Reef-Pi Controllers: New all-in-one with power bar, and the original Pico

[Bigtrout]

T

Check your ds18b20 wires and connectors closely. Several people have had issues and almost always root cause was a power or data line shorting to ground. Those are small wires. With a full circle of ground around it.

These base boards do make a pretty setup! Did you get the case that fits it as well?

This is what I like about reef-pi! A few short months ago @Mikeneedsahobby was starting on his build, not familiar with anything other than knowing that trolls ran his controller, and now he's able to help other people with troubleshooting. You learn quite a bit with a diy reef pi build, and the community as a whole is a great resource!

 

[Ranjib]

T

This is what I like about reef-pi! A few short months ago @Mikeneedsahobby was starting on his build, not familiar with anything other than knowing that trolls ran his controller, and now he's able to help other people with troubleshooting. You learn quite a bit with a diy reef pi build, and the community as a whole is a great resource!

When I started three years back I didnt knew anything . i recall being so stoked to see first LED blinking with code.

 

[theatrus]

Thanks for doing a great job with these boards @theatrus - really happy with how much neater (and more reliable) this setup is compared to my old homebrew Particle photon/breadboard rats nest.

Just a quick with a few observations (from using a Rev. D):

* It might be worth mentioning that the config.txt dtoverlay lines remap GPIO pin numbers in the readme? Based on the circuit diagrams for the board I wasted a little bit of time trying to figure out why setting up reef-pi jack with pins 13 and 18 wasn't working before realising they'd been remapped to 0 and 1 in config.txt.

* I'm seeing a fairly significant number of incorrect DS18B20 readings (dozens per day when sampling each minute). I didn't see these previously when the same sensor was wired up to a particle photon. Being honest I've not done any investigation into this yet (doesn't bother me hugely as I've got nothing automated on temperature) but thought it might be worth mentioning in case others see the same. I'll try to get hold of a new sensor at some point and look into it a bit more.

* The Jebao dosing pump slave units (e.g. the DP-4S) are easy to integrate with the pico for a cheap and cheerful dosing setup. The Jebao units have a simple RJ11 connection on the back with 12V on pin1 and negative side for each of the heads on pins 2,3,4,5 so simple to wire up to the Pico's DB9 connector. Just simple on/off control rather than PWM but it's accurate enough for my requirements and actually fairly neat setup (only a single cable ends up running to the dosing unit).

Chris.

That’s a pretty good tip on the dosing pump addon!

As for the sensors, wiring and water ingress have been the two main culprits.

I’ve had some probes that were just erratic out of the box and some fine, even before any water was introduced to them. If you do get to debug, simply trying another is the best workaround. In order to protect the Pi GPIOs, we also run them at 3.3V - some clones may not be tolerant of this.

 

[Bigtrout]

That’s a pretty good tip on the dosing pump addon!

As for the sensors, wiring and water ingress have been the two main culprits.

I’ve had some probes that were just erratic out of the box and some fine, even before any water was introduced to them. If you do get to debug, simply trying another is the best workaround. In order to protect the Pi GPIOs, we also run them at 3.3V - some clones may not be tolerant of this.

The quality of temp probes and the lm2596 buckpucks is very variable, but reading on electronic forums, seems that some of the batches of both these use copies instead of real chips, wonder if that is causing some the problems that we see.

 

[theatrus]

The quality of temp probes and the lm2596 buckpucks is very variable, but reading on electronic forums, seems that some of the batches of both these use copies instead of real chips, wonder if that is causing some the problems that we see.

Both are very commonly cloned parts, and there are a lot of possible clones variants. The clones are a lot cheaper even in the top-end side (e.g. https://lcsc.com/product-detail/Others_HGSEMI-LM2595S-5-0-TR_C316687.html) compared to retail from the real TI part. Combined with zero QA and a bunch of leftover scrap passives who knows.

 

[Bigtrout]

Thats why use meanwells:

Meanwell PS=reliable, UL listed, warranty

No name buckpuck=crapshoot and bad ones have been known to get hot enough to cause fires, and can be noisy because passives used dont always match the design.

A 3 amp 5 volt meanwell is all of 8 dollars, makes the choice pretty easy.

 

[theatrus]

I've completed the aforementioned e-mail blast - if you haven't gotten it, let me know via a DM and I can provide the information.

 

[Ranjib]

Yikes I posted and it asked me to sign in and it apparently posted 3 times! Sorry, hope one of you can clean that up.

@robottaway your issue is sorted now?

 

[Ranjib]

Thanks for doing a great job with these boards @theatrus - really happy with how much neater (and more reliable) this setup is compared to my old homebrew Particle photon/breadboard rats nest.

Just a quick with a few observations (from using a Rev. D):

* It might be worth mentioning that the config.txt dtoverlay lines remap GPIO pin numbers in the readme? Based on the circuit diagrams for the board I wasted a little bit of time trying to figure out why setting up reef-pi jack with pins 13 and 18 wasn't working before realising they'd been remapped to 0 and 1 in config.txt.

* I'm seeing a fairly significant number of incorrect DS18B20 readings (dozens per day when sampling each minute). I didn't see these previously when the same sensor was wired up to a particle photon. Being honest I've not done any investigation into this yet (doesn't bother me hugely as I've got nothing automated on temperature) but thought it might be worth mentioning in case others see the same. I'll try to get hold of a new sensor at some point and look into it a bit more.

* The Jebao dosing pump slave units (e.g. the DP-4S) are easy to integrate with the pico for a cheap and cheerful dosing setup. The Jebao units have a simple RJ11 connection on the back with 12V on pin1 and negative side for each of the heads on pins 2,3,4,5 so simple to wire up to the Pico's DB9 connector. Just simple on/off control rather than PWM but it's accurate enough for my requirements and actually fairly neat setup (only a single cable ends up running to the dosing unit).

Chris.

Welcome to reef2reef Chris. I love the jebao dosing pump tip. Do you have this documented anywhere?

 

[Mikeneedsahobby]

I've completed the aforementioned e-mail blast - if you haven't gotten it, let me know via a DM and I can provide the information.

Got mine!

This service level is beyond what I or my trolls deserve.

 

[robottaway]

@robottaway your issue is sorted now?

Replacement should be here tomorrow. Will swap, test and report back.

 

[theatrus]

Hi!

I'm interested in some feedback for those of you who do use the Pico board, or have looked at it so far. I'd like to make revisions based on what people want, and this is where you come in!

Drop some information in the form below:

https://forms.gle/oUrWWE2PzrAZHEBn9

 

[rushbattle]

Hi!

I'm interested in some feedback for those of you who do use the Pico board, or have looked at it so far. I'd like to make revisions based on what people want, and this is where you come in!

Drop some information in the form below:

https://forms.gle/oUrWWE2PzrAZHEBn9

Done, thanks for all of your hard work and therefore great products!

 

[theatrus]

Done, thanks for all of your hard work and therefore great products!

Thanks for the kind words, and all the responses so far.

I'm somewhat surprised at the number of people using the I2C port - I assume its for external PCA9685?

 

[DirtDiggler2823]

Thanks for the kind words, and all the responses so far.

I'm somewhat surprised at the number of people using the I2C port - I assume its for external PCA9685?

Was the pin labeling addressed in the new boards?

 

[theatrus]

Was the pin labeling addressed in the new boards?

Board silk screen is identical, so still reversed. This is correcting the I2C and SMC firmware only.

 

[DirtDiggler2823]

Board silk screen is identical, so still reversed. This is correcting the I2C and SMC firmware only.

Good to know. I'd already accounted for the reversal, and needed to know if it was still present. Thanks!

 

[DirtDiggler2823]

Board silk screen is identical, so still reversed. This is correcting the I2C and SMC firmware only.

I assume you want the "old" boards back, yes?

 

[dmolavi]

I assume you want the "old" boards back, yes?

Yes. That was in the email, I think.

 

[DirtDiggler2823]

Yes. That was in the email, I think.

Awesome.