cypho's 36" DIY reef (1 Viewer)
- Thread starter cypho
- Start date
nate_mcnasty
Senior Member
There are other articles that review the gran plot for seawater, but I did some further reading and it was only a small improvement from the inflection point method. Not worth the headache.
Hanna has standards to validate Alk checkers. Could you try to use these to validate your titrator?
The hana standards are worthless. Both to me and to users of Hana Checkers. You don't actually run the test on the standard to validate the reagents and procedure. You just stick the vial they send you in the checker and it confirms the color sensor is working.
I redid the lower half of my plumbing (between sump pump and bulkhead) a few weeks ago. It was originally 1" sch 40 pipe, I replaced it with 1.25" thin-wall pipe. It still drops down to a 3/4" bulkhead and then splits into 2x 3/4" sch 40 pipes for the rest of the run. Even though I only replaced a 30" section of pipe, the gains were immediately obvious. My drain could not handle the increased flow. Had to dial the pump back to quite a bit.
To accommodate the increased flow I replaced the 3/4" sch80 drain with 1" sch80 pipe. I calibrated the new flowmeter today and I can finally put a number on the gain. Old max flow rate was 690 GPH. New max flow rate is 925 GPH. Still nothing near the pumps 2700 GPH rating, but a pretty good improvement.
Yes I know it is backwards that I am using thin-wall on the higher pressure line and SCH 80 on the low pressure drain. The SCH 80 is purely for the grey color. Max head for the pump is something like 7 PSI. Actual working pressure is probably no more than half that. The thin-wall pipe is rated for 140psi. So the pipe is only 20x-40x stronger than needed. I think I can live with that....Especially since the interior diameter is nearly one size up.
To accommodate the increased flow I replaced the 3/4" sch80 drain with 1" sch80 pipe. I calibrated the new flowmeter today and I can finally put a number on the gain. Old max flow rate was 690 GPH. New max flow rate is 925 GPH. Still nothing near the pumps 2700 GPH rating, but a pretty good improvement.
Yes I know it is backwards that I am using thin-wall on the higher pressure line and SCH 80 on the low pressure drain. The SCH 80 is purely for the grey color. Max head for the pump is something like 7 PSI. Actual working pressure is probably no more than half that. The thin-wall pipe is rated for 140psi. So the pipe is only 20x-40x stronger than needed. I think I can live with that....Especially since the interior diameter is nearly one size up.
Last edited:
Crazy how much of a difference that makes. Guess you'll have to size up to a 1.5" next (o:
I have a 5' section if you're interested and I've got some great "do as I say not as I do" bulkhead drilling tips!
I have a 5' section if you're interested and I've got some great "do as I say not as I do" bulkhead drilling tips!
It is tempting to replace the 3/4" bulkhead with something larger. But drilling a new hole in the bottom of the tank while the tank is full of water seems very risky.
Davy Jones
Senior Member
There is no way i could ever talk myself into doing that. I would be shaking too bad to drill it correctly lol
See we'll just make this acrylic box to go on the inside with a rubber gasket and let it pressure seal to the area around the hole. Problem solved... what could go wrong?
Shortly after my last post I was fiddling with things and all of a sudden the raspberry pi went up in smoke. I ordered a new pi, and went looking for the source of the problem. I did find an explanation (one of the parts was configured to expect 5v instead of 3v), so I fixed that setting and when the new pi arrived I plugged it back in and guess what - it immediately went up in smoke again. So I ordered yet another pi. (good thing these pi zeros only cost $10) This time, I totally disassembled everything looking for the real source of the problem. I found a 12v wire that had broken loose and was touching a 3V wire. That would certainly explain the issues.
When I re-re-assembled the pi did not go up in smoke but nothing else was working. After quite a bit of troubleshooting I eventually ended up confirming that everything on the 3v circuit was toast. I had spare parts for everything except the isolation circuitry for the pH probe, so I re-assembled without that part. From my experience building my aquarium controller I found that Isolation was absolutely required with the probe in the sump with all sorts of other electronics, but with the probe in glass vial without anything else, it shouldn't be as necessary. But it turns out the magnetic stirrer causes enough interference pH readings all over the place. So I stole the isolation chip from my aquarium controller (no more pH on the aquarium controller for now - that part has a super long lead time).
You know how when one thing goes wrong, everything goes wrong. At the same time I was having electronics issues, I also ran out of acid. I wanted to order a bigger bottle this time to reduce the per-test cost, but the bigger bottle was out of stock. So I decided to give 0.1N acid a try. 10x as potent as what I was using before. If it works it will reduce the cost even more than ordering the acid in bulk would. But it added yet another variable and complicated the troubleshooting since I could not rule out the issues being related to the new bottle of acid.
And just to make things harder on myself, I figured it would be a good time to order a liquid filled glass pH probe. http://a.co/h4SAktj
Turns out that was a mistake. The liquid filled probe only lasted 2 days. I think it just out of liquid, but I don't have any liquid to refill it so I can't confirm that is the problem. I don't know if all liquid probes need to be refilled that often or if I just choose the wrong probe. But unless I build an automatic probe refiller to automatically refill the probe every day, that liquid probe is not going to work out.
And to make this even more of a disappointing purchase, with all of the other changes and troubleshooting going on at the same time, it was out of liquid before I had everything else working well enough to find out if the liquid probe was going to speed up the test. So that is back in the spare parts bin, and I went back to the gel probe.
And finally, after a month of pounding my head against the wall, this afternoon, I finally got a halfway decent test! Not yet back to the level of precision I was at before, but I finally feel like I am making progress again.
When I re-re-assembled the pi did not go up in smoke but nothing else was working. After quite a bit of troubleshooting I eventually ended up confirming that everything on the 3v circuit was toast. I had spare parts for everything except the isolation circuitry for the pH probe, so I re-assembled without that part. From my experience building my aquarium controller I found that Isolation was absolutely required with the probe in the sump with all sorts of other electronics, but with the probe in glass vial without anything else, it shouldn't be as necessary. But it turns out the magnetic stirrer causes enough interference pH readings all over the place. So I stole the isolation chip from my aquarium controller (no more pH on the aquarium controller for now - that part has a super long lead time).
You know how when one thing goes wrong, everything goes wrong. At the same time I was having electronics issues, I also ran out of acid. I wanted to order a bigger bottle this time to reduce the per-test cost, but the bigger bottle was out of stock. So I decided to give 0.1N acid a try. 10x as potent as what I was using before. If it works it will reduce the cost even more than ordering the acid in bulk would. But it added yet another variable and complicated the troubleshooting since I could not rule out the issues being related to the new bottle of acid.
And just to make things harder on myself, I figured it would be a good time to order a liquid filled glass pH probe. http://a.co/h4SAktj
Turns out that was a mistake. The liquid filled probe only lasted 2 days. I think it just out of liquid, but I don't have any liquid to refill it so I can't confirm that is the problem. I don't know if all liquid probes need to be refilled that often or if I just choose the wrong probe. But unless I build an automatic probe refiller to automatically refill the probe every day, that liquid probe is not going to work out.
And to make this even more of a disappointing purchase, with all of the other changes and troubleshooting going on at the same time, it was out of liquid before I had everything else working well enough to find out if the liquid probe was going to speed up the test. So that is back in the spare parts bin, and I went back to the gel probe.
And finally, after a month of pounding my head against the wall, this afternoon, I finally got a halfway decent test! Not yet back to the level of precision I was at before, but I finally feel like I am making progress again.
Have you considered wire wrap instead of push pin? After the second short across my DIY set up I build the entire thing using template PCB with all low volt signles through wire wrap. It's really easy (and stable) to secure wire wrap with solder and components on PCB.
I don't use any push pins. I have a few screw terminals on my main reef controller but other than those screw terminals, all of my connections are soldered. Wire-to-PCB using thou-holes or wire-to-wire by twisting the wires together and then applying a big pile of solder over the top to keep them from separating.
The loose wire problem was not caused by a loose connection. The wire broke, probably from heavy handed use of the wire stripper cutting too much into the copper.
Continuing to make software improvements.
I have not done same sample back to back testing to determine repeatability, but the curves using 0.1N acid are beginning to look pretty good!
Sent from my Pixel using Tapatalk
I have not done same sample back to back testing to determine repeatability, but the curves using 0.1N acid are beginning to look pretty good!
Sent from my Pixel using Tapatalk
Too many words in the last few posts. How about some pictures?
Sent from my Pixel using Tapatalk
Sent from my Pixel using Tapatalk
Last edited:
Sure looks good considering where it was a short time ago.
I finally got around to doing the back-to-back tests on the same sample using the stronger 0.1N acid. 3.129, 3.137, 3.125, 3.144, 3.131
So that is +/- 0.0095
As long as the test is more precise than my alk is stable there is no need for a more precision in the test. And right now my alk is bouncing around by 0.05 - 0.1. So although it is not as good as my early tests with 0.01N acid, I think it is probably good enough for now.
So that is +/- 0.0095
As long as the test is more precise than my alk is stable there is no need for a more precision in the test. And right now my alk is bouncing around by 0.05 - 0.1. So although it is not as good as my early tests with 0.01N acid, I think it is probably good enough for now.
When is the retail version set to be released 
Sent from my Pixel using Tapatalk
Sent from my Pixel using Tapatalk
Is your KHG still out of commission?
I'm not totally done with the project documentation, but it is mostly done. If you want to take a shot at building a DIY titrator, I have written up enough of the details to get started.
https://github.com/cypho/auto-titrator/wiki/Shopping-List
It's running and seems to be ok. CO2 levels are definitely causing interference in the results. I wouldn't mind another sig fig as well.
If I ever get enough free time, I may just start tinkering
Sent from my Pixel using Tapatalk
- Joined
- Jul 5, 2009
- Messages
- 2,523
cheezybuda how is CO2 impacting the test results I run a calcium reactor that pumps a lot of CO2 into my system and the results from the KHG are consistent with other three alk test kits I reference against?
I would like to tinker with building one of these units, thank you cypho for providing all of the documentation to start the project.
I would like to tinker with building one of these units, thank you cypho for providing all of the documentation to start the project.
I don't know if KHG has shared the details of exactly what it is doing. But if KHG is using the fixed endpoint method, then fluctuating CO2 levels could cause the results to fluctuate in theory.
However the impact is very small. My titrator does not use a fixed endpoint so it should not be influenced by CO2. I can see inflection point swing up and down a bit from day to night. But at least with the CO2 fluctuations in my tank, the effect is very small. The inflection point moves around by 0.05 pH which would correlate to an error of around 0.01 meq/l if I was using a fixed pH endpoint.
If your CO2 if bouncing around enough to be a real problem for the test, then the any alk swings it causes will be the least of your worries. The pH is probably getting so low that the low pH is a much bigger issue than the slight alk swing it causes.
The huge disadvantage of using a fixed pH endpoint from my perspective is that you have to carefully calibrate the pH probe. If you are using a fixed endpoint, a poorly calibrated probe is going to be a big problem, using the inflection point method you don't need to calibrate the probe at all, raw sensor data will work just as well as calibrated pH values.