Sensors & Controllers - Maybe DIY?

[drcrook]

Hey Folks,

So I'll start this off with I'm an engineer by trade and I build software for industrial controllers; so the programming, wiring and fabrication stuff is perfectly fine for me.

The main driver to this post is that I want to understand the parameters of my tank and preferably continuously in a digital format such that I can adjust my water parameters as needed.

I'm seeing the Neptune includes most of the sensors I think I should be interested in for $800; which sounded very high to me. I can purchase temperature sensors for $1 and build a housing fairly inexpensive too.

The salinity/conductivity, ORP and pH sensors from my vendor are a whopping $150-$200 each (call it $175). So that's $525. My controller, interface and housings would probably be another $150. So I'm figuring its roughly $675 to build my own or I could possibly just purchase for $125 more.

Questions.

1. Do you really need all of those sensors or which ones are most critical? I'm thinking Salinity, Temperature and pH. I'm not sure about ORP.
2. Has anybody built one before and if so; what sensors/vendors did they use? I use Atlas Scientific, albeit they are pricey.
3. This is pretty high for monitoring. How often do folks monitor without these solutions and what is the cost of that typically?
4. Are there additional benefits to the Apex?
5. Are there alternatives to the Apex primarily for water quality monitoring?
6. Are there sensors for Magnesium and Calcium?

 

[chipmunkofdoom2]

@Ranjib has done some really exceptional work on a Raspberry Pi based controller system called Reef-Pi. When I need a new controller, I'm going this route. There's no way I can justify the $800 for an Apex. I know about what relays cost and how easy it is to program a controller. It's not worth $800.

For which sensors are required, I would at least do pH so you can control dosers based on pH and temperature so you can control your heater. I wouldn't bother with ORP personally, unless you plan on running ozone. That's pretty rare for hobbyists these days though.

I've used the Atlas sensor for pH. It works just fine, but it's a bit expensive. If you're really good with electronics, you can DIY a sensor for pH. A pH probe basically outputs small amounts of voltage (in the mV range) based on what the pH of the fluid is. The typical vales are as follows:

PH : voltage (mV)
0 : 414
1 : 355
2 : 296
3 : 237
4 : 177
5 : 118
6 : 59
7 : 0
8 : -59
9 : -118
10 : -177
11 : -237
12 : -296
13 : -355
14 : -414

If you can create a circuit that can accurately read voltages that low and adequately isolate noise, then you just need a standard BNC connector and you're in business. I'm no expert with electronics, but I'd imagine this isn't trivial.

Apex is really the only game in town at the moment. As much as I love Reef Keepers, Digital Aquatics seems to be all but closed. If you don't want to use Reef-Pi or build your own, it's Apex or nothing. I can't speak to the benefits of an Apex since I don't own one. I don't personally like them for various reasons, but I'm the exception, not the rule.

Currently there are no devices for measuring calcium or magnesium. The Trident is coming soon, but I'd be surprised if it costs less than a thousand dollars. Plus you'll need an Apex to run it.

That's all I know though I'm sure others will chime in with what they know.

 

[drcrook]

Building a sensor from scratch is a bit more than what I want to do. You can already see how wonky commercially available ones are with their calibrations. I'll take a look at Reef Pi, I have probably 40 Pi v2's and 3's laying around. so it sounds like pH, Salinity and Temperature are my key sensors. I like the idea of DIY on this; because I pretty much build this kinda stuff all day and can hook up whatever as long as the price is reasonable.

 

[drcrook]

ReefPi seems to be lacking much of what I would have hoped for in the monitoring space. I'm open to contributing to Reef Pi if I can find affordable pH and Salinity sensors.

I did find this: https://shop.getfishbit.com/products/fishbit-system-plus Looks good; just not available yet.

 

[chipmunkofdoom2]

Agreed on building the sensor from scratch. While the Atlas sensor is a bit expensive, the effort that goes into isolating a sensor from interference and electrical noise can be nontrivial. It's not exactly cheap, but it's probably a decent compromise. There are a few Raspberry Pi/Arduino circuits for pH on Ebay which might be cheaper than the Atlas sensors, but they're not documented very well and I'd be concerned about how well they isolate noise.

With regards to salinity monitoring, this is personally not something I would consider critical. Measuring salinity via conductivity in reef aquaria can be a bit challenging. I personally would not feel comfortable automating any tank functionality using a salinity/conductivity reading. It might be a "nice to know" parameter, but I wouldn't rely on the controller.

For my DIY tinkering, I bought the Atlas sensor and isolated carrier board, but I bought a probe from a different vendor (Bulk Reef Supply). That might be the cheapest option if you're going with the Atlas sensors. Their all-in-one kits are really expensive.

Whichever route you choose, I'm with you on the DIY. Definitely keep us posted!

 

[Ranjib]

Excited to see more folks are interested in DIY controller.
I think temperature is most important one to monitor, followed by ph and then probably Ca/Mg. Salinity and ORP you can do, lot less useful. Most folks use it as a validation rather than something they can use to decide and act on. reef-pi currently support ph (atlas scientific), temperature, water level sensors (photo electric and float switches). We can support salinity and ORP, but I wont do it unless there are significant interest from others. Its expensive thing to do (unless you roll your own MCU based voltage to metric conversion), and I dont see the return of investment. pH support is already there, but it's not documented yet.

I have heard about Ca and Mg sensors for industrial setup, but they are very very expensive (no idea how one works).

There are more than one commercial controllers with similar capabilities, including Apex, GHL etc.
Most folks would prefer the commercial controller because its well testes, supported and above all they dont have to do the hard work to build one. Its time consuming, and prone to errors induced by the builder. Other than that, there are hundreds of DIY controller builds. Folks who come from software engineering tend to prefer Pi, those who comes from embedded systems, prefers arduino. Very few of the re-distribute the software in a way that others can use , and maintained it over prolonged time period, and generic enough to be useful by other. reef-pi is probably the only one that does currently, I hope that changes.

reef-pi uses standard ds18b20 for temperature sensor, pwm is generated using pca9684 adafruit breakout, relays are controlled using uln2803 darlington array, water level sensors are from dfrobot's. its written in go (backend) and react (front end) and it can do 8 outlets, 5 channels pwm (5 or 10v), 2 water level sensors,2 temperature sensor and 1 ph sensor under 30% cpu (300mhz), at 12w (12v 1 amp).

I could'nt spot anything other than ORP/Salinity that reef-pi cant do right now, among the features you mentioned here.
Sorry for the long post , reef-pi author here, so I am bit biased, I hope you can discard that bias in your consideration.

 

[Handegard]

Questions.

1. Do you really need all of those sensors or which ones are most critical? I'm thinking Salinity, Temperature and pH. I'm not sure about ORP.
2. Has anybody built one before and if so; what sensors/vendors did they use? I use Atlas Scientific, albeit they are pricey.
3. This is pretty high for monitoring. How often do folks monitor without these solutions and what is the cost of that typically?
4. Are there additional benefits to the Apex?
5. Are there alternatives to the Apex primarily for water quality monitoring?
6. Are there sensors for Magnesium and Calcium?

1. ORP is not critical, however, it’s one that can be very useful, and doesn’t cost much. Salinity and Temp are what I’d call critical, with pH being important as well.
2. You’ve seen ReefPi
3.This seems like a acceptable price to me. Your a engineer, so you’re looking at the component costs. But as a engineer, what’s your time worth? With the Apex, someone has already done the design, the software, and built a cloud platform. Not to dissuade you, I just see people in industries price themselves as free for home projects, which is true, but not valid for comparison [emoji6]
4. There are no benefits a determined engineer couldn’t replicate with work. But they’ve already developed a cloud platform, created plug and play modules, created a dashboard to display current monitoring, flow data, etc in a visually appealing way. I describe a Apex as a box of Legos. A lot of people want the gadget but don’t fully know why. But once you get it, you have all these parts you can build with. You’ll build something different than me. In the past few months, I’ve actually found a few new uses for my existing equipment. And I just convinced myself to splurge on another part that I don’t see the use for, just because I am so confident that once I have it, I’ll find uses I haven’t imagined yet.
5. GHL makes very nice systems that you may wish to look at. Negligibly more expensive, lacking in some features the Apex has, but also offering some features the Apex doesn’t (GSM modem would be nice Terrance)
6. At least Pinpoint does make a Calcium probe, however it’s not for continuous use, but just for spot testing. There are a couple of Kh monitors on the market, but they aren’t probes, they are automated tests using reagents, and they cost $650+

Neptune has the Trident in testing, and it does Ca/Kh/Mg all in one (automated reagent tests, not probes) and is supposed to be $599. Having the Trident, and a Neptune DOS will give you the ability to automate dosing based on testing, something that might be worth keeping a eye on.



I work in a LFS and I’m a Apex vendor. But my store is a Apex vendor because of how much I loved my personal Apex products, which I convinced the boss to bring in. I truly love mine. I love gadgets, and this was one that I’ll admit I initially bought because I like gadgets. And since then I’ve grown to believe a controller, even if it’s just a ReefKeeper Lite or a Apex Jr is a essential item for any aquarium. My Apex runs my lights, dosing, monitors flow, alerts me if my skimmer pump gets clogged or my return is off, and it runs the auto feeders. I even have it set to turn one of my lights red if there are any serious alarms.

Can you be successful without one? Absolutely. Nobody needs one (though I think a basic controller for heater redundancy is must when we are talking about thousands of $$ of livestock) but once you have one, especially if you are the tinkering type, you’ll keep finding new uses for it, and won’t want to go back.

 

[Ranjib]

1. ORP is not critical, however, it’s one that can be very useful, and doesn’t cost much. Salinity and Temp are what I’d call critical, with pH being important as well.
2. You’ve seen ReefPi
3.This seems like a acceptable price to me. Your a engineer, so you’re looking at the component costs. But as a engineer, what’s your time worth? With the Apex, someone has already done the design, the software, and built a cloud platform. Not to dissuade you, I just see people in industries price themselves as free for home projects, which is true, but not valid for comparison [emoji6]
4. There are no benefits a determined engineer couldn’t replicate with work. But they’ve already developed a cloud platform, created plug and play modules, created a dashboard to display current monitoring, flow data, etc in a visually appealing way. I describe a Apex as a box of Legos. A lot of people want the gadget but don’t fully know why. But once you get it, you have all these parts you can build with. You’ll build something different than me. In the past few months, I’ve actually found a few new uses for my existing equipment. And I just convinced myself to splurge on another part that I don’t see the use for, just because I am so confident that once I have it, I’ll find uses I haven’t imagined yet.
5. GHL makes very nice systems that you may wish to look at. Negligibly more expensive, lacking in some features the Apex has, but also offering some features the Apex doesn’t (GSM modem would be nice Terrance)
6. At least Pinpoint does make a Calcium probe, however it’s not for continuous use, but just for spot testing. There are a couple of Kh monitors on the market, but they aren’t probes, they are automated tests using reagents, and they cost $650+

Neptune has the Trident in testing, and it does Ca/Kh/Mg all in one (automated reagent tests, not probes) and is supposed to be $599. Having the Trident, and a Neptune DOS will give you the ability to automate dosing based on testing, something that might be worth keeping a eye on.



I work in a LFS and I’m a Apex vendor. But my store is a Apex vendor because of how much I loved my personal Apex products, which I convinced the boss to bring in. I truly love mine. I love gadgets, and this was one that I’ll admit I initially bought because I like gadgets. And since then I’ve grown to believe a controller, even if it’s just a ReefKeeper Lite or a Apex Jr is a essential item for any aquarium. My Apex runs my lights, dosing, monitors flow, alerts me if my skimmer pump gets clogged or my return is off, and it runs the auto feeders. I even have it set to turn one of my lights red if there are any serious alarms.

Can you be successful without one? Absolutely. Nobody needs one (though I think a basic controller for heater redundancy is must when we are talking about thousands of $$ of livestock) but once you have one, especially if you are the tinkering type, you’ll keep finding new uses for it, and won’t want to go back.

Very nicely put.

 

[Greybeard]

I am an industrial software engineer, and I, too, researched building a DIY controller. Compared to the systems I work on, a reef controller is a simple toy. I have no doubt I could build a fully DIY reef controller, with all of the functionality that Apex has, and perhaps even more, for way less money. Looked at ReefPI, as well, excellent project, if not quite what I'd build myself.

I bought an Apex.

Why? Time. The single most valuable resource I have. Took me 30 minutes to get the Apex up and running. It's a hobby... there are more enjoyable hobby related activities where I could spend my free time than creating a DIY controller.

I did DIY an Apex break out box. Simple, and the design of the factory BOB leaves a lot to be desired. Why on earth they used a single screw terminal for common across 6 inputs is beyond me. I picked up a cheap 6 x 2 spring terminal strip designed to use for speakers. Works great.

 

[drcrook]

I agree with a controller being essentially a toy compared to my normal work. I'll keep you all posted; but most likely I'm going to end up going a used Apex Controller, Trident or Fishbit. The cost difference between purchasing Apex and building my own is at most $200. The time however is likely at least 80 hours. That is basically like $2.5/hour. Once my kids are a bit older and able to learn, this might make a cool DIY project to do with them. For now, this type of DIY removes time from the family instead of adding to it. The other DIY projects tend to contribute to family time.

 

[Greybeard]

I agree with a controller being essentially a toy compared to my normal work.

Here's a shot of a fairly typical control panel I'd use at a job site. Might be a half dozen similar cabinets in a single install.

An Apex with a few level sensors? Might as well be one of these

I think you'll be happy with your choice, and highly recommend you go with the Apex. They're pretty good units, flexible, lots of 'stuff' you can add.

 

[drcrook]

Here is the latest I just got done with for day job. We basically mod big expensive stuff and add predictive analytics to it. I don't do the demos, I just build.