Meet Ellie Weinstein: 3D chocolate printer

Eight years ago, a mechanical engineering student thought that 3D printing on chocolate would be easy… that was when Ellie Weinstein’s journey into edible 3D printing began. It turns out that it isn’t easy at all – but that’s led to a machine that’s capable of reliably heating chocolate to within 0.1 of a degree, is food safe, and produces designs that just aren’t possible with traditional methods.

A finely detailed chocolate print of the word LOVE in extruded serif capital letters
Creating this shape in a mould would be incredibly tricky; with 3D printing, it’s easy

We spoke to Ellie about her venture, Cocoa Press, the art of designing for chocolate, and what it takes to get a blend of sugar, fat, and cocoa solids to bend to your will.

 So, before we get into the technical stuff: why 3D printing with chocolate?

Ellie Weinstein:  Yeah, so I guess there are a few reasons why 3D printing and chocolate go very well together. The first is obviously that it’s delicious, and it’s really fun to get to mess around with a whole bunch of different types of chocolate. 

But the two reasons that actually make sense are customisation and being able to make things that are not possible with traditional chocolate-making. So, in terms of customisation, you can make one-off chocolate bars, you can do things that chocolate shops just wouldn’t do, because they’re not going to buy a custom mould for it. Or, you can just take your entire library of moulds (some bigger chocolate shops have a room full of moulds). Now that all fits on an SD card, so you don’t have to keep all that physical stock out or buy each one individually. Similar to the original iPod, you have your whole album library on your hard drive now. 

Ellie Weinstein beside her Cocoa Press printer at the Midwest RepRap Festival, Goshen, Indiana
Ellie Weinstein at the Midwest RepRap Festival, Goshen, Indiana

The second advantage is being able to make things that are not possible with traditional chocolate-making. And a lot of that comes down to really cool textures and infills. I literally just use the PrusaSlicer gyroid infill, as my favourite example of that, because gyroid tastes so good. It cracks in your mouth differently. 

You can also do 3D designs with crazy overhangs that would be too delicate to make with the general flexible silicone-type chocolate mould.

That was going to be my next question: what does a 3D printer do that you can’t do with a mould?

There’s a lot of lead time for a custom mould. And custom moulds can be expensive: $500 plus, if you’re just making one or two of them, and the customer might not want to pay $500. But if it can be, you know, $1 of chocolate, and an hour of CAD time, that’s way easier.

Dewsbury’s own Jean-Luc Picard, one of the first models printed on the current printer
Dewsbury’s own Jean-Luc Picard, one of the first models printed on the current printer

Does the printing process affect the taste? 

Not at all, as long as you don’t mess up the temperatures. If you mess up the temperature too badly, then you’re going to cause that bloom issue

Bloom. What’s that?

Bloom is when the fat or sugar separates out. When the cocoa butter separates out from the cocoa solids, it’ll get a white, powdery finish on the outside. 

It also might not fully solidify. If you’ve ever had a chocolate bar that was left in your car or something, you might take it out, it might still be fine, but it’s still a little bit bendy, or it’s not quite as solid as it was before. Or, like if you snap it, it doesn’t have that nice, crisp snapping sound – it’s more of a bend and break. And that means the chocolate is now out of temper. 

It’s still perfectly safe to eat, but it just changes the texture, the mouthfeel, and sometimes the taste, if it’s a more extreme bloom. But no, the printing process doesn’t change it, as long as you’re within normal parameters.

And how do you control those parameters? How do you make sure that the chocolate doesn’t just melt as soon as it comes out onto the print bed?

Very precise temperature control is the answer. We have a multi-zone heating system. And we have an accuracy of a little high than about plus or minus 0.1 degree Celsius on the extruder. 

Cocoa Press has gone through loads of revisions; here’s the first version, made of wood
Cocoa Press has gone through loads of revisions; here’s the first version, made of wood

Chocolate comes out of our machine at just below body temperature; that compares with around 200 degrees Celsius for PLA printing.

I don’t know exactly how accurate the temperature of a Prusa extruder is, for example, but I guess it’s to within at least two degrees. I do know that on Prusas, and anything running the Marlin 3D printing firmware, generally, temperature is stored as an integer, so you literally can’t set it to any decimal point. That makes our temperature control at least ten times more accurate, in terms of what you can set it to.  

Is that as a software control thing, or hardware? And is that what makes Cocoa Press Cocoa Press?

It’s partially on the software side, partially on the custom electronics that we have. And then, partially, just on good mechanical design to retain heat in our extruder, and to control two separate zones: one right near the nozzle, and one that’s the main body part. 

This is a silly question, but I guess the chocolate isn’t on a spool, like PLA?

It is not. I can never come up with a good analogy, but it’s like a big cigar shape. And so, all of the chocolate for the print is heated up at once. And then we push on it with, I think, about 40 pounds of force. 

We’re heating it up to just below body temperature, the lowest temperature where it is somewhat liquid to extrude it, and then it’s able to solidify immediately after extruding. Or at least within a few seconds. 

Nine square milk chocolate prints with beautiful abstract patterns

What kind of chocolates do you use? Is there a specific blend that you’ve made for printing with?

This is where we’re starting to change what we’re doing. In the past, we’ve had our $10,000 printer, that is fully enclosed, and had active cooling in it too. So, not only were we heating multiple zones, we were also cooling the build chamber. And, with that, you could use a wide array of chocolate: we used Belgian chocolate, I used to use Callebaut, then I moved to Guittard, because they’re US-based and it was easier to get it. At one point I was using whatever I could get because, like everything else, chocolate has had supply shortages as well.

It’s crazy. I’ve been calling up chocolate suppliers… minimum order quantities have gone way up, and sometimes chocolate is just sold out. 

On our new version of the printer, it’s just an early prototype that we showed off at Midwest RepRap Festival [in June], we have our own special blend which solidifies at 80 degrees Fahrenheit [26.7 °C] which is, you know, warm for a room.

We’re not going to lock anyone out of using whatever chocolate – if they have a cold room, they can stick the printer in the room. But the cost of this new printer is going to be approximately a fifth of the cost of the other one. 

The main downside is not having that active cooling on board. But, with this new chocolate, it’s going to work. It’s just an early prototype, but the initial tests are incredible. 

I’m not a big chocolate expert, but I do know that a lot of UK chocolate can’t be labelled as chocolate in the EU because there’s so little cocoa solid to it; I’ve tried some American chocolate and it tastes completely different again. How do different chocolate blends print?

The sugar content doesn’t seem to affect it that much; it’s really the cocoa butter percentage or cocoa solid percentage. And palm kernel oil versus cocoa butter being the main fat in it. I also think it’s just a shift in the larger chocolate industry; I’ve been doing this now for seven or eight years. And I did not come at this from a chocolate background, so I’m learning like everyone else. But I think there’s a shift in the chocolate industry to use more palm kernel oil, as opposed to cocoa butter. And so, it’s just been interesting material science, food science happening to keep the taste really good while switching the ingredients. 

You mentioned infills a while ago. Does that influence how you design the chocolates?

I don’t know if it influences how I design the files because I can just take any STL that I can print on a plastic printer and use it. 

Overhangs are a little harder, as you might imagine. So I generally like to stay in the 30–40%, infill range… with a 10% infill, sometimes the bridging is a little too far. And, you know, I can go higher on percentage infill, but I like prints to take less time and use less chocolate. Also, because we’re limited by the amount of chocolate that’s in a cartridge, I like to see if I can maximise the size of the object that I’m printing with one cartridge.

What sort of nozzle size do you have to use? I guess you can’t go as fine as you would with hot plastic.

Yeah, so we use a point eight nozzle. I want to experiment with more at some point, but it’s not the most important thing.

That’s still smaller than I imagined.

Yeah, and I have some worries. We have played around very briefly with a point four and had some more clogging issues. I have worries that going bigger will cause it to take longer to cool, and could actually give us worse results. It’s just a theory – I’ve never tested it. But it’s something I want to play with for sure. Because it could also let us go faster. Who knows?

Cocoa Press chocolate printer

What made you want to print in chocolate?

I kind of stumbled into it. It was a project in Intro to Engineering class [Ellie has a background in mechanical engineering]. And I was just, like, I wanted to build a 3D printer. And that project got denied because they said, ‘Hey, look, we have a 3D printer right there – do something unique!’

I was inspired partially by the PancakeBot printer to try printing food. And I thought chocolate would be easy to work with…so it’s been about seven and a half years, and I’m still realising how difficult chocolate is to work with.

But the other side of it is that I started bringing it to events, New York Maker Faire, Philly Maker-Meetup, etc. And people were asking how to buy it. And so I started having more serious conversations, and realised that there might actually be a market for it. 

What is the market for it? Is it 3D printing people? Or, is it to chocolate people?

So far, chocolate shops are mainly who I’m talking to. With this new printer, we’ll see. That will be pitched a little more towards makers, but our high-end one is still towards chocolate shops. 

The difference is mainly: are you trying to teach the chocolate side, or are you trying to teach the 3D printing side? Because, most chocolate shops don’t have any experience with 3D printing. And I find the chocolate side easier to teach because I had to learn the chocolate side myself, and I just think we have a good system that makes the chocolate side way easier – you don’t have to temper it yourself, it comes in pre-tempered cartridges etc.  

So, we’ll see whether this new machine reaches new customers, but, in the past, it’s been chocolate shops.

Finely detailed strawberry-shaped print in white and pink chocolate

What else have I missed – temperature of the print bed, the nozzle, the pre-extrusion heating – all these are different from printing with plastic. Is there anything else that you’ve had to take into consideration?

I’m extremely happy with the general food safety of the whole system. The chocolate only touches four things in total: the cartridge body; the nozzle, which unscrews so it’s easy to clean; there’s one plunger piece on the top that acts as a seal, and then we have our silicone baking mat on the bottom, which is our print surface. All of those can be removed without any screws, and washed in the sink or whatever. So, it makes it extremely easy to clean and keeps the whole thing food-safe. 

That sounds like very much a mechanical engineer’s system.

Definitely, definitely. But that’s my background. 

We have a printer set up in the other room here, which is a certified food establishment space, so we can legally sell prints out of it, or do our chocolate. 

We still test in this room, but for our actual setups, and for all of our Christmas orders, we’re doing them in the food-certified room. (And that certification is specifically for the room, not for the device itself.)

There are two of us full-time at the moment – it’s gone up and down in the past. We’ve been selling the chocolate, and now we’re getting back to selling the printers. And that’s why I’m excited about this new version – it probably won’t be out until late 2022, early 2023, but I’m excited about it, because we’re gonna get back to selling the hardware. And I feel like it’s not every day that we can cut the cost by, you know, 75% and keep a similar print quality. So I’m just really, really excited about this new version.

HackSpace magazine issue 58 out NOW!

If you liked this interview, check out more in the latest issue of HackSpace magazine.

Each month, HackSpace magazine brings you the best projects, tips, tricks and tutorials from the makersphere. You can get HackSpace from the Raspberry Pi Press online store or your local newsagents.

HackSpace 58 cover

As always, every issue is free to download in PDF format from the HackSpace magazine website.


James avatar

My wife would be smiling too and perform a couple of back flips.

Vaneesa avatar

ummmm, nice

Comments are closed