ESP32, a temp probe, and a fish tank - Electrical Guru's

[hilt_ctn]

For some learning.

I have a spare ESP32 kicking around and was thinking about adding a 18B20 temperature sensor to keep an eye on a fish tank temperature. That way if the heater ever packs up ( who checks the heater ?! ), I'll get an alert on a Home Assistant dashboard blah blah. The ESP would be powered by an AC adapter

Here's the question. What's to stop a ( unlikely ) 230v spike running through the AC adapter, the ESP and the sensor, turning the water live for a millisecond and all the fish to sushi before the distribution board breaker drops ? Or is everything such low voltage in the ESP system, everything between it and the water would fry first, posing no danger to the little swimmers

Just curious

 

[Imoe]

Have built something similar to measure my pool and outdoor temperature with the following components:

ESP32 integrated with esphome
2xDallas sensors one for the pool water and one for the outside temperature.
esp powered using a 10 000mAh power bank
power bank being is charged by a solar panel connected to a hkd lith charger
all components added to a waterproof enclosure

Works pretty well and all low powered

 

[hilt_ctn]

Have built something similar to measure my pool and outdoor temperature with the following components:

ESP32 integrated with esphome
2xDallas sensors one for the pool water and one for the outside temperature.
esp powered using a 10 000mAh power bank
power bank being is charged by a solar panel connected to a hkd lith charger
all components added to a waterproof enclosure

Works pretty well and all low powered

Hiya. I have a near identical setup for the pool. I'm only missing the sensor for the outside temp. All the electronics are housed at the pool DB with the water temp sensor on the end of a 10m cable ( with the help of a resistor ). Works a treat yes. And no risk of turning the pool into a giant vat of hotdog water !

So now I'm essentially trying to find out what would happen in an absolute worst case if there was a high voltage supply. My guess is the AC adapter would take the hit before anything dangerous could even reach the ESP. But that's my caveman understanding

 

[Gnome]

For some learning.

I have a spare ESP32 kicking around and was thinking about adding a 18B20 temperature sensor to keep an eye on a fish tank temperature. That way if the heater ever packs up ( who checks the heater ?! ), I'll get an alert on a Home Assistant dashboard blah blah. The ESP would be powered by an AC adapter

Here's the question. What's to stop a ( unlikely ) 230v spike running through the AC adapter, the ESP and the sensor, turning the water live for a millisecond and all the fish to sushi before the distribution board breaker drops ? Or is everything such low voltage in the ESP system, everything between it and the water would fry first, posing no danger to the little swimmers

Just curious

This is a very long topic and I can talk you through it, but to distill it down, if you are super paranoid, power your ESP from this:
(choose the voltage you need)
5v: https://za.rs-online.com/web/p/ac-dc-adapters/7910967
12v: https://za.rs-online.com/web/p/ac-dc-adapters/7910985

Both are medical grade power supply and they have very high isolation requirements because of certification for use with medical equipment hooked up to patients.

ESP32 runs off of 3.3v, but most boards take a USB input and then drop that down on the board to 3.3v which is why I linked to 5v.
I've not really seen 12v powered ESP32 but just in case.

If you need 3.3v, I still recommend you look for a board that is powered from USB instead or else build a 5v -> 3.3v regulated circuit because 3.3v power supplies become quite a bit more expensive as they are much less common. The 5v -> 3.3v will make no difference to your isolation.

EDIT: Medical isolation is as good as it gets for low prices (and well within the safety norms). But if want to see someone go to extreme measures to reduce the coupling between AC -> DC, this guy went all out for incredible high precision equipment. FYI this is not relevant for medical or even animal safety as the amount of isolation needed for the precision equipment he is using is sensitive to voltage that could be measured from walking across the room (ie. moving through air generates stray voltages). The equipment is incredible sensitive in other words, but he does a very good job of going over the problem and ways to reduce it (but again as you can tell from his video it is impossible to get rid of, except for things like light energy transfer or DC batteries in isolated systems, etc)

 

[hilt_ctn]

This is a very long topic and I can talk you through it, but to distill it down, if you are super paranoid, power your ESP from this:
(choose the voltage you need)
5v: https://za.rs-online.com/web/p/ac-dc-adapters/7910967
12v: https://za.rs-online.com/web/p/ac-dc-adapters/7910985

Both are medical grade power supply and they have very high isolation requirements because of certification for use with medical equipment hooked up to patients.

ESP32 runs off of 3.3v, but most boards take a USB input and then drop that down on the board to 3.3v which is why I linked to 5v.
I've not really seen 12v powered ESP32 but just in case.

If you need 3.3v, I still recommend you look for a board that is powered from USB instead or else build a 5v -> 3.3v regulated circuit because 3.3v power supplies become quite a bit more expensive as they are much less common. The 5v -> 3.3v will make no difference to your isolation.

EDIT: Medical isolation is as good as it gets for low prices (and well within the safety norms). But if want to see someone go to extreme measures to reduce the coupling between AC -> DC, this guy went all out for incredible high precision equipment. FYI this is not relevant for medical or even animal safety as the amount of isolation needed for the precision equipment he is using is sensitive to voltage that could be measured from walking across the room (ie. moving through air generates stray voltages). The equipment is incredible sensitive in other words, but he does a very good job of going over the problem and ways to reduce it (but again as you can tell from his video it is impossible to get rid of, except for things like light energy transfer or DC batteries in isolated systems, etc)

This is great info. Thank you both for your inputs