Another Unique Automation Project - Aquarium Automation

chrishet

Member
I am looking to control my reef aquarium usung an ELK M1 system. As per my previous post I was looking for an inexpensive controllable powerstip, for pumps, lights...and I think I found a good solution...
http://www.digital-loggers.com/EPC.html I will connect to this device using an ELK M1XSP.

I have a couple of other design challenges, perhaps this group might have some solutions....

1) I need a good temp probe...one that can be used in water (salt water!)
2) It would be nice if I could also monitor the water pH values..there are rules that I could write that would adjust the waters pH values by activating certain pumps to dose the aquarium with the correct buffers...any ideas on how to interface a pH probe to an M1? Perhaps there are pools systems out there that are controlled by an M1. Not a huge deal as there are dedicated pH controllers out there but it would be nice to be able to control other rules that involve pH values. I did find this device..close but I don't think it will work as it is a parallel device...
http://www.drdaq.com/products.html


Thanks in advance!
 
I dont think you will get the ELK to do everything you want, so interfacing this with a PC and some automation software may be a soloution
 
I think I can get most of what I need just by writing rules to control the power strip and using zones for specific sensors. I have read the doc on most of the dedicated aquarium controllers out there and their automation rules are very simple. If I can't get a good water temp sensor then I will probably ditch the idea and use one of the dedicated controllers out there. Here is what I have so far:

AQRules.jpg
 
I dont think you will get the ELK to do everything you want, so interfacing this with a PC and some automation software may be a soloution

Todd may be right. You seem to have planned your rules carefully. But your matrix does not show whether you want your temperature and pH sensors to give you [a] continuous value readings, or threshold triggers.

If you use sensors that, on their own, give a simple on/off indication based on a set threshold level, and so provide a switch-type open/close signal to the M1 zone, then you are good.

On the other hand, if you expect to attach analog sensors directly to a zone and get useful readings, well... ...we'll all be "rooting for you."

Dave
 
One of the 'weaknesses' of the M1 is the lack of flexibility for temperature sensing. That said, I think if you just have 1 or 2 sensors you need it can still be done reasonably. Disclaimer: I have not used or tried these so I may be wrong but I believe the theory is correct. Personally, I am going to use a Datanab module with their sensors and connect it via CQC.

Ok, so the M1 can either use their own temp sensor or a 1-wire sensor via the M1KAM. The Elk temp module is available with a remote probe but idt the probe is waterproof or can handle the salt water. You might be able to get by bedding the remote probe in epoxy or something without affecting its sensitivity and still go that route. That would be the easiest. The other possible route I think may work is to use a sensor like this with the KAM. If that would work you could do 1 sensor per KAM. I'm pretty sure I read before that the Elk sensor can be bedded in something to make it waterproof, so that may be the best option for 1 sensor like this.

Otherwise, I would agree with previous post that perhaps you are better off with something like the datanab and having a variety of sensors available and just using it with other software. At some point you may want some supplemental software anyway. If, for example you went that route you could still trigger rules in the M1 from the datanab sensor. With CQC, since there is a datanab driver and M1 driver, CQC would get all the readings then you could create actions to do stuff in the M1. Granted you would not want to buy something like CQC just for that but if not perhaps you can get some of the free or cheap apps like J9 or vCrib or someone to create a datanab driver.
 
I dont think you will get the ELK to do everything you want, so interfacing this with a PC and some automation software may be a soloution

Todd may be right. You seem to have planned your rules carefully. But your matrix does not show whether you want your temperature and pH sensors to give you [a] continuous value readings, or threshold triggers.

If you use sensors that, on their own, give a simple on/off indication based on a set threshold level, and so provide a switch-type open/close signal to the M1 zone, then you are good.

On the other hand, if you expect to attach analog sensors directly to a zone and get useful readings, well... ...we'll all be "rooting for you."

Dave


Thanks Dave, from my research it sounds like the m1 does not report temps very accurately so that may be a challenge. My rules will be based on thresholds, i.e. whenever temp >80 then turn off heater...etc...I have written rules using the ELK temp sensors but never really payed that close attention to the accuracy of the module...I am not even sure you can calibrate the ELK sensors. I see that HAI makes a water temp sensor
http://www.automatedoutlet.com/product.php...at=0&page=2

not sure it will work with the m1 though...
 
My rules will be based on thresholds, i.e. whenever temp >80 then turn off heater...etc...I have written rules using the ELK temp sensors but never really payed that close attention to the accuracy of the module...I am not even sure you can calibrate the ELK sensors.

I don't know about the pure accuracy of the Elk temperature sensors. I have one, in the attic, that is satisfactory, at least.

Anyway, it's not necessarily a question of calibration since you can deal with that in other ways. You can just compare the Elk reading with your calibrated thermometers, right? The challenge, I guess, would be the waterproofing. If you use one of the Elk probes you can perhaps pot it, or seal it up in some other way, or just (gasp :rolleyes: ) dunk it in the water?

So while you can in principle get your temperature measurement using an Elk-made temperature sensor, once you beat the waterproofing, pH measurement would be more involved. You may or may not know that the Elk temperature sensors communicate with the M1 in an odd sort of way, using a serial communications protocol; they do not use the zone in a conventional manner.

As Steve said, Elk is missing an important Home Automation piece: some sort of "analog attachment base" expander for the M1, connecting to the RS-485 bus and providing a set of terminals for various analog inputs. Accepting continuous voltage devices, current loops, or wide resistance ranges would be a valuable addition.
 
I would also consider the Ocelot. It supports temp probes, supports complex rules, and ascii strings. You can have it talk to the Elk M1 using an M1XSP (that's how I do IR with the M1).
 
One of the 'weaknesses' of the M1 is the lack of flexibility for temperature sensing. That said, I think if you just have 1 or 2 sensors you need it can still be done reasonably. Disclaimer: I have not used or tried these so I may be wrong but I believe the theory is correct. Personally, I am going to use a Datanab module with their sensors and connect it via CQC.

Ok, so the M1 can either use their own temp sensor or a 1-wire sensor via the M1KAM. The Elk temp module is available with a remote probe but idt the probe is waterproof or can handle the salt water. You might be able to get by bedding the remote probe in epoxy or something without affecting its sensitivity and still go that route. That would be the easiest. The other possible route I think may work is to use a sensor like this with the KAM. If that would work you could do 1 sensor per KAM. I'm pretty sure I read before that the Elk sensor can be bedded in something to make it waterproof, so that may be the best option for 1 sensor like this.

Otherwise, I would agree with previous post that perhaps you are better off with something like the datanab and having a variety of sensors available and just using it with other software. At some point you may want some supplemental software anyway. If, for example you went that route you could still trigger rules in the M1 from the datanab sensor. With CQC, since there is a datanab driver and M1 driver, CQC would get all the readings then you could create actions to do stuff in the M1. Granted you would not want to buy something like CQC just for that but if not perhaps you can get some of the free or cheap apps like J9 or vCrib or someone to create a datanab driver.

I had an interesting conversation with the owner of a company that specializes in large aquarium automation, and he was familiar with the m1 gold. He thought it would be possible to integrate a lab quality temp and pH probe to the m1 but cost will be an issue, as would reporting sensitivity. The m1 temp zones report using voltages and it is these voltages are calibrated as temps that are displayed and used in rules. The problem is that the m1 does not report an accurate enough temp to be used for my needs as a pH probe (i.e. down to the tenth of a degree). Here is a device that he suggested that would take and convert probe inputs to specific output voltages for the m1 zones.

http://www.automatedaquariums.com/7011.htm

Bottom line..this is looking too complex for the m1 to handle on its own...so I will be looking into integration with PC tools or most likely one of the standalone aquarium automation devices.

Thanks for all the help!
 
The m1 temp zones report using voltages and it is these voltages are calibrated as temps that are displayed and used in rules.

If you mean the connection between an M1ZTS and an M1 zone, you are absolutely, totally, completely wrong.

Here is a device that he suggested that would take and convert probe inputs to specific output voltages for the m1 zones.

Too bad you are not taking this path, because I would like to know how such an interface would work.
 
The m1 temp zones report using voltages and it is these voltages are calibrated as temps that are displayed and used in rules.

If you mean the connection between an M1ZTS and an M1 zone, you are absolutely, totally, completely wrong.

Here is a device that he suggested that would take and convert probe inputs to specific output voltages for the m1 zones.

Too bad you are not taking this path, because I would like to know how such an interface would work.

It does sound interesting to take on...but expensive! $500 for a temp/pH solution is too much $ for my budget....

on the ZTS, I was taking my assumptions from the engineer at Aquatics Automation...you are correct, the temp sensor sends the zone a binary pulse to the zone, not variable voltages...Thx for correcting me....

I just got off the phone with AO and ELK and they seem to think the remote temp sensor may work for my use, I need to see if the metal is compatible with my salt water tank.
http://www.automatedoutlet.com/product.php...=335&page=1
 
the temp sensor sends a binary pulse to the zone, not variable voltages

Yah, Spanky explained that it uses Manchester encoding at 1 bit per second.

Good luck with this interesting project Chris. We are interested to know your result. Photos would be nice.

Dave
 
Back
Top