Ever wanted to measure soil moisture? Common “soil moisture meter module arduino raspberry compatible free shipping” PCBs might deceive you with their ascetic looks. Today, [Raphael (@rbaron_)] is here to teach us (Twitter, unrolled) what it takes to build a soil-embedded sensor that can actually survive contact with a plant.
As the picture might hint, waterproofing is of paramount importance, and soldermask doesn’t quite cut it. Raphael describes his journey of figuring out approaches and coatings that would last, starting from simply using nail polish, and ending with the current option – a rotisserie-like device that rotates sensors as the coating applied to them dries, mitigating a certain kind of structural failure observed long-term. With plenty of illustrative pictures and even a video of the rotisserie device in action, you’ll quickly learn things that took time and effort for Raphael to figure out.
This isn’t the first time Raphael shares some design battlefield stories and lessons with us – he has taught us about overall capacitive moisture sensor principles, too! If that interests you, we’ve covered quite a few moisture sensor designs, from cheap but hardy two-nails designs to flip-dot-equipped ones, and some of us take the commercial designs and upgrade them!
We thank [Chaos] for sharing this with us!
When I started working on my soil moisture sensor, I had little idea about how harsh an environment the soil really is. In this 🧵, cool failure modes & solutions when sticking electronics in the ground: pic.twitter.com/ROGvLSPgk9
— Raphael (@rbaron_) February 9, 2022
>> their ascetic looks
ascetic
/əˈsedik/
adjective
characterized by or suggesting the practice of severe self-discipline and abstention from all forms of indulgence, typically for religious reasons.
“an ascetic life of prayer, fasting, and manual labor”
I think the word you wanted there is “aesthetic”.
“Ascetic” is exactly the world that fits there, substitute “simple” if you have trouble understanding it. “Aesthetic looks” would be a tautology.
If “ascetic” was troubling enough, I doubt “tautology” is going to be very helpful.
I do appreciate any opportunity to use the word “tautology” in a sentence, however.
Another hack only for native english speakers :D
Nice tests.Love to see some experimental DIY Granular Matrix sensors: https://agsci.oregonstate.edu/mes/irrigation/granular-matrix-sensors
Those GMSs look permeable. Anything permeable will become a home for life forms.
I’ve gone through this exactly. Conformal coating all the way to a thick epoxy coating. I determined that you basically cannot make these sensors surviveable. The problem is that any moisture that gets through will eventually permeate through the fr4 and destroy the calibration. If you try to make an ultra thick coating it destroys the sensitivity of the sensor.
I went through numerous iterations and was never able to get one of these types of sensors to stay calibrated for more than a month. Consequently I don’t recommend them.
Assuming that there is watering involved, you can easily recalibrate automatically whenever the signal goes above some threshold, so this doesn’t seem like a major problem to me.
You have to solve the software half too, it isn’t enough to just fix the hardware.
Long time ago, I used two 1/4-20 stainless bolts set in an plastic box with a silicone seal, it lasted years-never wore out. Recently bought a bunch of wireless moisture sensors to hack, the two-part soil probe appears to be aluminum.
No, ascetic also works. The cheap, poorly-made PCBs are constructed very simply, sparsely, in a way which could be compared to e.g. a monastic cell: only the bare minimum.
Anyone who would stick a PCB coated or not in water or dirt should stay away from doing anything electronic related.
Learn how to do something before you just start guessing.
“Learn how to do something before you just start guessing.” — No, stuff your elitism. One of the best ways of learning is trying and failing.
Your attitude is atrocious.
He’s not wrong though.
Also, people are _selling_ these doomed-to-fail sensors, and people are buying them up.
https://www.aliexpress.com/item/32840409177.html
Eh , I’ve had one in an indoor pot now for 2 years and it’s been fine. It’s not nearly as moist as some soils are though. It’s not easy finding a sensor that connects via rs485 modbus in that price range. Sure for 4x the price I can find one from Seeed Studio
https://www.tindie.com/products/miceuz/modbus-rs485-soil-moisture-sensor-2/
Likewise, I used one for about a year, no signs of any corrosion – the plant died for other reasons, so the sensor got retired.
Key to long life I believe is only turning the sensor on when you want to take a reading?
The type of corrosion in the image above would happen regardless of being powered or not. If you really want long life you would have to remove it from the soil. But then if you get one that chirps when you need to water it won’t alert you so….
Since my product has been mentioned, I am obliged to chime in.
I know this challenge very well, been working on it for over 7 years already.
At first I have used acrylic based spray, but alas, acrylic is hydrophilic! It worked fine, but degraded in a year or so. Then I have used non-polyester epoxy – high end one, hell to work with as you have to correctly maintain humidity, temperature etc, etc. This resulted in bulled proof sensors, but proved to be economically non-viable. Then in parallel I have started using polyurethane spray – I must say, it’s a pretty good compromise – takes a bit of care to properly apply and sensors stay operational for over a year. Actually I have one reporting reliably for over 3 years in my flower pot.
I’ve used all of the products mentioned here, including miceuz’s (very well-made) rugged sensor. One thing that hasn’t been discussed much is potting medium: the coarser the medium, the faster sensors will fail—regardless of what coating is used. Bonsai akadama clay? Forget about it.
Are you suggesting that you never learn anything from your projects?
Offcourse you can learn from failures.
However, most failures can be avoided if you just stop and think a bit before you do something.
Reading about what has been done in the past also helps to avoid making the same mistakes.
Firstly, this thing measures resistance, not moisture. Althouh there is a correlation between the two, there is not necesarily a good way to convert the one into the other, it is at best a guestimatye and, if it is used outside, you will get non-sensical measurements once it starts freezing.
Secondly, for any long-term measurements using such a method, you should avoid DC and use AC. Long term DC means you will be basically electroplating your electrodes and therefore changing the measurements over time.
I was confused by your reply until I understood the ambiguity of my message.
You parsed it as;
“Are you suggesting that one never learns anything from their projects?”
While I meant;
“Is this person suggesting that they never learn anything from their projects?”
That’s not the harsh environment. It’s an electrochemical reaction. I once dissolved a perfectly fine stainless steel spoon when using ist as a soil sensor electrode….
Capacitive soil moisture sensors don’t have the same problem. Resistive ones are the ones that have that electrochemical reaction and are why when I first got some I stayed clear of resistive sensors. These days I just use resistive sensors with replaceable stainless steel probes.
Conductive sensor are as they name imply, conductivity sensors, which are not moisture sensor per se.
Any change in dissolved ions will impact resistivity even if moisture content is the same.
I’ve been working on EC sensors in my spare time, and yeah, water will eventually find it’s way into the FR-4… Sealing the edges helps a lot. But it will even creep through soldermask slowly. Stainless steel electrodes help, but are a *** to solder!
Try the type that have replaceable probes. That way you only need to solder once to the female coupling/nut side. And use dialectic grease on the threads
dialectic grease is perfect to cure logorrhea. Can´t think of a better logical lubricant.
I’ve buyed cheap HCL based gel flux (for $4) and it was quite easy to solder wires to stainless cuttlery using regular SnPb solder. There are even better fluxes made specificaly for stainless which work even better, while being in similar price range.
Did this get covered in the hackaday winner Vinduino project? I remember him mentioning sensor designs and how they failed. Seems he used stainless steel bolts. https://hackaday.io/project/6444-vinduino-a-wine-growers-water-saving-project
Maybe someone should try carbon electrodes, connected with spring contacts
https://www.vegetronix.com/Products/VH400/gindex.phtml
Different tech avoids this problem to some degree.
If you coated these parts in Cerakote. This would never happen. You can apply Cerakote over nearly any surface and it will not degrade. I recommend Cerakote Elite Series.
Here’s s video showing the corrosion protection:
https://youtu.be/IL_bXx_qtfo
2034 hours is just over 96 days. A bit over 3 months.
I have had the same problem with f****** coating on the chinese sensors.
I ordered the original sensors from:
https://www.ramser-elektro.at/shop/bausaetze-und-platinen/giesomat-kapazitiver-bodenfeuchtesensor-erdfeuchtesensor-mit-beschichtung/
There is a polyurethane coating on it.
No problems anymore.
The sensors on the linked shop are the original one!
They was developed in a forum:
https://www.mikrocontroller.net/articles/Giess-o-mat