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

[Sral]

@PlantedAquaChicago
Alright, that sounds to me like there is a potential difference between the tank and your circuit's ground. Might either be from the tank OR from your circuit, since voltages are relative potential differences.

As soon as you connect your tank to your circuit's Ground with a low resistance wire the potential difference is drastically decreased, as evident by the pH measurement accuracy. That means that whatever is causing the potential difference can't supply a lot of current, as per usual for these kind of disturbances.
The residual potential difference that remains is whatever current the disturbance source can deliver times the (very) low resistance of the connecting wire. Or in other words: the internal series resistance of the disturbance source is much higher than your wire's resistance, so the voltage in that circuit drops mainly over the disturbance source, not the wire between the circuit and the tank.

Try for giggles to measure the voltage between the circuit's ground and your water with a multimeter. Would be funny if it is about the 1 V I calculated earlier ^^
If that is so, ground your tank as earlier with a wire, but put a resistor in series. Measure the voltage again. If you can find a resistor that drops the voltage about half, that's the internal resistance of the disturbance source.

My solution, if you don't want a ground wire in your tank, try to ground your circuit instead. If the circuit's ground is floating, as you mentioned earlier, it's something you might want to do anyway. I would recommend to connect your circuit's Ground to your AC socket's earth (the large earthing rod for US plugs, not the shorter Poles !) over something like a 1k resistor. The value should be much lower than the disturbance source's internal resistance I mentioned earlier, but high enough to limit power flow to something non-destructive.

One thing however: the whole thing sounds like your pH circuit is not properly isolated. If it were it shouldn't be disturbed that much by this potential difference in my oppinion. I am however not a professional, so I might be wrong on that ^^

 

[Sral]

I'm not sure what you mean...
I'm using a pH module from diy more, a cheap probe with no name and a ads1115 analogue to digital converter.
What should be galvanic isolated?

I'm quick fixing it atm with a macro that fill a small 3d printer camber then measure pH
But after i had to add a "wait" in the macro it won't trigger any macro

BTW, he meant that your pH module should be galvanicaly isolated from the rest. In your case, the ADS1115. This can be done by something like this.

If you don't do something like this you typically end up with readings like @PlantedAquaChicago

 

[Matt Carden]

1st install says wrong credentials. Any way to fix/change log in credentials?

 

[PlantedAquaChicago]

@PlantedAquaChicago
Alright, that sounds to me like there is a potential difference between the tank and your circuit's ground. Might either be from the tank OR from your circuit, since voltages are relative potential differences.

As soon as you connect your tank to your circuit's Ground with a low resistance wire the potential difference is drastically decreased, as evident by the pH measurement accuracy. That means that whatever is causing the potential difference can't supply a lot of current, as per usual for these kind of disturbances.
The residual potential difference that remains is whatever current the disturbance source can deliver times the (very) low resistance of the connecting wire. Or in other words: the internal series resistance of the disturbance source is much higher than your wire's resistance, so the voltage in that circuit drops mainly over the disturbance source, not the wire between the circuit and the tank.

Try for giggles to measure the voltage between the circuit's ground and your water with a multimeter. Would be funny if it is about the 1 V I calculated earlier ^^
If that is so, ground your tank as earlier with a wire, but put a resistor in series. Measure the voltage again. If you can find a resistor that drops the voltage about half, that's the internal resistance of the disturbance source.

My solution, if you don't want a ground wire in your tank, try to ground your circuit instead. If the circuit's ground is floating, as you mentioned earlier, it's something you might want to do anyway. I would recommend to connect your circuit's Ground to your AC socket's earth (the large earthing rod for US plugs, not the shorter Poles !) over something like a 1k resistor. The value should be much lower than the disturbance source's internal resistance I mentioned earlier, but high enough to limit power flow to something non-destructive.

One thing however: the whole thing sounds like your pH circuit is not properly isolated. If it were it shouldn't be disturbed that much by this potential difference in my oppinion. I am however not a professional, so I might be wrong on that ^^

I will try all of that. The logic seems sound for sure.

Thanks!

 

[JachPot]

Ok, I'm about to jump in the deep end. Is Robo Tank still the best way to just purchase a premade Reef Pi setup?

 

[bishoptf]

Ok, I'm about to jump in the deep end. Is Robo Tank still the best way to just purchase a premade Reef Pi setup?

That is a good option there are others like @Schreiber Leviathon offers, and I am sure there are some other stuff out there. Hit me up via PM and I can give you some other things to think about etc...

 

[trustychords]

Ok, I'm about to jump in the deep end. Is Robo Tank still the best way to just purchase a premade Reef Pi setup?

Yes, absolutely. RoboTank makes setting up Reef-Pi hardware so much easier. It is basically plug and play. I am not sure I would have continued with Reef-Pi if it wasn't for RoboTank.

Plus, I like supporting a "startup" business---great for the Reef-Pi ecosystem!

 

[robsworld78]

Yes, absolutely. RoboTank makes setting up Reef-Pi hardware so much easier. It is basically plug and play. I am not sure I would have continued with Reef-Pi if it wasn't for RoboTank.

Plus, I like supporting a "startup" business---great for the Reef-Pi ecosystem!

Right on, glad you're enjoying it and thanks for the kind words and support.

 

[DWKDiver]

That is a good option there are others like @Schreiber Leviathon offers, and I am sure there are some other stuff out there. Hit me up via PM and I can give you some other things to think about etc...

I have Rob's board and it is great. Get the latest full blown version, better than the earlier version I have. I was 99% complete with my 2nd diy controller and saw what he offered and haven't regretted a second. Reef pi updates seem to be far in-between these days and 5.2 is last version I can use due to later version not working with 12v dc dosers and a few other issues for awhile now. Autism on fixes myself.....

 

[elysics]

after having my last azdelivery ds18b20 fail on me, what's next? If i get a genuine maxim chip, or the wired ones from sparkfun, for a several times higher price than the other ones, can i expect it to run for years and years and years (properly sealed of course) or will they fail too after a while?

 

[elysics]

after having my last azdelivery ds18b20 fail on me, what's next? If i get a genuine maxim chip, or the wired ones from sparkfun, for a several times higher price than the other ones, can i expect it to run for years and years and years (properly sealed of course) or will they fail too after a while?

disregard that post, chip is still fine, the wire rusted inside the ****** white glue they put in there... I thought leaving it in and sealing it with silicone was enough, apparently not, this time i'll clean it up and pot it properly

 

[Sral]

disregard that post, chip is still fine, the wire rusted inside the ****** white glue they put in there... I thought leaving it in and sealing it with silicone was enough, apparently not, this time i'll clean it up and pot it properly

Interesting info, can you send pictures of the damage and how you fixed it ?

I heard that the DS18B20 are prone to fail but never knew why. Thanks a lot in advance !

 

[elysics]

Interesting info, can you send pictures of the damage and how you fixed it ?

I heard that the DS18B20 are prone to fail but never knew why. Thanks a lot in advance !

Didn't take pictures at the time sadly but the VCC line was completely rusted through. The others were spotless. The initial failure mode was that i was getting temperature spikes that started at 39-40°C and then shot up to 85°C, sometimes they stayed there until i wiggled the cable, i thought my connector was bad, sometimes it went away on it's own. Eventually, it was constantly reporting 85°C and no amount of wiggling changed that.

Here are some of the remains of the white glue, you can still see traces of rusted copper.

What i did to fix it was completely remove that white stuff, shorten the cable and to-92 legs until i was left with shiny metal, and soldered them back together. Heatshrinked the individual lines and another heatshrink around the outer cable isolation the exposed lines and half of the to-92 package. Then i potted the whole thing in a plasticstraw shroud with some hot melt, temporarily until i get my hands on some silicone.

If you are going to buy new ones, buy the to92 chips and do the cabling and potting yourself from the start, much cheaper and less ******.

 

[DWKDiver]

Didn't take pictures at the time sadly but the VCC line was completely rusted through. The others were spotless. The initial failure mode was that i was getting temperature spikes that started at 39-40°C and then shot up to 85°C, sometimes they stayed there until i wiggled the cable, i thought my connector was bad, sometimes it went away on it's own. Eventually, it was constantly reporting 85°C and no amount of wiggling changed that.

Here are some of the remains of the white glue, you can still see traces of rusted copper.

What i did to fix it was completely remove that white stuff, shorten the cable and to-92 legs until i was left with shiny metal, and soldered them back together. Heatshrinked the individual lines and another heatshrink around the outer cable isolation the exposed lines and half of the to-92 package. Then i potted the whole thing in a plasticstraw shroud with some hot melt, temporarily until i get my hands on some silicone.

If you are going to buy new ones, buy the to92 chips and do the cabling and potting yourself from the start, much cheaper and less ******.

I sealed mine years ago before using. Used heat shrink tubing on top of rubberized under coating paint at the top of metal piece. It seals and STIFFENS the wire so it doesn't bend easily, which I believe helps even more. ALWAYS use an alarm for water temperate just in case. I also have a backup heater in the tank NOT on controller, just set the ON temp a little below set temp of controller heater. All this from experience!

 

[DWKDiver]

I have Rob's board and it is great. Get the latest full blown version, better than the earlier version I have. I was 99% complete with my 2nd diy controller and saw what he offered and haven't regretted a second. Reef pi updates seem to be far in-between these days and 5.2 is last version I can use due to later version not working with 12v dc dosers and a few other issues for awhile now. Autism on fixes myself.....

Correction I use 5.1. 5.2 won't run dc dosers at all for me.

 

[elysics]

Correction I use 5.1. 5.2 won't run dc dosers at all for me.

Yep, same here

 

[elysics]

I sealed mine years ago before using. Used heat shrink tubing on top of rubberized under coating paint at the top of metal piece. It seals and STIFFENS the wire so it doesn't bend easily, which I believe helps even more. ALWAYS use an alarm for water temperate just in case. I also have a backup heater in the tank NOT on controller, just set the ON temp a little below set temp of controller heater. All this from experience!

Yeah, i never had real problems, i just got annoying email spam and it messed up my graphs because the chart minimum and chart maximum settings don't work properly. But i also never let the room cool out completely in the winter, i have unheated ambient temperature tanks and stuff standing around.

 

[bishoptf]

Didn't take pictures at the time sadly but the VCC line was completely rusted through. The others were spotless. The initial failure mode was that i was getting temperature spikes that started at 39-40°C and then shot up to 85°C, sometimes they stayed there until i wiggled the cable, i thought my connector was bad, sometimes it went away on it's own. Eventually, it was constantly reporting 85°C and no amount of wiggling changed that.

Here are some of the remains of the white glue, you can still see traces of rusted copper.

What i did to fix it was completely remove that white stuff, shorten the cable and to-92 legs until i was left with shiny metal, and soldered them back together. Heatshrinked the individual lines and another heatshrink around the outer cable isolation the exposed lines and half of the to-92 package. Then i potted the whole thing in a plasticstraw shroud with some hot melt, temporarily until i get my hands on some silicone.

If you are going to buy new ones, buy the to92 chips and do the cabling and potting yourself from the start, much cheaper and less ******.

These are the shizzle in regards to temp probes if you want the best @theatrus has the best ones you can purchase and they are real DS18B20 - https://blueacro.com/collections/se...or-reefvolt-and-reefpi?variant=32680136835153

 

[AquaKey]

@Ranjib @robsworld78 - trying to get the lighting setup with reef pi and robo tank but running into an issue. Tested with my multimeter and looks like reef pi is only changing the voltage on pin 8 (ch1) all other pins/ch only output 10v no matter what I change in reef pi. Any thoughts?

 

[robsworld78]

@Ranjib @robsworld78 - trying to get the lighting setup with reef pi and robo tank but running into an issue. Tested with my multimeter and looks like reef pi is only changing the voltage on pin 8 (ch1) all other pins/ch only output 10v no matter what I change in reef pi. Any thoughts?

What controller do you have and which reef-pi version are you running? I wonder if the connectors aren't setup correctly. Here's the link for the manuals, find the one for your controller, then in the manual go to the PWM ports page and at the bottom will be the pin numbers. For what it's worth also go to the admin tab in reef-pi and press the reload button, maybe that will sort it out.