Reef-PI build thread, go big or go home

So it appears I had some corruption. I wasn't showing any 1 wire devices in my PI so I disabled 1 wire and rebooted then reenabled it and rebooted. Then I went back to Reef PI and reloaded it and bam temp sensors showing up again and functional. Thanks for all the help @Bigtrout .
 
Well I guess all is still not well, my temp sensors are again now showing up and my temp is reading 32 constant. Maybe I damaged the sensor itself?
 
temperature sub-system. Failed to read sensor Temp 1. Error:open /sys/bus/w1/devices/28-00000a98e329/w1_slave: no such file or directory
temperature sub-system. Failed to read sensor temp 2. Error:open /sys/bus/w1/devices/28-00000a9b6bb9/w1_slave: no such file or directory
These are the errors I'm getting in reef PI.
 
When they were working did they read correctly? Maybe u have a wiring issue, double check that.
 
They where reading correctly when they where working I even took my temperature and compared it to a hospital grade digital thermometer lol. I'm going to reset the database and see if that helps.
 
They where reading correctly when they where working I even took my temperature and compared it to a hospital grade digital thermometer lol. I'm going to reset the database and see if that helps.
Did some digging...seems intermittent 0c or 32F readings suggest an interference issue...make sure your temp sensor wires are not close to any 120v wiring...

If you still get sensors showing up with a serial number, I doubt that they are damaged or that the pi is damaged...what all did u change when u unplugged them...maybe an ac wire ends up too close to them...something to that effect maybe?
 
Well I found one wire on the connector that appeared to be a weak connection and soldered it again. Rebooting the PI now and waiting for it to load. Getting a long loading screen in the reef-pi though that may be because I reset the database for it.
 
Bleh now I can't even get to the reef pi screen on my desktop. So frustrating, just going to shut it down and reinstall everything on the PI in the morning including formatting the entire drive and reinstalling from scratch. I can't even begin to express how frustrating this is for me. Right now I'm seriously wondering just how many more bugs and problems I'm going to encounter once this is up and running on an actual tank.
 
Bleh now I can't even get to the reef pi screen on my desktop. So frustrating, just going to shut it down and reinstall everything on the PI in the morning including formatting the entire drive and reinstalling from scratch. I can't even begin to express how frustrating this is for me. Right now I'm seriously wondering just how many more bugs and problems I'm going to encounter once this is up and running on an actual tank.
I have had mine running for weeks and not a problem...its just a learning experience, dont get fruatrated. Just remember, dont hot plug things anymore.
Start with a clean card, reinstall raspian(lite) and reef pi and see if that does the trick.
 
Yeah I'm finishing up the noobs reinstall now, hoping this solves my issues.
 
This can not be a software issue. Particularly the ds18b20 is always a circuit issue. You can reset the database, but its highly unlikely to solve this. We know this can be frustrating, as the probe wire length, the voltage and wiring .. all introduces some chances of errors :-/

All of this makes me feel I or someone else should just set up a tindie (or whatever appropritae) store to sell these sensore pre-wired and tested.
 
Yup still not showing any temp sensors in my reef-pi. Made sure I had them connected while the PI was shutdown this time. So now I can be pretty sure it's not software related.
 
Ok @Ranjib if we can guarantee it's not software related, what steps should I pursue to troubleshoot the issue. I'm using
https://smile.amazon.com/gp/product/B01C3QTS5Q/ref=ppx_yo_dt_b_asin_title_o03__o00_s00?ie=UTF8&psc=1
Then I am connected directly to the DROK Temp Sensors which I no longer see on amazon they are the three meter version. On the female side of my connectors I've got less than a foot of wiring running from them directly to my hat.
What voltage you are using to power the ds18b20 probe, 3.3V or 5V ? what resistor value? If you are using 3.3V and 4.7K, then I;ll recommend trying 5V and 10K resistor just to test things out.

Check hardware.

1) Power down the system and check for continuity across the ds18b20 wires, across all the female audio connector ends (or whatever connector you are using)
2) Power up the system and check voltage reading across the female connector ends. both data and +ve ends.
3) Check `dmesg` log for any kernel error related to ds18b20 (just type dmesg in your terminal)
4) Is the probe getting hot? sometime the wiring is reveresed and it can really heat up the probe (other than the fact that it does not get detected). Be careful, i have burned my finger by this :-) , and I know another person in r2r also experienced the same.
 
Check continuity (with power down) between your permproto pins/holes to female connector ends. I didnt notice you have a 1 foot wire there, this is new . I have not seen that with any other build. I am suspicious about this (but not certain). You are using 3 meter long cable thats also a big error prone thing for 3.3 v based setup.
 
What voltage you are using to power the ds18b20 probe, 3.3V or 5V ? what resistor value? If you are using 3.3V and 4.7K, then I;ll recommend trying 5V and 10K resistor just to test things out.

Check hardware.

1) Power down the system and check for continuity across the ds18b20 wires, across all the female audio connector ends (or whatever connector you are using)
2) Power up the system and check voltage reading across the female connector ends. both data and +ve ends.
3) Check `dmesg` log for any kernel error related to ds18b20 (just type dmesg in your terminal)
4) Is the probe getting hot? sometime the wiring is reveresed and it can really heat up the probe (other than the fact that it does not get detected). Be careful, i have burned my finger by this :-) , and I know another person in r2r also experienced the same.

Using a 5V pull-up will subject the Pi’s pin to 5V - not suggested (it’s going to be shunted by the body diodes of the processor on the Pi and degrade them over time - they’re only there for ESD not continuous use). Powering the probe with 5V is ok however, just leave the pull-up at 3.3V.

I’d also suggest the opposite in resistor values - use lower (stronger pull-up) values. Electrically, you want to overcome the increased capacitance of the longer wire by speeding the rise time of the signal. Use a 2.2k, or just two 4.7k resistors in parallel.
 
I am using the @Michael Lane Hat which I believe uses 5v and a 10k resistor, I'll double check that however. The probes are not getting hot and are verified wired correctly as they where reading just fine before I unplugged them and plugged them back in. I find it highly unlikely the connectors on three separate probes would generate a fault in the wiring itself; however, I have pulled the connectors apart and checked all of the solder points for weakness. Worst case scenario is I blew something on the board when I did this, again though it seems unlikely that I would damage a transistor in this way.
 
I am using the @Michael Lane Hat which I believe uses 5v and a 10k resistor, I'll double check that however. The probes are not getting hot and are verified wired correctly as they where reading just fine before I unplugged them and plugged them back in. I find it highly unlikely the connectors on three separate probes would generate a fault in the wiring itself; however, I have pulled the connectors apart and checked all of the solder points for weakness. Worst case scenario is I blew something on the board when I did this, again though it seems unlikely that I would damage a transistor in this way.

Are the sensors wired in a star (all connected back at the Pi)? And there are three of them?

You’re most likely in trouble from wire length. I’ll look at the Hat schematics and suggest some other values. The pull up should not be at the 5V level though :(
 
Using a 5V pull-up will subject the Pi’s pin to 5V - not suggested (it’s going to be shunted by the body diodes of the processor on the Pi and degrade them over time - they’re only there for ESD not continuous use). Powering the probe with 5V is ok however, just leave the pull-up at 3.3V.

I’d also suggest the opposite in resistor values - use lower (stronger pull-up) values. Electrically, you want to overcome the increased capacitance of the longer wire by speeding the rise time of the signal. Use a 2.2k, or just two 4.7k resistors in parallel.
Take @theatrus 's suggestion here, he is more competent than me on these things :-)
 
I am using the @Michael Lane Hat which I believe uses 5v and a 10k resistor, I'll double check that however. The probes are not getting hot and are verified wired correctly as they where reading just fine before I unplugged them and plugged them back in. I find it highly unlikely the connectors on three separate probes would generate a fault in the wiring itself; however, I have pulled the connectors apart and checked all of the solder points for weakness. Worst case scenario is I blew something on the board when I did this, again though it seems unlikely that I would damage a transistor in this way.
Well if you are using Michael's hat then bulk of these suggestions wont apply, i guess
 

IF YOU HAD TO TAKE A REEFING EXAM, WOULD YOU PASS?

  • Yes!

    Votes: 32 45.7%
  • Not yet, but I have one that I want to buy in mind!

    Votes: 9 12.9%
  • No.

    Votes: 26 37.1%
  • Other (please explain).

    Votes: 3 4.3%
Back
Top