AHT10 humidity sensor over Software I2C

[mrburnette]

it turns out my pcb humidity sensor thought experiment is more than just a thought experiment
the real implementation of this concept is called *capacitive soil moisture sensors*
https://www.mdpi.com/1424-8220/20/19/5644
https://www.aliexpress.com/wholesale?ca ... capacitive

And, as expected, the inductance side is also covered.
https://www.hindawi.com/journals/js/2013/894512/

[ag123]

Touching on this topic analog interfacing e.g. capacitance, inductance leads to a lot of other unrelated areas that falls into google search results.
Notably capacitance and inductance sensing, a lot leads to many non-trivial, complicated non-linear sensing topics such as capacitive touch sensing, capacitive, inductance sensing such as humidity, soil moisture etc and others like proximity sensing capacitive and inductive etc. including converting such sensors into force sensing, pressure sensing etc. a very broad arena that often don't have straight equation / formulas for the analog to measurements conversions.

one of those things that i've wanted to do with inductive sensing is to make voice coil accelerometers, but didn't get down to trying it
http://roevalley.com/newsbrowser/interface/quake.htm
https://www.eeweb.com/earthquake-detect ... lerometer/

[blue-man]

one of those things that i've wanted to do with inductive sensing is to make voice coil accelerometers, but didn't get down to trying it
http://roevalley.com/newsbrowser/interface/quake.htm

That looks interesting - but the picture does not really show details of the construction of his "voice coil accelerometer".
The idea to use old hard disks is really good.

I think an GY-521 mounted on something will not be so sensitive.

Now we have really left the humidity theme ...

[ag123]

Accordingly, that kind of 'moving coil accelerometer' has a name for it, it is called a geophone
https://en.wikipedia.org/wiki/Geophone
http://www.geol.lsu.edu/jlorenzo/Reflec ... phones.htm
https://www.essearth.com/the-geophone-h ... the-earth/

They are sold for quite a premium
https://www.sparkfun.com/products/11744
https://www.aliexpress.com/wholesale?ca ... t=geophone

but I'd guess, for 'fun' purposes it is possibly feasible to create a 'ghetto' one that is basically a coil with a magnet e.g. the neodymium 'coin' types.
And we have pretty much a 'similar' thing.

The 'cheap' ones are our 'cellphone' semiconductor (mems?) accelerometers
accordingly, ubc has pioneered it:
https://myshake.berkeley.edu/
then closer today google does it as well
https://www.science.org/content/article ... arthquakes

I'd guess the moving coil types is likely more sensitive as in part we can control the design, e.g a small magnet suspended from a string.
But as usual, unlike 'digital' sensors, we'd need to convert that voltage we measured into 'acceleration, velocity' etc. It is an 'analog' world we live in.

one of those aspects that had me procrastinating about making a ghetto geophone is that for 99.9% of times, it probably detect nothing and arbitrary noises. then 0.1% of times, maybe it is a real earthquake.
edit: but a few more google searches later i stumbled into this
https://create.arduino.cc/projecthub/te ... ure-702324
and this could be that ghetto design
http://www.vibrationschool.com/mans/DBS ... dxr03.html

[blue-man]

They are sold for quite a premium
https://www.sparkfun.com/products/11744
https://www.aliexpress.com/wholesale?ca ... t=geophone


one of those aspects that had me procrastinating about making a ghetto geophone is that for 99.9% of times, it probably detect nothing and arbitrary noises. then 0.1% of times, maybe it is a real earthquake.

$60 for 0.1% fun is heavy, so the HDD variant is really more interesting.

[ag123]

$60 for 0.1% fun is heavy, so the HDD variant is really more interesting.

well a ghetto design would likely be something like this
http://www.vibrationschool.com/mans/DBS ... dxr03.html
the parts to overexercise brain cells for it (i.e. think) is that spring, magnets, the mass, then convert the analog voltages first to velocity then to acceleration and how to power it and capture only the meaningful data as it does 'nothing' 99.9(99) % of times
a thing though is small neodymium magnets are still quite affordable for such a ghetto effort
https://www.aliexpress.com/wholesale?ca ... ext=magnet
as for the 'spring', if space is less of a constraint and say you need 10hz or even smaller say 1 hz a solution is here
https://en.wikipedia.org/wiki/Pendulum
the pendulum is the magnet, then you'd still need to figure out how to do all that damping
oh and maybe to optimise the return-on-investment, the pendulum can work as a clock

[blue-man]

well a ghetto design would likely be something like this
http://www.vibrationschool.com/mans/DBS ... dxr03.html
the parts to overexercise brain cells for it (i.e. think) is that spring, magnets, the mass, then convert the analog voltages first to velocity then to acceleration and how to power it and capture only the meaningful data as it does 'nothing' 99.9(99) % of times
a thing though is small neodymium magnets are still quite affordable for such a ghetto effort
https://www.aliexpress.com/wholesale?ca ... ext=magnet
as for the 'spring', if space is less of a constraint and say you need 10hz or even smaller say 1 hz a solution is here
https://en.wikipedia.org/wiki/Pendulum
the pendulum is the magnet, then you still need to figure out how to do all that damping

Yes - but with an defective HDD you get all of this for free - two big neodym magnets in a good parallel holder and a big detection coil centered with a perfect ball bearing.

[ag123]

HDD magnets are good if they are available, but i'd think tiny ones like those 50 magnets for a dollar types may be adequate for this ghetto effort.
All it takes is some coils (if you are lazy to wind them, i'd guess inductors for radios works)

[AndrewBCN]

Sorry for posting this a few months later, but I just wanted to note that I have deprecated the use of the AHT10 in the STM32 GPSDO project and I now recommend the use of the AHT20 instead of the AHT10, basically because it costs essentially the same and is slightly more accurate. The AHT10 and AHT20 modules are pin-identical and have similar dimensions too, and the Adafruit library AHTX0 handles both sensors identically (no code change necessary).

I have not had the time to investigate the potential problems caused by the AHT10 on the I2C bus but to me, this is a moot problem at this stage: just use an AHT20 instead, if you want to avoid any potential problems.