a particle. This distorts the measured values.
smingo · 1h ago
The entire device is tiny: 20mm x 5mm x 5mm approx. It looks to work using a small laser and a lens which focuses the beam about 5mm above the device. The sensor unit contains a photodiode(s), and algorithms count the particulates. It looks to need a heat sink.
progbits · 36m ago
I mean this is cool but the size claims are somewhat misleading. They only sell you a sensor element on flatflex PCB, then you need to add own metal cage for thermal dissipation and a lens, so the overall size is significantly larger.
This is great for flexibility but very impractical especially for DIY space. Hopefully small fully integrated modules on something like edgeconnector PCB come out soon enough.
alliao · 2h ago
love this... while the politicians busy angle themselves over stances on whether air filtration is required for public space during a pandemic engineers and companies come up with products and solutions ready to deploy.. it is my dream that our last generation gave us clean-ish water, and we can somehow pass down clean-ish air...
arnejenssen · 3h ago
Cool. So we could get air quality apps for the phone or smartwatch?
mjs · 2h ago
Might need to wait a few generations. It costs ~30€ for a single unit (maybe half that at volume?). And not waterproof.
progbits · 32m ago
It needs an external lens anyway so you could shove it into the camera assembly, waterproofing is not an issue.
Cost is one, but this isn't something you can measure quickly on demand, you want to keep averaging multiple measurements over time. So unless you want to hold out your phone for minutes to get a measurement it doesn't seem practical.
> Laser light is emitted from the sensor and focused by the sensor lens at approximately 5 mm from the top of the sensor’s lens surface.
> Particles traveling in free space due to the natural ambient airflow are detected when passing through the laser focal (sensitive) region.
> Due to the interaction between particles and light, the light scatters in different directions; a fraction is back-scattered towards the sensor, where the integrated photo-detectors detect it.
> The back-scattered signal is processed by unique algorithms (based on particle counts, particle relative velocity, probed air volume during measurement) to derive the particulate matter mass concentration.
Curious if there are any maintenance requirements for the bosch sensor.
How does this model deal with this?
You can detrend for high humidities, but once water condenses, the only way round this would be to add a drying instrument.
It's that fog is being detected as a particle. This distorts the measured values.
This is great for flexibility but very impractical especially for DIY space. Hopefully small fully integrated modules on something like edgeconnector PCB come out soon enough.
Cost is one, but this isn't something you can measure quickly on demand, you want to keep averaging multiple measurements over time. So unless you want to hold out your phone for minutes to get a measurement it doesn't seem practical.