Not all DS18B20 are Created Equal

[Dewd]

I have up until now used the generic DS18B20 and they have always had varying temp reading.  Like these https://www.banggood.com/Wholesale-DS18B...rehouse=CN  

But I've never been able to get perfectly stable readings.  Then recently I have had one go bad and it would lock the I2C bus and I would lose all my temp sensors.

So I decided to replace them all (3). I looked around and eventually broke down and decided to buy the Drok units.  I have read some good reports on their quality.  So I bought 2 of these

https://www.amazon.com/gp/product/B01MY8...eter+Probe

and they arrived two days ago.  Ad expected, they look like much better units.  Silicone cable and properly double crimped tip.  I added an additional waterproofing layer and then connected them up.

and

perfectly stable temp readings on the graphs - Very happy with the result.

[Rob F]

Those do look nice, I changed the .com to .ca and $40 here, I guess I won't be offering them but good to know they are available. I've never experienced unstable readings and don't think I've really heard that issue, how unstable were they?

A bad sensor can hang the 1-wire bus (I2C is pH and same thing can happen to I2C bus), you can configure the Pi for multiple buses which would prevent that. If you plug in your SD card to computer you can open the config.txt file and edit it.

You'll see the following line. This enables the 1-wire bus using the default GPIO pin 4 on the Pi.

       dtoverlay=w1-gpio

If you change above line to the first line below it will keep GPIO 4 as is and add a 2nd, 3rd or more lines for extra buses. I've tested 3 buses but should work for most GPIO pins. 

       dtoverlay=w1-gpio,gpiopin=4
       dtoverlay=w1-gpio,gpiopin=17

And as a bonus the Pi reads all the buses and puts the files needed for reef-pi in the same location so reef-pi will look the same if you have 3 sensors running on 3 buses.

I know you don't have controller yet but the 3 temp sensor ports do share the same bus due to shortage of GPIO pins but it's possible to use any of the 6 sensor ports for extras.

[Dewd]

Those do look nice, I changed the .com to .ca and $40 here, I guess I won't be offering them but good to know they are available. I've never experienced unstable readings and don't think I've really heard that issue, how unstable were they?

They would always vary up and down by as much as 0.5 degC.  Never had a stable reading over time.  Graphs were never a smooth line like now.  I'll see if I can find a screenshot of it somewhere.

[pwest]

This is very interesting.  As documented in this thread: https://www.robo-tank.ca/forum/Thread-Un...0-Readings I have been unable to use the standard DS18b20 cables on my system, presumably due to EMI caused by my LED lighting (probe typically works when lights are off, then fails after lights-on).  In hopes that these cables might somehow be better, I ordered a couple.  Right out of the bag, I was impressed w/two things: their apparent quality and the presence of a braided shield.

I wired up the probe, connected the shield to a ground pin, put some heatshrink and RTV around the metal probe and it's been running for 2 days so far.

-Phil

[Rob F]

Hi Phil, thanks for sharing, those definitely do look nice, glad it's been working out. Did you get them on Amazon? I see that listing is out of stock and couldn't find any others.

[pwest]

I ordered these:
https://www.droking.com/2-PCS-LOT-DS18B2...ption=true
And they were delivered in less than two weeks.

[Rob F]

Nice, thanks, I might have to start getting some of these as the price is much better than I expected. I wonder if the sale is really a sale or always running.

[pwest]

The new sensor has now been running for over a week w/no evident issues. It's quite nice to have a reliable temperature probe in the tank again!

-Phil

[Rob F]

That's great news, I'm glad you were able to find one.