JCA34F:
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.
I agree.
Nick_Pyner:
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.....
No worries, I am not the only one doing this. In fact, as long as you have not added yeast to the wort, you have not done anything wrong. Like many other countries, you can buy beer kits, malt, hops, etc. etc here. Just the yeast is missing. Also, thanks to corona, government announced that alcohol production in Iran does not require any license or permission until further notice. Even corona has a bright side
I really enjoyed reading your last post!
You seem like someone who likes to experiment. Therefore, consider making your own temperature sensor by sealing a regular silicon diode inside a glass tube. They have always been used to make temperature sensors. The back bias current flow will increase with temperature. I have not looked, but Google should be able to find more information on "diode temperature sensor".
Paul
Diode temperature sensors are forward biased, usually with a constant current source, since the
forward voltage depends on temperature through the diode equation - however it takes careful
analog processing to read this small voltage accurately enough for good temperature resolution.
The DS18B20 wraps this up for you and does the hard task of converting and digitizing the
small voltages involved, and comes pre-calibrated. The laser trimming for calibration and
writing the unique ID is what makes it more expensive than many chips, as the device has to be measured
at several different temperatures, then programmed.
You can pretty much guarantee a DS18B20 from a non-reputable source is fake these days, the mark
up for the fruadsters is just too tempting.
Arduino ADC is 10 bit. With 1 V reference it is about 1 mV/LSB. Tempco of diode forward drop is said to be 1-2 mV/°C, resolution with Arduino ADC is about 1°C. I think it should be enough. But is the precision good enough?
So it turns out that my fake DS18B20s have interesting behavior. The probe handles up to 65 C for more than 2 hours (at least and as long as I did monitor it) no problem and behaves normal, when I go above 65 (near 67) it starts to "blink" between -127 and the correct temp. If I keep it there, it eventually stops working (maybe after 15-20 mins). If I remove the probe and let it cool down for a few minutes it comes back to life and again remains functional under 65 C and it works perfectly fine and accurate. So, apparently these fake ICs are fine for applications where temperature fluctuations remain within the 50% of the nominal max and min values.
Then you likely have genuine factory made devices, but they are from the batches that could not pass the test for quality.
Pau
Honestly I am not sure, I didn't bother checking the ROM and registers of the probe cuz I was brewing last night. And I managed to getaway with this faulty probe.
In Iran electronic parts has it's own "bazar" like pretty much any other thing. In downtown Tehran there is a very long street and is the bazar of electronic parts and musical instruments. You can go and check all of those 100s of shops one by one, show them the part number you are looking for and hope to get what you want. And often, if the 10th shop you checked didn't have the part you want then you probably are not going to find it, or they all together have the same part form the same factory or mass supplier. So, I guess here there are a few mass importers that import everything, from ICs to cellphones, fabric, tires, i dont know you name it, and distribute it in the market. So often you have to wait 4,5, 6 month for the next imported batch to arrive.
But how this "fake IC industry" works? I guess counterfeiters actually buy bare dies from the "original" mass producers and then package them. If so then does packaging matter so much?!
I tried to explain that they are not counterfeit devices, but devices the failed the factory tests.
Paul
Smajdalf:
Arduino ADC is 10 bit. With 1 V reference it is about 1 mV/LSB. Tempco of diode forward drop is said to be 1-2 mV/°C, resolution with Arduino ADC is about 1°C. I think it should be enough. But is the precision good enough?
it is not a huge deal for a home brewer. but still it is better to have the best you can afford and I have a Bluepill gathering dust in my drawer. So, it is time to put this baby into some action.
Paul_KD7HB:
I tried to explain that they are not counterfeit devices, but devices the failed the factory tests.Paul
Yeah I got your point and that is also highly likely. It is very possible that they go and buy QC-failed parts for probably one-third of the real price. Or the Chinese guy ceases this opportunity to get rid of their garbage.
You know, 98% of Iranians suffer from sanctions but that other 2% benefit from sanctions and I mean millions of dollars for each of these vultures.
If you have a logic analyzer or a scope you can try to look at the communication. Maybe the chip is working but "out of specs". Too slow or too fast communication? Or it is unable to pull the line low enough or leaking too much current (you can try if adding stronger/weaker pull-up helps).
I don't know Arduino one-wire library well enough to guess where a problem may be. But one-wire is complicated timing and noise sensitive protocol. There may be the problem and if so there may be possibility to solve it.