Temp and pH Controller

[jd]

Hey guys, this project works together with Wes' controller but I thought I'd start a new thread to keep hardware and software separate.

This project is for a cheap temperature and pH controller and monitor. Shooting for under $25, but that won't get you the big screen I have

The Idea:

To build a simple but fail-safe controller to monitor pH and trigger an output once to meets a limit. This could be a solenoid for your reactor or an outside air pump.

To build a redundant temperature controller using two separate probes whose values are compared to one another(with an alarm that warns if they are too far apart) and averaged together many times a second to give an accurate temperature reading.

How it is done:

The setpoints will be set(heh) by simple potentiometers. Simply press and hold one button to enter the "SET" mode. One pot for pH, one for low temperature limit and one for high temperature limit.

You will select if you'd like your pH relay to trigger above or below your setpoint ( but I haven't found a reason to trigger above it yet )

At your low temp setting a transistor will bias allowing you to supply your own voltage to power your relay for your heater, likewise for the high limit to turn on chiller or fans.

Any temperatures 2 degrees above or below your set points will sound an alarm (or bias a transistor)

The beauty of this system is, even if power is removed, once it is restored the controller already knows your set points and will take the correct measures to adjust the temp or pH.

If there is significant interest I may include logging capabilities, this will add a few dollars to the cost.

Without further adieu!

(yes, it is cold in my house)



Have a great Holiday everyone, I'm looking forward to your feedback!

[Rayman]

very cool, iiiiiiiiiiiiiiiiiiiiiiii like it!

[rayviv]

JD, I don't understand it but I like it. Great idea coming together, impressive! I'll take ten!! Or maybe just one to start

[slandis3]

Its comeing along very nice. how much would it cost to add the screen

[cyberwollf]

Nice. I'm starting to get excited about the upcomming ozone gen diy!

You'll have to bring me up to speed on the arduino on sat

[CoralBeauties]

I would be interested in the temp controller.  I want to run a larger heater with my controller (aquajr) but want to have a failsafe incase the heater would become stuck on for some reason.
Jeff

[jd]

very cool, iiiiiiiiiiiiiiiiiiiiiiii like it!

Thanks

JD, I don't understand it but I like it. Great idea coming together, impressive! I'll take ten!! Or maybe just one to start

Its pretty simple Ray, you hook up your heaters and your fans to it and it keeps your tank temperature within a set range, and warns you if something goes haywire. The pH side can he as simple as simply giving you a realtime reading.

Its comeing along very nice. how much would it cost to add the screen

My goal was to us a 8x2 screen, that is 8 characters across on two rows. That way I could format the screen like

pH 8.20
79.2 OKY

But with this larger screen you could display multiple temps, like the temp in your hood or ambient temp. I got the bigger screen because I'm going to use it for other things also. Bottonm line is the larger screen will cost you about $5 more depending on the color and backlight.

Nice. I'm starting to get excited about the upcomming ozone gen diy!

You'll have to bring me up to speed on the arduino on sat

Me too, O3 is supposed to make your water crystal clear.

This language is super easy, here is a (non-working) program

/*
pH and temp controller
 */

// include the library code:
#include <LiquidCrystal.h>

// initialize the library with the numbers of the interface pins

LiquidCrystal lcd(12, 11, 5, 4, 3, 2);

int sensorPin = 0;
int temp;
int count = 5;
int average = 0;
int var;

void setup() {
  Serial.begin(9600);
  // set up the LCD's number of rows and columns:
  lcd.begin(16, 2);
  // Print a message to the LCD.
  lcd.print("  Hello, Ohio Reef!");
}

void loop() {

while (count > 0){
    int reading = analogRead(sensorPin);
     float voltage = reading * 5.0 / 1024;
     float var = (((voltage - 0.5) * 100) * 9 / 5) + 32;
     
      average = var + average;
        count --;
    Serial.println(count);
    Serial.println(average);
    Serial.println(reading);
    
  delay(5000);}

 temp = (average / 5);
 count = 5;
 average = 0;
  
  // set the cursor to column 0, line 1
  lcd.setCursor(0, 1);
  // print the number of seconds since reset:
  lcd.print(temp);
}

Writing serial data is as easy as Serial.print("message");

The LCD is much the same also:

lcd.print("  Hello, Ohio Reef!");

I would be interested in the temp controller.  I want to run a larger heater with my controller (aquajr) but want to have a failsafe incase the heater would become stuck on for some reason.
Jeff

Great! I am all about redundancy also. This code will be tested an absurd amount to make sure everything checks out

[RTN_H8ter]

Nice. I'm starting to get excited about the upcomming ozone gen diy!

You'll have to bring me up to speed on the arduino on sat

I have everything to build my ozone generator. I'm just too lazy to put it together. Plus lighting fluorescent bulbs 5 feet away kinda freaks me out.

[lazylivin]

Great job guys, it is coming together very nicely

[jd]

Great job guys, it is coming together very nicely

Thanks Brian. Not to shabby considering I only for the uC yesterday night at about 7

[Blown76mav]

Looks good, to bad it probably won't work with a MAC

[cyberwollf]

Looks good, to bad it probably won't work with a MAC

not trying to speak for Josh, but I think we are leaning more towards running it like a normal controller with the CPU only as an added logging/display device. So you're in luck

[jd]

Looks good, to bad it probably won't work with a MAC

Actually Lonnie, this is completely hardware based so no computer is required. If I implement logging, you're still in luck! the IDE (development software) runs in *nix, MAC and PC so it will work fine. Logging will be pretty rudimentary at best, plug in the controller, run the IDE software, tell the controller to transfer the data and then it just spits out the readings with some time/date info along side it.

not trying to speak for Josh,

Isn't that like speaking for yourself also? 

[hamiltro]

sign me up for one, cant wait to see the finished project

[jd]

I don't know how happy I am with these temperature sensors..

Moment to moment they vary by about +/-1.5* F. However since most errors on things like this go positive as equally as they go negativity I have been averaging 60 reading together to get a temp. I don't have anything really accurate to compare too. I will do the ice-water test once I have one water proofed . Digital sensors exist but are a pain to use. It takes about 5 lines of code to read this sensor, digital ones take about 50. 

Once I have some sensors water-proofed I'll get more accurate results to refute or confirm my suspicion. In any case these are working for now so work has begun on the set-points and alarms. Will post details when they exist.

Anyone want to confirm they'll buy one? 

[jd]

Set-points... uhh... set.

How come I always work when everyone else is asleep? That make me like Santa?

[HUNGER]

no ur the elf 

[jd]

no ur the elf 

I see how it is....


HI and Lo set-points are done, as well as status messages. Status messages are OKAY for temp in range. WARN for temp above or below the set-points, when the heater/chiller outputs turn on. and ALRM when the temp is 2 degrees over or under the setpoints. All of these measurements are taken and compared to the status multiple times to prevent frequent cycling of your equipment. The heater/chiller output stay on until the temp reaches the middle of your setpoints

Now that I got 4 hours of sleep, I'm off to work...

[cyberwollf]

OK, heres a pic of my test.  Not getting the exact results i want:

1. No flow, so water temp is not uniform.
2. I covered the sensor in epoxy, heatshrink tubing, and more epoxy (overkill maybe, but its watertight  ) But i think it is insulted too much, rising temperature readings lag behind my meat thermometer.

Very simple drive circuit (about 99% of the crap on the breadboard is just there becuase im too lazy to disassemble other projects.

Reading a Temp probe, is exactly like reading a pH and other probes, so the project is actually making much bigger jumps than just temp.

I'll post a graph when i get the kinks worked out.

[slandis3]

I have titanium tubing and 2part epoxy to seal the probe in. I have 1 pc sitting here now and Ill get anothe couple pieces tomorrow at work.

[Wall_Tank]

OK, heres a pic of my test.  Not getting the exact results i want:

1. No flow, so water temp is not uniform.
2. I covered the sensor in epoxy, heatshrink tubing, and more epoxy (overkill maybe, but its watertight  ) But i think it is insulted too much, rising temperature readings lag behind my meat thermometer.

You should insulate the probe, as to slow the response to temperature changes.   Also, remember you have a huge heater for your water volume, and no flow.   So I would very much expect you to have control problems that you'll never see in a tank.    I would worry about things like......calibration....and does it go off and on when you want.

[cyberwollf]

You should insulate the probe, as to slow the response to temperature changes.   Also, remember you have a huge heater for your water volume, and no flow.   So I would very much expect you to have control problems that you'll never see in a tank.    I would worry about things like......calibration....and does it go off and on when you want.

Of course this would be worse case.. I in the tank +100 gallons of water make a pretty nice thermal body. 

Ok i moved it to a bucket with a pump and a little water, but i think the pump was putting out enough heat to keep the water at temp (only a few degrees above room temp)

I'll try it again tomorrow with a full bucket of water and let it run longer. Heres the graph from today: "On or Off" just shows when the heater was on or off, The numbers are arbitrary, just wanted to log its state. 

all the logic works great  Also I am only sampleing one every 5 seconds. I could get better readings by taking many every second and averaging them.

[jd]

Got some more work done on my controller (which has 100% more LCD screens than Wes' does  )

It works in the following manner, I'd like to hear your comments on it.

You set your low temp point to 74* and your high temp point to 84*

If your tank falls below 74* your heater turns on.

Your heater remains on until the temperature reaches 79* (this is the point halfway between your two set points)

This of course all works in the opposite manner as well.

Anyone have any features they would like to suggest?

Will had a good suggestion of a "cold house" setting. If our home is cold due to you being away and one heater isn't enough, the controller would see how long it has been since it turned it on, and turn on another one to speed the heating of your tank.

What temperatures are you running your tanks at? How much of a 'swing' would you like?

[lazylivin]

I have my controller set to turn the heater on at 78.5 and below so there is no high point to avoid swings. Tanks is always 78.5
If temperature reaches 80 fans kick on until temperature drops to 79.5 so have high and low point
If temperature reaches 82 lights shut off and wont turn back on unless I reset to signal tank temperature was breached

With temperature consistency is key. So if tank wants to naturally peak at 80 I would recommend setting heater to 80. A fluctuation of 4 degrees in 24 hour period is to much. Some may get it away with that however the fish and corals would do better with more stability.

[cyberwollf]

On the idea of controlling multiple heater independently it would be fairly simple to implement some fuzzy logic: Depending on how far you are from optimal temp determines how many heaters turn on. I.e. It may only take one heater to maintain temp, but if you add cold top off water or for some other reason temp drops lower than normal multiple heaters kick in to keep it at setpoint.


I don't really understand the your idea on the setpoints.... If it turns back off at 79 why isn't that the setpoint? What are you trying to do that 2 setpoints a few degrees apart can't?

[jd]

I have my controller set to turn the heater on at 78.5 and below so there is no high point to avoid swings. Tanks is always 78.5
If temperature reaches 80 fans kick on until temperature drops to 79.5 so have high and low point
If temperature reaches 82 lights shut off and wont turn back on unless I reset to signal tank temperature was breached

With temperature consistency is key. So if tank wants to naturally peak at 80 I would recommend setting heater to 80. A fluctuation of 4 degrees in 24 hour period is to much. Some may get it away with that however the fish and corals would do better with more stability.

Thanks for your input Brian. The 10* range was a poor choice on my part, i assume it would be more like 78-81 or the like. How much would you say your temp swings in a given day?

I don't really understand the your idea on the setpoints.... If it turns back off at 79 why isn't that the setpoint? What are you trying to do that 2 setpoints a few degrees apart can't?

The two set points are to allows people to fine tune their temps. You know that if have your heater turn on at 79 and off and 79.1 it will cycle on and off almost constantly. I guess I could have one set point and have heaters on at 1.5* below that and coolers on at 1.5* or something similar.

[lazylivin]

I am not sure how the reefkeeper controller I have does it but for the heater port it only gives you an on temperature. There is not a high setting. The tank temp is always, 24 hours a day 78.4 - 78.6. The only time it goes out of the range is if I manually turn on the frag tank lights while the display tank lights are on. I have seen in that case it go to 80 and the fan kick on.

Think of it like this Josh. The heater itself has its own built in thermostat. It is likely that it is cycling on and off with tolerance less then .5 degrees. We could test that, fill a bucket with water, stick a heater, small pump to circulate the water and a digital thermometer. Turn the heater temp above the ambient room temperature and see if the temp fluctuates.

[cyberwollf]

Brian, i understand joshs worry. You don't want your heater short cycling every few seconds to keep water within 0.01 degrees. There is some hysterisis needed to prevent that.  The more I think of it now unless you a super high watts to gallons ratio of heaters. The transition will be inherently slow. I bet we could get away with 0.5 degree setpoint.  Thanks for the RK input. Intresting design idea to simplefly user input. I'm sure they just take your input and setpoints +-0.2 or something close.

All these questions should be answered when we get a prototype all working. We'd be able to monitor the cycling frequency on all outputs and make adjustments as needed. I know I don't want a 10 degree temp swing with this thing on my tank

another idea I have been kicking around for what I care about in a temp controller. With 3 Normal heaters all set at 79 they keep my tank +-1 degree.  I don't know if I want the complexity or addiontional possible failure of an independent controller. What I'd really like is only a normally closed relay that would open to kill the heaters if they got stuck on, so it would prevent cooking the tank. And of course overtemp light shutdown etc. But leave the temp control up to the actual heaters. Anyone had a problem with a heater other and being stuck on?

[jd]

I'm sure they just take your input and setpoints +-0.2 or something close.

That is exactly what I think they do also.

I know I don't want a 10 degree temp swing with this thing on my tank

It was a poor example! It was just what they happened to  be when I wrote that!

But leave the temp control up to the actual heaters. Anyone had a problem with a heater other and being stuck on?

You trust your heater thermometers more than ours? I think the inherent problem with most heaters is, their temp sensor is very near the element. Unless your heater is in a very high flow area, this can effect its operation. Also, how do you expect anyone to use one of these the one of its developers doesn't even want to?! 

[lazylivin]

You are likely right that there is a inhernet +\- built in to the controller and he brand controller probrably handles it diffrently

[jd]

Yes, I'm not so much worried about the cycling of the heater as I am of the relay, and SCRs are expensive...

Maybe I should just get a RK and be done with it

[lazylivin]

Yours is going to be much cooler.

[cyberwollf]

I say just code up the +-0.2 and be done with it. Allows a user to only have to input one setpoint.

[jd]

Well, that frees up some more IO pins. I may do a .2 over set is one relay one and .4 over is another one. That way you could use fans for .2 over or turn your chiller on at .4. That'd save energy and wear and tear on chillers. How does that sound to everyone?

[cyberwollf]

Mine is bigger....

[jd]

Yeah... uh... mine is... uh... bluer..

[HUNGER]

looks good guys 

Mine is bigger....

[jd]

Wes and I spend the evening talking about phase detractors and stators. We also managed to take a peak inside of a Pin-Point pH monitor. You'll be angered to know that the pH and ORP probes only differ by what appears to be 1 part. Also, the designer at Pin-Point forgot the importance of protection diodes and the reason why you use sheilded cables. 

New temperature probes are on the way, and should yield much better results.

[UDJustin]

where did you get the pinpoint ph monitor from

[jd]

Wall_Tank loaned it to us. Thanks again Wall.

[cyberwollf]

You'll be angered to know that the pH and ORP probes only differ by what appears to be 1 part.

He meant the monitor, not the actually probe itself.  And yes, they just add a jumper/resistor and sell it as ORP. BTW VERY POOR craftsmanship.. As Josh said they have crap just soldered in that they apparntly forgot about.  And the whole purpose of using a shielded coax cable with BNC connectors is voided by the little wire that is jumpered from one side of the board DIRECTLY to the ADC pin...

[UDJustin]

I had a pinpoint ph controller but it got wet and while it still turns on the number read out makes no sense lol.

[jd]

I had a pinpoint ph controller but it got wet and while it still turns on the number read out makes no sense lol.

Do you still have it?

[slandis3]

Do you still have it?

I have it at my house. I can bring it to the meeting so you can look at it.

[jd]

I have it at my house. I can bring it to the meeting so you can look at it.

If you (or Justin) have written it off as dead. I'd love to take it all apart and reverse engineer the circuit.

[Wall_Tank]

If you (or Justin) have written it off as dead. I'd love to take it all apart and reverse engineer the circuit.

Or at lease reverse engineer what not to do.

[jd]

Or at lease reverse engineer what not to do.

Heh, no kidding. That seems to me likely. They do a few things that I'd like to find the reasoning behind.

[jd]

Any by this week you mean the end of the month?? lol  Man, this project is actually super simple.. we are just lazy lol

did you ever figure out if we can burn arudino to a chip without the overhead of usb and all the other crap on a arudino board?

The USB section is a small part of our overhead, we have to use it because it includes all the delay, acd and lcd protocols.


See if you think this will work, I can't tell if it will do +/-2V or not. Its a shame the one I found yesterday is so dang small.

http://ww1.microchip.com/downloads/en/DeviceDoc/21409b.pdf

[cyberwollf]

im not familar with the design.  from wiki:

A Sigma-Delta ADC (also known as a Delta-Sigma ADC) oversamples the desired signal by a large factor and filters the desired signal band. Generally a smaller number of bits than required are converted using a Flash ADC after the Filter. The resulting signal, along with the error generated by the discrete levels of the Flash, is fed back and subtracted from the input to the filter. This negative feedback has the effect of noise shaping the error due to the Flash so that it does not appear in the desired signal frequencies. A digital filter (decimation filter) follows the ADC which reduces the sampling rate, filters off unwanted noise signal and increases the resolution of the output. (sigma-delta modulation, also called delta-sigma modulation)

[jd]

You totally dodged my question I'm reading a new datasheet now, it looks promising.

[cyberwollf]

Another idea to simplify design.  Ususaly you are only running a probe to monitor the tank OR a reactor. So if you know which case you will use exclusively we can set the onboard ADC to read it.

for Cal reactor are are running ~4 Ph.  That would be ~-0.18V output.  If you run that through an inverting buffer you get 0.18V and just set the ADC from 0-0.25V and you should get excellent resolution

for tank monitoring ~8 pH.  that would be ~0.10V.  If that was put through an NONinverting buffer and set the ADC for 0-0.15V you should get excellent resolution.

I would think it would be rare to measure both tank and reactor with a single probe. Is that assumption correct? That was our problem. We had to be able to scale the voltage to all positive without corrupting a very small signal.  At the very least this REALLY simplifies the dedicated Cal reactor controller.

[cyberwollf]

The above idea would allow 2 probes to run on independent ADCs with a single 0.25v reference voltage.

One port would be for the acidic probe while the other was for the basic probes. 

[jd]

If we can find a ADC that will handle all of our application without having to throw switches, I'd rather do that. This way, we can use one ADC and multiplex a large number of inputs while reusing the same conversion code, yeah know?

EDIT: two ADC's add cost....

[j80]

Bump, could i get a parts list and where to buy them for this? Thanks

[cyberwollf]

Like everything me an Josh did, this never got finished lol.  pH is a pita becuase you have to multply the probe output by -17 (if i remember correctly).  What are you trying to do?  honestly in the time since we played around with this project the new reefkeepers lites are SOO cheap now, its almost not worth it