reef-pi :: An opensource reef tank controller based on Raspberry Pi.

[Ranjib]

Is there a specific interface needed or just a generic one wire probe?

reef-pi uses ds18b20 based probe with one wire interface

 

[Ranjib]

There was a picture of one on here. It was coated in black plastic.

Yeah, checkout the canakit or sparkfun ds18b20 probes

 

[bishoptf]

Yeah, checkout the canakit or sparkfun ds18b20 probes

Now that I looked, Amazon has lots of variations....If I need to extend the length I assume a normal audio extension will be fine, not sure how far they can drive, but will probably need to extend a couple of mine.

 

[Ranjib]

hi guys I started to build my reef-pi but I immediately ran into a problem I mounted an 8-relay card
I created the sockets taking care that the gpio pins were the correct ones I go into the equipment I create the physical grip and upon activation the input to the raspberry arrives (I see the green light flashing)
but the relays do not start ...
I saw, however, I miss the ULN2803A chip (already purchased waiting for me to arrive) because of this chip I can not activate the relay?
Thank you

Can you start a build thread ? We need lot more details including your relay type and circuit etc . It can be a lot of different things if you are not following the guides as it is

 

[Ranjib]

You weren't rude at all. I knew there was a specific use case involved here, just didn't know if the code would be able to run on other architectures. I do agree that it's good to do one thing very well instead of a lot of things "just ok"....

reef-pi binaries are compiled for arm 6/7 (pi zero/pi 3) hence they won’t work on intel/x86 architecture as it is. Although I can’t recall anything specific to arm, so it should be relatively easy to compile it for intel boards such as the one you suggested

 

[Ranjib]

Well been a little slow on the updates, I needed to rebuild my bench PC and had to do some major surgery on it, plus it was almost 60Deg yesterday and was really nice to be out. I soldered up everything last night, always takes me much longer to get things soldered but I think the PWM expansion board is ready to go, here are some pics. I started testing my small led and setting the frequency to 300, wow what a strobe effect. Funny but I am actually back to using 100, lol go figure but while ti doesnt seem to cycle, it produces a flicker free light about at the proper voltage.

I haven't tested with the big light yet but will be doing that soon. I did also manage to crash reef-pi a couple of times this morning when playing with the frequency setting and doing reloads a bunch of times. Not sure which logs you store info in, Ranjib if you can tell me where to look I am curious as to what issue it was having.

Can you elaborate what you observed when you say it crashed? reef-pi logs are available via

systemctl -u reef-pi.service

 

[Ranjib]

Can you elaborate what you observed when you say it crashed? reef-pi logs are available via

systemctl -u reef-pi.service

And you can use —since / —until to see logs from a specific time
https://www.loggly.com/ultimate-guide/using-journalctl/

 

[Ranjib]

Now that I looked, Amazon has lots of variations....If I need to extend the length I assume a normal audio extension will be fine, not sure how far they can drive, but will probably need to extend a couple of mine.

I have both steel and powder coated (canakit version) probes running for over two years now without any issue. My probes have 5 feet’s wires at max. I know @Erica-Renee had tried some longer length probe wires

 

[DirtDiggler2823]

Yeah, checkout the canakit or sparkfun ds18b20 probes

Sparkfun. That was the brand.

 

[bishoptf]

Can you elaborate what you observed when you say it crashed? reef-pi logs are available via

systemctl -u reef-pi.service

Pretty familiar with journalctl, not a systemd fan but it is what it is. I'm usually ssh'd into reef-pi when playing with stuff and have journalctl running to see the output. The first time it happened it didn't come back from a reload, hung for some reason, restarted the service and all was fine. The second time I think pi was actually core dumping although I haven't looked for a core, the hdd light went solid and from ssh console I had nothing. I did a cursory glance through journalctl but will look further later. I need to go do my weekly water changes on my tanks and do some cleanup.

 

[Erica-Renee]

I have both steel and powder coated (canakit version) probes running for over two years now without any issue. My probes have 5 feet’s wires at max. I know @Erica-Renee had tried some longer length probe wires

yes . I have reached 20 feet . 15 feet 15 feet and two 10 feet all at the same time.

You have to set up the Master (your 1-wire connection. and first slave (Probe furthest from master) then each of the others need to break that connection with the probe on the longest leg being slave 2 .. position . closest probe to the master should be the last probe you connect.. ..

You guys prob know more about the reasoning about this but i think it has to do with data crashes on the signal .. The master learns the distance of each probe based on the first slave being the furthest... Thou i might be wrong ... I have probe in my Display tank that is connected to the reef-pi way down in my basement .. eventually i am going to put the reef-pi computer stuff in my network closet with all the other electronics ..

 

[bishoptf]

yes . I have reached 20 feet . 15 feet 15 feet and two 10 feet all at the same time.

You have to set up the Master (your 1-wire connection. and first slave (Probe furthest from master) then each of the others need to break that connection with the probe on the longest leg being slave 2 .. position . closest probe to the master should be the last probe you connect.. ..

You guys prob know more about the reasoning about this but i think it has to do with data crashes on the signal .. The master learns the distance of each probe based on the first slave being the furthest... Thou i might be wrong ... I have probe in my Display tank that is connected to the reef-pi way down in my basement .. eventually i am going to put the reef-pi computer stuff in my network closet with all the other electronics ..

Question, I have three probes but did not see any where to set the master/slave relationship. How do you do that?

Thanks

 

[DirtDiggler2823]

Yeah, checkout the canakit or sparkfun ds18b20 probes

Probe isn't even registering at this point. I've already ordered the sparkfun probe, and a new 3.5mm jack for it.

 

[DirtDiggler2823]

yes . I have reached 20 feet . 15 feet 15 feet and two 10 feet all at the same time.

You have to set up the Master (your 1-wire connection. and first slave (Probe furthest from master) then each of the others need to break that connection with the probe on the longest leg being slave 2 .. position . closest probe to the master should be the last probe you connect.. ..

You guys prob know more about the reasoning about this but i think it has to do with data crashes on the signal .. The master learns the distance of each probe based on the first slave being the furthest... Thou i might be wrong ... I have probe in my Display tank that is connected to the reef-pi way down in my basement .. eventually i am going to put the reef-pi computer stuff in my network closet with all the other electronics ..

How do you tie multiple probes together?

 

[theatrus]

How do you tie multiple probes together?

Just attach more probes to the signal line in parallel. 1-wire allows multiple devices and each device (should) have a unique address set by the factory.

Bummer on the dead probe - that’s a weird failure. If you still have the dead one, can you cut the metal pill apart? I’m curious if the failure was water ingress.

 

[DirtDiggler2823]

Just attach more probes to the signal line in parallel. 1-wire allows multiple devices and each device (should) have a unique address set by the factory.

Bummer on the dead probe - that’s a weird failure. If you still have the dead one, can you cut the metal pill apart? I’m curious if the failure was water ingress.

I still have it. I might try to pull it apart to see what happened.

 

[Bigtrout]

I

I have both steel and powder coated (canakit version) probes running for over two years now without any issue. My probes have 5 feet’s wires at max. I know @Erica-Renee had tried some longer length probe wires

The ones I got, disappeared from amazon, but the wire is 3 meters long, so about 10 feet and they are working fine. They were drok branded, made in canada, with red high temp silicone insulation and even shielded. They are reading within .1F of each other and within .1F of my ITC1000 with an 10k ntc thermistor. I have a different 3 meter coming from drok but its not the same version this time but still has a 3 meter length and black insulation like the million others on amazon. I will report if its a decent option when it arrives.

 

[Erica-Renee]

Question, I have three probes but did not see any where to set the master/slave relationship. How do you do that?

Thanks

You don't . AHH.. I am not good at Explaining Technical stuff So here is where i got my Info for my testing .

Right from the source
https://www.maximintegrated.com/en/app-notes/index.mvp/id/148

 

[Diamond1]

Just for a reference for everyone I got a a ds18b20 from adafruit. It was the first probe I got and was coated in silicone before I put it into use. I started getting erratic readings in the reef-pi interface and took it out to check it. Apparently the silicone didn't do it's job very well because I noticed it had some rust on it. I just took it apart to check it to see if the sensor was intact and found this.
These are supposed to be waterproof and the only reason I used silicone was to keep the probe from rusting. After inspecting the sensor I found that whatever they used to seal the probe and keep it water tight had a big gap in it that allowed water to get into the probe and short it out.

 

[DirtDiggler2823]

Just for a reference for everyone I got a a ds18b20 from adafruit. It was the first probe I got and was coated in silicone before I put it into use. I started getting erratic readings in the reef-pi interface and took it out to check it. Apparently the silicone didn't do it's job very well because I noticed it had some rust on it. I just took it apart to check it to see if the sensor was intact and found this.
These are supposed to be waterproof and the only reason I used silicone was to keep the probe from rusting. After inspecting the sensor I found that whatever they used to seal the probe and keep it water tight had a big gap in it that allowed water to get into the probe and short it out.

I'll probably find something similar