A while ago it occurred to me that I could build an inline carbonator much like Blichmann Engineering’s QuickCarb™ and I started to figure out the parts list, which I posted here. Shortly after that post I purchased all the parts and eagerly await the slow boat from China to deliver my new project. A few agonisingly long weeks later it had all finally trickled into my mailbox and I was ready to get my build on.
There isn’t actually a whole lot to this build, most of it was figuring out how to get things connected right, like using the carbonator cap with a barb on it to allow me to easily connect the gas line. I took pictures of the steps anyway so here they are:
Step 1 – Assemble the Carbonator Tee
First install the diffusion stone in the tee. It is important not to touch the stone with your bare hands as apparently the oils from your skin can clog the tiny holes. As I always do, I liberally apply pipe thread tape.
Step 2 – Connect the Pump
I used a short length of tubing to connect the pump to the liquid out ball lock connector. This will connect directly to the keg and should be as short as possible as it serves no purpose to have extra line length here.
Use another short length of tubing to connect the output of the pump to the carbonator tee. Again extra length here serves no purpose, so keep it short.
Finally attach a long length of tubing to the output of the carbonator tee. This piece of tubing is where the magic happens. The long length increases resistance for the moving beer and forces the CO2 into solution before being pushed back into the keg through the gas in post. I used about 7ft of tubing on the output.
Final Parts List
So you don’t have to go hunting for my parts list, here is the final parts list for what I used:
- DC 12V 70W Food Grade Diaphragm Pump
- 12V, 10A power supply
- Inline Diffusion Assembly: Stainless Tee + 2x 3/8″ Barb + Inline Carbonator
- Gas/Liquid QD Set, with Swivels and 3/8″ Barbs
- Carbonator Cap with 1/4″ Barb
- 8ft Thin Wall 3/8″ ID X 1/2″ OD racking hose
How Does It Work?
So the first time I tried to use it on a keg full of water things didn’t go smoothly. What follows is based on a mistaken belief as to how the pump worked.
I believed that what I saw the pump doing was part of some sort of priming procedure as it would crank once, suck up a little liquid, and then do it again. I expected that once it picked up enough liquid, it somehow sensed it and would start running normally. This happened the first time I ran it with a keg of just water, but it was terribly inconsistent in how it worked. I began to suspect a faulty pump.
With HomeBrew Con this week I was suddenly faced with needing my Club Night beer carbed and ready to be dropped off to our transport all in one afternoon. It was time to try the carbonator again. I set the pump up with some StarSan in a bucket, it did its little suck up some liquid a little at a time routine, and then when it got to the pump it started to run. That had to be how it worked, so I flush the StarSan out with CO2, hooked it up to the keg and low and behold it did the same thing with the beer and then it ran! My line carbonator was carbonating! Well, it ran for a few mins before I tried to adjust the position to close my keezer and it stopped and went back to moving a little liquid at a time.
After fully disassembling the pump to figure out if I could bypass this weird prime routine that didn’t seem necessary it dawned on me – this isn’t a smart pump with a sensor. The pump is dumb. I am dumb. What was happening was that the power supply I was using that should have been giving enough power to run the pump was not. Maybe the pump draws a little more than the rated 70W or maybe the power supply outputs a little less than the rated 72W, but either way the pump was not getting the power it needed.
A Properly Working Inline Carbonator
To get through needing the beer carbed ASAP to be ready to send to HomeBrew Con I figured out that I could remove the motor from the pump assembly, switch it on (it ran fine that way, which was how I figured out it was a power supply issue), and quickly shove it back into the pump assembly. This got over the initial large current draw where the motor had to work against the magnets of the pump and get it running smoothly. Since the motor drives a magnet that pushes and pulls the pump assembly I could do this without worrying about breaking the pump, at least a few times.
I picked up a more robust power supply from Amazon, 12V @ 10A instead of the original 12V @ 6A, and tested it today to find that the pump kicks on perfectly every time. Problem solved, carbonator working!
A Few Things to Adjust To
After running the carbonator for a couple of hours on the beer I poured a taster to see how it was going. Low and behold it was carbonated! It wasn’t, however, carbonated to the level I was hoping. I had set the PSI a few above what my serving pressure would be for the temperature the beer started at. What I didn’t account for was that the pump got warm, and after running for a couple of hours the beer had heated up a few degrees as well. I estimate it got up to about 45F, which is higher than the 38F I would normally carb and serve at. So while I wanted about 2.5 volumes I was probably more like 2.2 to 2.3 volumes.
The second time I ran it I put a fan in the keezer and the pump wasn’t feeling as hot. I think it would be good to add a fan if your pump gets warm during operation. I’m going to play with dialing it in, attempting to limit the temperature increase, and/or at least measure it so I can predict what PSI I need to set to properly carb the beer.
The very same beer I was attempting to carbonate, while good – you can taste it at HomeBrew Con if you find the AABG booth on Club Night – doesn’t have quite the bold flavour I was hoping to get out of the spruce in the mash. I am borrowing a randall from a club member to add a little extra spruce to the beer on-site, but then I had a thought: can’t I hook a randall up inline and run the pump to quickly add flavour from hops or other sources? That’s basically what Sierra Nevada’s torpedo is, right? I think this is well worth investigating in the future and if anyone has any experience doing this on the homebrew scale I’d love to hear from you.