What I Learned Selling the 3D Printed Soda Adapter for Six Months on Tindie

| Comments

Putting the SodaStream Terra adapters up for sale in my Tindie store was an accident. My friend Alex designed the adapter. He got busy with real life and didn’t want to deal with the hassle of selling them on Etsy any longer, so he asked if we would like to take over.

Chris had just started setting up her Etsy store the week before, and she only had one item for sale. The timing seemed good, and he was selling one or two adapters every day. It seemed like a good way to get some initial sales onto her store, so we took on the task of printing and selling 3D printed soda adapters.

There was some lag between Alex running out of stock and us adding the item to Chris’s store, so there were immediately a bunch of orders. Chris paid for labels and shipped those out, then more orders came in, and she paid for labels and shipped those out.

Then Etsy closed her store. Etsy didn’t say why. Etsy didn’t respond to her emails. The store is gone, and Chris never got paid for the inventory she shipped out. It was quite a bummer.

So we dropped the item on my existing Tindie store.

tl;dr I just want a SodaStream Terra adapter!

I am no longer selling the adapters. As has been the case for most of the time the adapters have been in my store, you can download the 3D model of the soda adapter from printables.com and make your own.

The harder part is acquiring the rubber o-rings. They’re easy to get in quantities of 100 or 200, and they’re easy to get in assortments of hundreds of o-rings. The trouble we’ve had with the assortments is that not all assortments are measured the same way!

I have a whole mess of o-rings left over. You can find the correct o-rings in my Tindie store.

In my opinion, you should skip the 3D printed adapter and get the metal soda adapter from Amazon. I’ve been recommending this in my Tindie store since the product became available. It is a much more robust solution!

Why was I hesitant to sell the adapter?

Alex called me up one day and explained that he wanted to use his 3D printer to make an adapter to connect the old-style SodaStream CO2 canisters to the SodaStream Terra. I told him it was a bad idea, and that it couldn’t be done.

We drove to Target, bought a SodaStream Terra, and got to measuring. We had a basic part designed and printed in a couple of hours. It didn’t work, because SodaStream designed the new fitting to be difficult to connect to. Even though the adapter worked for Alex, I suspect SodaStream’s purposefully convoluted engineering has been trouble for our some of our customers.

It took him a few iterations to get the air directed to the correct places, but he did get it working.

Just because it was working doesn’t mean it is a good solution. I’ve been designing 3D printed parts for eight years. I know that 3D prints are weakest along their layer lines. I know PLA and PLA+ aren’t the ideal material to stand up to this sort of pressure.

Seeing it work and hearing that his customers were excited about using their adapters helped ease my concerns here.

There’s also the fact that SodaStream made it difficult to adapter their connector on purpose. I could write 2,000 about this part alone!

The failure rate is just too high

I sold 240 adapters over roughly six months. I’ve issued refunds or send replacements for around 30 orders. Why are they failing so often? Let’s start with the problems that may qualify as user error.

More than a few people have managed to cross-thread the adapter. If you are at all mechanically inclined, it is really obvious that this is about to happen. It is also pretty difficult to do accidentally, but if you do, most people are plenty strong enough to destroy PLA+ threads.

At least a few people seem to have trouble trimming the plunger to the correct length. It is a bummer that the plunger has to be trimmed to account for different bulk CO2 kits.

We suspect that many failures happen because the customer doesn’t screw the adapter on tight enough. If you don’t compress the large o-ring enough to make a good seal, CO2 can escape. Once the CO2 starts escaping, it has a much larger surface area to push up against.

This ties in with another problem. Some folks have most definitely managed to tighten the adapter way too much! The adapter is only 3D printed PLA+, so a person is definitely strong enough to break things. Especially if they put a wrench on it!

There’s no good way to document this for the average customer. Saying, “You have to tighten it enough, but don’t tighten it too much!” just isn’t terribly helpful.

There is also a good chance that some people’s SodaStreams are just built to slightly different tolerances than the machine Alex designed the adapter against. If the machining on Alex’s unit leaned towards the tighter side of the tolerances, then there’s a good chance that folks with machines leaning towards the looser side would have leaks.

Mitigating the weaknesses of 3D printing

At first, I was 3D printing with the default PrusaSlicer profiles just like Alex. As the failures came in, I started making tweaks.

Alex tried increasing the infill percentage, but that doesn’t make parts all that much stronger. I started by adding as many perimeters as would fit. Then I started slowly increasing the temperature and extrusion multipliers.

Hotter plastic tends to have better layer adhesion, at least up to a point, but it leads to stringier prints. I’d rather the adapters work than attempt to completely avoid stringing.

The slightly higher extrusion multiplier also helps keep gaps out of the layers, which helps with adhesion. I doubt either of these changes make a huge difference, but every inch counts!

The increased extrusion multiplier also has the side effect of the tolerances a bit tighter. That means the small o-ring is tighter in its slot, and the plunger pushes on it just a little harder. That ought to make it less likely to leak. The correct way to tighten up the tolerances would be editing the model, but that wasn’t really my goal. It was just a happy accident.

Why not try a different material?

This is where we get to the fundamental problem of Pat selling soda adapters.

I don’t drink soda. I am not a soda enthusiast. I am not excited about SodaStream machines. Printing with a very different material would require testing, tweaking, and more testing.

If this were my hobby, I would be diving right in. It isn’t my hobby, though, so I am just not excited about pushing the design into new materials.

Especially now that the all-metal soda machine adapter is available. There’s no beating that solution with plastic!

Since PLA+ works more than 80% of the time, I am confident that nylon would survive more than 99% of the time. Nylon is a pain to print with an FDM 3D printer. I sure don’t want to be doing that every day!

Expectation vs. reality

Most customers found my Tindie listing by way of Alex’s video about his adapter design. Alex’s video is pretty positive. He is proud of the work he did, as he should be, and he made those videos before significant number of people got adapters in their hands.

I’ve tried to keep a balanced description product description on Tindie. I don’t hide that there are failures. I made sure to point everyone towards the solid metal bulk CO2 adapter.

I believe most people understood what they were ordering. I think at least a few people were expecting some sort of unicorn to arrive in their mailbox.

Why continue to sell the plastic adapter when the metal adapter exists?!

I expected that I would be discontinuing the product as soon as the metal adapter was in stock. Surely everyone is using the 3D-printed adapter for bulk setups, right?!

Some people definitely continued to use the 3D-printed adapter for bulk CO2. A few people ordered adapters before messaging me to ask which bulk-CO2 kit they should buy! I told them they shouldn’t, and they should order the parts that match the all-metal adapter. If they told me that’s what they wanted to do, I refunded their money.

Most of my customers just want to be able to plug the SodaStream canisters from ALDI into their SodaStream Terra. They’re the reason I decided to keep on selling these adapters.


There aren’t any soda adapters in my Tindie store, but things are still chugging along. I am still cutting carbon-fiber ducts on the CNC pretty regularly, and I added a new carbon-fiber backpack hacking item to my store recently. I am pretty excited about those no-sew backpack straps, but I don’t have a good way to put them in front of the people who would want to use them. I don’t even have a good name for them!

I am sorry to see the extra revenue go. The extra money has actually made a real difference for us this year, but the ratio of happy to unhappy customers just isn’t high enough for me to feel comfortable. I am much happier selling over-engineered carbon-fiber doodads than plastic bits that have to stand up to 1,200 PSI!