From Gas to Electricity

This article is a bit technical, but in short; here is how we ended up with a fully electric sample roaster. Like Thomas Alva Edison, we had several tries until we achieved what we wanted.

Let's begin!

Prototype 1: Electric heating

Our first prototype back in 2014 was designed for electric heating. The primary heat source was a 3300W heating element designed to produce hot air. We also included twelve 50W heating cartridges in the outer drum wall for pre-heating. The purpose of the cartridges was to heat the machine faster, but we soon realised that they were unnecessary and that they just created a huge spagetti of wires.

The first prototype. 

The first prototype. 

Our main problem at this stage was the heat transfer from the heating element to the drum. Even though we had a very powerful fan, the response was very slow and we burned up a variety of heating elements in the search for the optimal setup. We didn’t find it.

Then, one Sunday, after Trond had been out camping and he brought with him a Primus camping gas to the workshop. We changed the direction of the heat inlet from horizontal to vertical and put the 2000W gas burner below.

After working all those hours with electricity, we were amazed by the sudden response. The instant heat was too much for the coffee beans, but we changed the rotation of the paddles and suddenly we actually had a decent manual sample roaster.

Satisfied with our improved roasting results, we went up to Morten Wennersgaard in Nordic Approach to get an expert opinion. There, we roasted about 10 samples of coffee and blind cupped them along with 4 roasts from their sample roaster.

Morten selected 2 of our roasts as the “winners” of the cupping. Needless to say, that gave us a boost to start producing the next gas-powered prototype.


Prototype 2 to 4: Fully gas-powered

In prototype 2 to 4, we used a 3kW pilot gas burner. Pilot means that it was designed to ignite larger gas burners, not to roast coffee. It was far from ideal for our use, but it worked.

For control, we used two Fuji PIDs and we logged through a programme called Artisan. That also worked. We were able to roast consistent, to a certain degree, and the coffee cupped OK.

Our main challenge in this period was mechanical. We knew how we wanted to roast the coffee beans, but we needed a reliable solution for dropping the beans.

Production drawings prototype 1 to 5

Production drawings prototype 1 to 5

Prototype 5:  Gas burner and custom electronics

For prototype 5, we made a custom gas burner and custom electronics. That changed everything, we got control, consistency, and the roasting was greatly improved.

This was also the prototype where we solved our mechanical problems and the first prototype with our award winning design. It was time to get out of Sverre´s basement!


Prototype 6: Testing and doing changes

With the good results from prototype 5, we were so enthusiastic that we started development of improved custom electronics and decided to make 10 units together with four pilot customers in Norway.

Getting a few of the prototypes up and running was done in a few months. Testing and doing changes for optimal control and consistency took a lot longer.

To keep the costs down, we tried to find our own ways of controlling the gas within the frames of the strict gas regulations. Not easy. Existing air-gas ratio control units would do the trick, but would increase the price too much.


From prototype 7 to 7x: A new concept with electricity

Prototype 7 was the most “prototype” of all the prototypes. It just consisted of the roasting chamber and a gas burner. It was made to test a few new engineering solutions. At this stage, most of our time was spent on improving the electronics, mainly on how to better control the gas burner.

ROEST sample roaster v0.1

ROEST sample roaster v0.1

In November 2017, after a few weeks of frustration with gas and the gas directive, Sverre decided to try a new concept with electricity. We still had plenty of parts from prototype 1 in the basement, and in three days prototype 7 was transformed into a fully electrical version called 7X.

With a new fan system, our previous problems with heat transfer in prototype 1 were solved.

By the end of the week, we were cupping identical to the gas-powered prototype and what really blew our minds was that we had increased responsiveness and the roaster was much easier to control. The power consumption went down to 1/3 of the consumption in prototype 1, so around 1100W.

This was so good that we started focusing 100 percent on electricity.  

Finally, we have the product that we wanted to create! Elegant design combined with optimal performance.