PO4 Controller possible?

Sherman · Aug 2, 2018

Recent KH controller has help many to resolve the Alk issue .
The other big problem I think is high PO4 issue.
Is it possible to DIY a PO4 controller ?

Gareth elliott · Aug 2, 2018

Im not the technology is there yet.

Every po4 test on the market has a lot more involved than most alkalinity tests. Also phosphate readily bonds, leaches, changes forms.

Perhaps a similar idea would be an auto feeder controlled by DOC. Would make keeping certain corals a lot easier.

WWIII · Aug 2, 2018

One day we will have it. One of the issues is the levels we are trying to maintain in a reef tank are so low. As of now the hanna ulr phosphorus checker is about as good as it gets. To make that automated, I'm sure is possible. To do it well, being very accurate to ppb levels, and for a reasonable price may be the challenge. I would personally rather have a phosphate monitor over magnesium and probably over calcium as well. As one said here on R2R, alkalinity monitoring was the holy grail, but phosphate monitoring would be a nice icing on the cake!

mitch91175 · Aug 3, 2018

WWIII said:
One day we will have it. One of the issues is the levels we are trying to maintain in a reef tank are so low. As of now the hanna ulr phosphorus checker is about as good as it gets. To make that automated, I'm sure is possible. To do it well, being very accurate to ppb levels, and for a reasonable price may be the challenge. I would personally rather have a phosphate monitor over magnesium and probably over calcium as well. As one said here on R2R, alkalinity monitoring was the holy grail, but phosphate monitoring would be a nice icing on the cake!

Agree 100% on monitoring calcium and magnesium. Those aren’t as critical as alkalinity. Magnesium doesn’t change that much and calcium can be handled easily with dosing kalk and bi weekly testing if that. Magnesium testing once a month if that.

ZaneTer · Aug 3, 2018

Personally I would monitor calcium and magnesium daily but only if you are dosing. This can help you detect a failure in your dosing system whether it is under or over delivering. I am tempted to buy a second hand Hanna ULR PO4 meter and hack it to be automated. In the end the cost will probably come to around $200 in my case but I have a custom controller so the costs are relatively low. Reagent costs are reasonable so I am not too worried. Coupled to a PID loop and an accurate doser( to 1/100th of ml, this can even be done with a $5 dosing head controlled by a PLC operating at <1ms) you should be able to achieve near perfect PO4 stability. Additions can be done by dosing pump and subtractions by increasing the lighting schedule on a fuge or by dynamically dosing vinegar.

Sherman · Aug 3, 2018

ZaneTer said:
Personally I would monitor calcium and magnesium daily but only if you are dosing. This can help you detect a failure in your dosing system whether it is under or over delivering. I am tempted to buy a second hand Hanna ULR PO4 meter and hack it to be automated. In the end the cost will probably come to around $200 in my case but I have a custom controller so the costs are relatively low. Reagent costs are reasonable so I am not too worried. Coupled to a PID loop and an accurate doser( to 1/100th of ml, this can even be done with a $5 dosing head controlled by a PLC operating at <1ms) you should be able to achieve near perfect PO4 stability. Additions can be done by dosing pump and subtractions by increasing the lighting schedule on a fuge or by dynamically dosing vinegar.

Hi ZaneTer,
Your idea sound interesting.
Hana ULR is only monitoring. I just wonder how to hack it to automate the dosing.
Maybe you like to share your idea .....

ZaneTer · Aug 3, 2018

Do you want the long description or the short one? ;)

ZaneTer · Aug 3, 2018

Which part do you want the details of? The Hanna hack or the automatic dosing and compensation? Or both? I am happy to describe both.

Sherman · Aug 3, 2018

A brief description about the Hanna hack.
I belief this is the first step to automation

ZaneTer · Aug 3, 2018

Give me half an hour or so, just helping some Dutch people get to their destination here in China, they can’t speak any mandarin.

Sherman · Aug 3, 2018

Take your time. Thank You

ZaneTer · Aug 3, 2018

Do you have WeChat?

ZaneTer · Aug 3, 2018

For the Hanna unit:
1) I need to attach in a better power supply than a battery. I don’t like batteries.
2) Replace the button on the front and connect it to my controller so I can signal it when I need to. Single digital output to a zero volt interposing relay.
3) modify the lid so I can add and remove a sample but using injection needles (this increases accuracy as drop size decreases). Connect three lines to the head. One for mixed fluid, one for waste removal and the last for RO water flush line.
4) Prepare a mixing chamber where reagent will be prepared. Connected will be a water sample line, waste line and mixing line. Each line has its own dosing head. They are cheap at about $5 per head. I will explain accuracy later. I’ll probably just construct this from a short piece of acrylic tubing.
5) Connect the BCD display to my controller so I can read the status and final figure.

Next is the long part

ZaneTer · Aug 3, 2018

1)I will turn on the checker by using my digital output for a pulse. I can read back the display to know if it came on or not and to then trigger an alarm.
2) I will add a sample of 10ml +- 0.01ml of tank water.
3) Trigger my output again so that it reads C1.
4) When I see the display feedback I will drain the cuvette. I will prepare another 10ml sample into my mixing holder.
5) I will meter out my dry reagent ( I may use a high accuracy load cell to milligram accuracy)
6) I will run my circulation pump for 2 minutes
7) Transfer to my cuvette
8) Trigger my output again to start the test and wait for the end result.
9) Read feedback and transfer the current level to my PID loop to adjust output of PO4 dosing pump or refugium lights.

Please remember this is just it in brief and I am using an industrial controller (S7-315) AND automation is my day job.

If you have questions or comments please just send them my way.

Thanks
Zane

ZaneTer · Aug 3, 2018

I forgot to add that I will have to do a flush sequence immediately after to keep the mixing container and cuvette clean. This will be done with RODI.

Running a hack like this will allow me to run a test whenever I like as well as reliably track and trend the data. For 10 years if I like

ZaneTer · Aug 3, 2018

Accuracy:

I mentioned this because currently GHL have the top spot for dosers but I see their accuracy as horrendous. They use a beautiful stepper motor but it is both expensive and truly not the most accurate as I will explain.

I use a cheap amazon $5 peristaltic doser. My particular doser operates at 100ml/min

Here is the cool part. My controller operates at <1ms but for ease of use I run it at 1ms exactly. This means that I can reach per millisecond dosing.

100ml/min divided by 60000 milliseconds per minute equals 0.0017 ml/ms

I have to run my code for around 6ms or 6 cycles to be able to dose 1/100th of a millitre.

GHL dosers aren’t looking quite so good anymore at this point.

Cost:

I will not include the controller as it costs around the same as a profilux P4. Lets have some fun calculating this. We will do the calculation based upon adding 32 individually controlled dosing heads.

Siemens parts:
1 x 32 digital output card @ $250
32 x $3 interposing relays @ $96
32 x $5 dosing heads @ $160
12v power supply @ $40
Wiring @ $50
Time ?????
Total $544

GHL
8 x 4 head dosing slaves ($400 each)
Total $3200

It’s a funny world we live in when $550 system beats the hell out of a $3200 system.

As always please feel free to ask any questions or if you have any comments. I will always happily answer your questions to whatever you technical level may be.

ZaneTer · Aug 3, 2018

Accuracy:

I mentioned this because currently GHL have the top spot for dosers but I see their accuracy as horrendous. They use a beautiful stepper motor but it is both expensive and truly not the most accurate as I will explain.

I use a cheap amazon $5 peristaltic doser. My particular doser operates at 100ml/min

Here is the cool part. My controller operates at <1ms but for ease of use I run it at 1ms exactly. This means that I can reach per millisecond dosing.

100ml/min divided by 60000 milliseconds per minute equals 0.0017 ml/ms

I have to run my code for around 6ms or 6 cycles to be able to dose 1/100th of a millitre.

GHL dosers aren’t looking quite so good anymore at this point.

Cost:

I will not include the controller as it costs around the same as a profilux P4. Lets have some fun calculating this. We will do the calculation based upon adding 32 individually controlled dosing heads.

Siemens parts:
1 x 32 digital output card @ $250
32 x $3 interposing relays @ $96
32 x $5 dosing heads @ $160
12v power supply @ $40
Wiring @ $50
Time ?????
Total $544

GHL
8 x 4 head dosing slaves ($400 each)
Total $3200

It’s a funny world we live in when $550 system beats the hell out of a $3200 system.

As always please feel free to ask any questions or if you have any comments. I will always happily answer your questions to whatever you technical level may be.