DS18B20 durability and reliability problem

Hello all,

I've been trying to make am all-in-one brewing system which has been a success so far (at least on the software side) but I have problem with DS18B20 that I am using as the temperature probe for my project.

Recently, since I put the system together and tried an actual brew, the sensor failed halfway through the mashing step (duration 1 hour, 65 deg C). I then changed the probe for a new one and pretty much same thing happened. then I tried different sensors, those who come inside a metal sheet, also the "naked" sensor itself, bla bla bla and all seem to fail after 40 minutes inside the hot water. (and by fail I mean it shows -127C)

However, when I take the sensor out of the hot water, a few seconds later it comes back to life and works normal, it seems to me that it just cannot handle high temperatures for longer than a couple of minutes. Has anybody faced the same issue? Am I making a foooolish mistake here?!

Please help!

Thank you!

mramir: Am I making a foooolish mistake here?!

Quite possibly. The first thing you need to consider is that DS18B20s are good for 125C, very popular, and they have been around for years. So, if they are really as unreliable as you suggest, everybody would know about it.

So what's the problem? God only knows. Maybe you've got a bad, untypical, batch. Maybe the problem is at the other end of the cable. I'm sure you are aware that -127 means no connection.

I had one on a water tank which went up to 60C and ran for over a year. It was dry though.

Nick_Pyner: So what's the problem? God only knows. Maybe you've got a bad, untypical, batch. Maybe the problem is at the other end of the cable. I'm sure you are aware that -127 means no connection.

What is actually what is puzzling me is that why it begins to fail above 65 C?? or some other times above 70C?? or after a couple of minutes?! I can't figure out what is going wrong here, I bought different sensors from different sources, I even have both maxim and dallas sensors.

Any idea for how long a DS18B20 can be exposed to temps above 70C?

According the datasheet:
Operating Temperature Range … -55°C to +125°C
Storage Temperature Range … -55°C to +125°C
the time is not specified so it simply means the lifetime.
Are you sure you bought genuine sensors?
People frequently refers problems with chinese counterfeit sensors from ebay or Ali.

I’m using DS18B20 for my home heating system. For an example, boiler sensor is working in conditions up to 80-82°C for years.

Budvar10: People frequently refers problems with chinese counterfeit sensors from ebay or Ali.

Since I am living in Iran have no way to verify the real origin of the sensors, but I am assuming they are coming from China 99%. Having that said I have sensors that reads"DALLAS/18B20/1916C4/+051AG" is it possible to check their genuinity using that information?

How do you power the sensors? Parasitic through the data pin or do you power it via the 5V pin?

is it possible to check their genuinity using that information?

Try it on their web maximintegrated.com.

Read this https://electronics.stackexchange.com/questions/456096/which-one-of-these-ds18b20-temperature-sensors-is-fake the #6 is interesting.

Budvar10: Read this https://electronics.stackexchange.com/questions/456096/which-one-of-these-ds18b20-temperature-sensors-is-fake the #6 is interesting.

Indeed it was! and just by appearance I can tell that my sensors are probably fake. Well, pretty much everything is fake in this country.

Thanks for the help!

Smajdalf:
How do you power the sensors? Parasitic through the data pin or do you power it via the 5V pin?

5V pin. Because the datasheet says parasitic could become noisy and problematic for higher temperatures.

You could try to buy from some reputable eshops like Mouser, Digi-key, Farnell, TME, SOS .... However, I do not know if some of them sell to customers from Iran. There are sanctions since 2018 started by USA and EU accepted it also, but I have no idea which goods are concerned.

Sadly none of them ship to Iran. The unit itself doesn't seem to be sanctioned (or maybe it is, as Trump's sanctions have reached almost our underwear).

Maybe I will have to buy a whole box of this sensors and replace them after every brew. Price wise, it is not that expensive compared to hops and other consumables. Or I have to go deep inside the black market of electronics in Tehran where they sell salvaged parts and sometimes you can find very rare and genuine parts and look for a decent temperature probe salvaged from a old hot liquor tank. I actually managed to find a genuine Japanese circulating pomp.

What about using some other sensor? LM75 which is I2C. Or analog - LM35 or a simple thermistor? Also any diode junction voltage drop may be used to measure temperature.

mramir: and all seem to fail after 40 minutes inside the hot water.

Have read somewhere that sensors in metal tubes are only splash/damp resistant, but not waterproof. Must use those sensors inside some second barrier (longer tube, filled with RTV?). Leo..

Smajdalf: What about using some other sensor? LM75 which is I2C. Or analog - LM35 or a simple thermistor? Also any diode junction voltage drop may be used to measure temperature.

Surely these are other options, but given my situation I prefer to use something without semiconductors.

Why? The DS18B20 is expensive chip. I think it is almost sure you will get a counterfeit if you buy it at less reputable seller. I would try some less advanced one, both LM75 and LM35 has max. temperature 150°C which may be true even for a generic copy. You can always test it in a cheap setup i.e. boil it in water for 3 hours and monitor if it works.
Or a thermistor - it should survive elevated temperature easily.
Also you can check this. Diodes should also survive elevated temperature but I don’t know if long term stability would be good.

mramir: Well, pretty much everything is fake in this country.

Geez, that sounds risky, and the post complete with a photograph. I don't suppose Chinese face-recognition technology is held back by sanctions. You could have six ayatollahs on your doorstep tonight, coming to take you away, and your sanction-busting brewery as well.....

If your sensors are submerged, that's probably the problem. No such thing as waterproof, water resistant, maybe, but water proof in hot water...... Nahhh.

Also, as you have discovered, the microcontroller in the sensor must have the power off in order for it to reset after reaching 125C.

Paul

Smajdalf: Why? The DS18B20 is expensive chip. I think it is almost sure you will get a counterfeit if you buy it at less reputable seller. I would try some less advanced one, both LM75 and LM35 has max. temperature 150°C which may be true even for a generic copy. You can always test it in a cheap setup i.e. boil it in water for 3 hours and monitor if it works. Or a thermistor - it should survive elevated temperature easily. Also you can check this. Diodes should also survive elevated temperature but I don't know if long term stability would be good.

The article was very interesting, I will definitely try this approach, but I guess maybe it is better to use something with better A/D resolution, I mean better than 8-bits.