Over-engineering a peristaltic ATO for basement reservoir

[Chee-tomorpha]

It's been 2-3 years since I acquired a Kamoer X2SR unit to help with water changes. I'd been procrastinating on setting it up, but about a month ago, I finally started the journey of drilling a decent-sized hole from the basement through the main level, right beside my RS 500 Peninsula.

While running the tubing for the AWC system, I realized how much I hated going up and down the stairs to fill the freshwater jug for my Tunze Osmolator ATO sitting beside it. I figured I should run another line and relocate that ugly 5-gallon plastic jug hidden behind the couch down to the basement mixing station.

Little did I know this idea would turn into a massive project. A month after drilling the hole, I finally have the AWC working, but my ATO system was struggling. Just prior to the drilling, my original ATO pump failed, so I purchased Tunze's new upgraded pump. I thought it would be more than capable of moving water from the basement floor to the main level.

You all probably know where this is going: the upgraded pump didn't have the power to push the fresh water to my main level, and that's because it had to run through 50ft of horizontal run and 10ft of vertical head.

After some research, I found a few people in the community talking about swapping their standard ATO pumps for a peristaltic pump. My AWC system was already using them to push the same path. Since peristaltic pumps create far more pressure, I figured it was worth a shot. But once I got my first prototype built, I got a little too excited and ended up completely over-engineering the heck out of it, stopping just short of adding a microprocessor.

I am now at Version 4 of my over-engineered peristaltic ATO engine, designed to be retrofitted into almost any existing ATO system. Instead of a simple 3D printed box with a motor wired directly to my Tunze, my current build features:

1. Digital Trigger Logic Board: This isolates the motor's power draw from the ATO unit. It allows the pump to be triggered by any ATO system without risking frying the ATO's brain, because the motor pulls from its own dedicated 12V power source.
2. PWM (Pulse Width Modulation) Speed Control: A dial to seamlessly increase or decrease the motor's RPM to adjust the exact water flow rate.
3. A 3-Way Selector Switch: Expanding the unit's usability with different modes:
   • Operation I (Auto): The default mode that waits for a trigger signal from your ATO's brain to turn the pump on.
   • Operation II (Override/Smart): An "Always On" mode. This is perfect for manually pumping water between containers, or—since the outlet controls its on/off state—plugging it into a WiFi smart plug for app control.
   • (Upcoming V5 Feature): I'm adding a flush-mounted port to allow a standard, "dumb" float switch to act as the trigger. Because the internal board isolates the power, the micro-current won't burn or melt the float switch contacts.

Right now, I'm just waiting on a couple final parts to arrive in the mail to complete the Version 5 vision. It will remove the odd DC barral jack to be flush mount, the trigger wires will also get flush mounted as an audio jack, then the third port for the dumb float switch feature.

Since I had to buy a lot of these internal components in bulk packs of 5 or 10, I’ll probably have enough leftover materials to assemble 3 or 4 extra complete V5 units once I’m done - I will have to source the motor though. If anyone is interested in grabbing one of the extras to test out on their own basement runs, let me know!

Anyways, has anyone else gone down this rabbit hole for a basement relocation? Are there any specific failsafes or features you'd want to see on a box like this?



Here are some pictures of the v4 assembly before the testing tonight, and some modeling and 3D printed prototyping parts






Here's a quick video of the v4 test run using the Operation II (always on mode) and showing the PWM to control speed

Here it is in Operation I, the default where the Tunze brain controls the pump.

 

[Mebbid]

I've been considering doing something similar tbh, which tunze pump did you try that didn't work?

Are you planning on putting the peristaltic pump near the tank pulling water upwards or near the reservoir pushing water? From my understanding, the pushing is far more efficient than pulling and you might run into issues with the pump struggling to draw the water up.

 

[Chee-tomorpha]

I originally bought the new Tunze Turbelle High Jet. I thought it could handle the basement run, but since it couldn't, I'm keeping my custom unit downstairs right next to my 50-gallon freshwater barrel.

I've got it set up to push the water all the way up to the main floor. I've had it running a few times throughout the development phase, and it pushes the water like a champ.

I haven't tried setting the unit upstairs to pull the water up from the basement. Even though peristaltic pumps are self-priming and technically could do it, I have a feeling the vacuum created by that much vertical head and the 50ft run would just suck the soft silicone adapter tubing flat and collapse it.

Pushing from the basement just seemed like the much safer, more reliable route for this application!

If someone did want to try a 'pull' setup upstairs, the trick would be reducing that small section of soft silicone tubing—the part that converts the pump barb to the RO quick-connect—to be as minimal as possible. You'd basically want to butt the rigid quick-connect fitting right up against the pump's barb, so the silicone just acts as a tight shrink-wrap sleeve over the two. With zero gap of unsupported silicone in the middle, it physically can't collapse under the vacuum.

 

[gastone]

I'm having the same issue. RODI unit in the basement, ato reservoir on the main level. Tired of carrying buckets. How is this coming? Are you going to have extra units? I'm interested.

Garrett.

 

[Chee-tomorpha]

I'm having the same issue. RODI unit in the basement, ato reservoir on the main level. Tired of carrying buckets. How is this coming? Are you going to have extra units? I'm interested.

Garrett.

Sorry for the delayed response. I've been out of the country on vacation and am just getting back this week.

The good news is that the unit was set up a few days before I left and ran flawlessly the entire time I was away. Zero issues, which was a great stress test.

I will have some extra parts on hand, but I will need to source the actual pump based on your specific needs. My V4 components should be waiting for me at the house. I'm also actively working out the details for V5—which will operate as a fully standalone ATO system with dual floats (high and low)—but that version will require a few additional parts and some more testing before it's ready.

To help me figure out the right pump size for you: How big is your total system, and what is the exact rise and run from your freshwater reservoir to the display? Also, do you already run a controller on your tank, or are you looking for a fully standalone setup?

I'll post an update with the final V4 build and V5 design details as soon as I get settled!

 

[gastone]

Sorry for the delayed response. I've been out of the country on vacation and am just getting back this week.

The good news is that the unit was set up a few days before I left and ran flawlessly the entire time I was away. Zero issues, which was a great stress test.

I will have some extra parts on hand, but I will need to source the actual pump based on your specific needs. My V4 components should be waiting for me at the house. I'm also actively working out the details for V5—which will operate as a fully standalone ATO system with dual floats (high and low)—but that version will require a few additional parts and some more testing before it's ready.

To help me figure out the right pump size for you: How big is your total system, and what is the exact rise and run from your freshwater reservoir to the display? Also, do you already run a controller on your tank, or are you looking for a fully standalone setup?

I'll post an update with the final V4 build and V5 design details as soon as I get settled!

Glad to hear everything worked well. I don't utilize a controller currently. I'm looking at 10' or head height and about 20' of run. Standalone would be nice, but I've been exploring a controller. I'd go with either hydros or ghl. Have an apex at the school where I teach and I'm not completely satisfied. Appreciate the response and I'm looking forward to what you've come up with.


Garrett.

 

[Chee-tomorpha]

Update: V4 is Built, Tested, and Survived the Ultimate Stress Test!

Just wanted to share a quick update with some photos of the completed V4 build.

I tested the prototype for a few days right after I finished putting it together in my last update, and then left for a spring break cruise with my family for a week and a half. I had no access to WiFi or 5G and couldn't check my tank except for a couple of days when we were in port.

Leaving the unit completely unattended for that long is terrifying, especially with no access to keep an eye on things. But it ran flawlessly the entire time. No leaks, no failed triggers, and my tank levels stayed perfectly locked in.

I got my DC barrel plugs in and completed the V4 build. Here are some shots of the final V4 enclosure and the internals:




For those following along, here is what this little box is currently packing to replace my stock Tunze pump for the basement run:

• The Muscle: A Kamoer peristaltic pump that effortlessly pushes water up the 10ft vertical head and across the 50ft horizontal run from my downstairs reservoir.
• Isolated Digital Logic: A dual high-power MOSFET board. This takes the 12v trigger signal from the Tunze Osmolator but isolates the power draw. It prevents the stronger peristaltic motor from frying the Tunze brain. (This can also be used with almost any trigger signal, turning a cheap Amazon ATO unit into an absolute powerhouse).
• Adjustable Flow: Integrated a Mini PWM controller on the side so I can dial the flow rate up or down perfectly.
• 3-Way Operation Switch:
  • Position 1: Standard ATO mode (waits for the Tunze trigger).
  • Position 2: Off.
  • Position 3: Manual Override (Always On). This is great for manual water transfers, or plugging the dedicated 3-amp power supply into a smart plug for WiFi control.

I managed to get the CAD files smoothed out, so the 3D-printed enclosure is looks and feels clean with the flush-mounted barrel jacks on the back panel.

What's Next (V5):Now that V4 is proven and reliable, I’m digging into how to add standalone functionality using a microcontroller. The goal for V5 is to make the unit a completely standalone system using direct inputs for high/low dumb float switches, completely bypassing the need for a 3rd party ATO controller altogether.

Let me know what you guys think of the wiring and the enclosure. Any feedback on the current design is welcomed.

Also, any experiences with microcontrollers would be awesome too. I've done a little bit of researching and am thinking about using an ESP32-S3—which has WiFi, Bluetooth, Matter, etc. (rabbit hole!)

 

[Chee-tomorpha]

Glad to hear everything worked well. I don't utilize a controller currently. I'm looking at 10' or head height and about 20' of run. Standalone would be nice, but I've been exploring a controller. I'd go with either hydros or ghl. Have an apex at the school where I teach and I'm not completely satisfied. Appreciate the response and I'm looking forward to what you've come up with.


Garrett.

It's going to take me a bit to get the standalone feature locked in, I think. I fell down the microcontroller rabbit hole when looking up how to work with a two-float ATO mechanically. It turns out I really need a smart controller to handle the logic—mostly to ensure it has a timeout fail-safe so it doesn't flood the floor if a mechanical float switch gets stuck.

Since you mentioned Hydros, that is actually what I run on my tank. I currently use my Hydros as a fail-safe to cut power to my Tunze Osmolator if the water hits the Hydros high-level sensor.

If you need a solution right now for that 10' head and 20' run, this current V4 version can work with almost any trigger. You could literally grab a simple $20-$30 generic ATO unit with a weak pump off Amazon, use its brain for the water sensing, and plug it into this box to turn it into an absolute powerhouse.

 

[PPBlimpy]

That’s one way to do it. Love the DIY!

I had a couple different ideas- little less DIY
First was using an old JBJ ATO brain and switches that I am currently using but running a long extension cord into the basement to an adjustable peristaltic pump I picked up off Amazon.


The other idea is I am going to try to use a second dos and eb832 on my apex and a long custom aquabus cable. One side will be for my ATO and the other will be used to Dose Kalk from my 50g storage drum that I will relocate to the basement instead of next to my tank.

I am probably going option 2 as I just picked up another eb832. Already have the extra dos. So now I can power and monitor my mixing station and can expand further later to ad a salinity probe and temp probe

 

[gastone]

It's going to take me a bit to get the standalone feature locked in, I think. I fell down the microcontroller rabbit hole when looking up how to work with a two-float ATO mechanically. It turns out I really need a smart controller to handle the logic—mostly to ensure it has a timeout fail-safe so it doesn't flood the floor if a mechanical float switch gets stuck.

Since you mentioned Hydros, that is actually what I run on my tank. I currently use my Hydros as a fail-safe to cut power to my Tunze Osmolator if the water hits the Hydros high-level sensor.

If you need a solution right now for that 10' head and 20' run, this current V4 version can work with almost any trigger. You could literally grab a simple $20-$30 generic ATO unit with a weak pump off Amazon, use its brain for the water sensing, and plug it into this box to turn it into an absolute powerhouse.

I do have an osmomolator currently that I use for my ATO. I also have about 10 other atos laying around. Mostly optical sensor types. I think currently I've got 10 nano tanks up and running, down from 15 or 16.

 

[SantaMonica Verified Sponsor]

Get 10 or 20 built and put them up for donation purchases, and see what the sweet spot price could be.