Swipe to the left

The Story of Focus Go Flow

By Mike Oxley January 5, 2016 4 comments

The story of Go Flow starts about two months ago. We were making our preparations for CES 2016 where we planned to launch Quantum8 EEG. The problem was our newest prototypes were just not good enough. So we decided we needed a Plan B for CES.

Around the same time, there was some discussion in the tDCS community around the reliability of lower cost DIY tDCS devices. Many people had discovered that the devices were not delivering accurate current levels. This meant many people who thought they were trying tDCS, were not really getting tDCS.

Plan B: Build an accurate reliable, lower cost tDCS device and launch it at CES.

We spent the next week working out the details. Choosing a standard 9v battery as the power supply was obvious. This removed the need for a recharge circuit but allowed us to design the step-up converter. Electrode resistances vary and we felt 25V was a good maximum voltage to implement.

Most studies use current levels between 1.5mA and 2.0mA but the DIY devices often offer setting as low as 0.5mA. So we decided we should also offer between 0.5mA and 2.0mA.

Focus devices have always included bluetooth and other modes, but we wanted to keep the cost and support issues down, so these were left out.

Safety is critical so we included dual current protection features - both software and hardware control. Ramp up and ramp down also seemed important as it makes stimulation more comfortable.

Finally we knew a timer was an essential safety feature that had to be included.

Go Flow was the name of an early DIY tDCS project that didn't quite reach production. We always loved the name and thought that the spirit of that earlier project was reflected in these requirements - so Go Flow became the working project title.

From these requirements we set to work on R&D and electrical design. The electrical part was quite straightforward as we have the experience from four previous brain stimulator designs to build upon. The main constraint was designing the Bill of Materials to only include parts we could source in time to manufacture before CES.

R&D was a green field project which is always fun. The key early decision was to use a three way rocker switch. This influenced the physical design as well as laying out the user interface and user experience.

Rocker switch shown in top left of PCB.

Once we had PCB designs we flew to China to assemble first prototypes. Physically being in China makes it much faster to turn around changes and indeed we did three PCB revisions in the first trip to make the electrical fit the mechanical.

To mass produce plastic products requires a steel mould be tooled. This generally takes 4 weeks from design sign-off to t0 parts. So we left China and tooling began. It was going to be tight, but if no problems we could have a product ready for CES.

Go Flow Mould - family tool making all plastic parts in one shot

Once tooling has started a quiet calm emerges. Its now impracticable to make any changes, so all you can do is wait and hope what you designed is going to work. Our uncertainties were around battery connection, switch hat usage and LED appearance.

With the BoM now fixed, tooling costs paid and most elements decided it was time to make a decision on pricing. We posted our plan on Reddit and asked the community for pricing feedback. Suggestions ranged from $49 to $199 but the consensus seemed to be around $99. Internally we had a lot of discussion around price point and couldn’t come to a final decision.

Go Flow PCBs Mass production.

It was time for our next trip to China to test the final PCBs and also review the t0 plastic parts. Nothing ever runs smoothly and we had two problems. Firstly, there had been a misunderstanding about who was sourcing the battery connection parts - so we had none. Secondly we found some of the Chinese components were fakes. Fake components can be a big problem in China so its important to always test everything. We quickly found and replaced the two fake chips as well as buying the battery connector parts.

Missing metal parts, no longer missing.

Plastic mould tools are not designed to produce perfect parts first time. It is always a process of imrpoving the tools, correcting tolerances and general fine tuning. focus are extermely lucky to work with one of the best tool makers in China. This project was delivered from nothng to production parts in only 20 days. Incredible.

Seemingly wanting to outdo themselves, they then spun up a production line literally the next day. Whilst the UK enjoyed Christmas dinner and boxing day leftovers, our Chinese factory mass produced our first batch of focus Go Flow.

Go Flow Ultrasonic Welding - Final step of mass production process.

Every device was tested three times. First test was performed during flashing of the IC. This is when the firmware is loaded and locked to the device. Then a full circuit test was performed to ensure output levels are within tolerated ranges using a factory mode activated by the test jig. Finally a post assembly check was performed.

Go Flow Test Jig

Devices were then packed for shipping to the UK where they will be tested again prior to boxing. The remaining parts of the Go Flow package such as box, cable and hydrogel pads are being manufactured by our existing suppliers.


Go Flow Stimulators Packed for Shipping to UK.

So finally the price. $99.95 is probably the right price considering the quality and value delivered by Focus Go Flow.

We've had a lot of fun making Go Flow and we hope you like it as much as we do.

John Humphrey January 6, 2016 at 5:57 AM
Just... Wow!
Rita Carter January 6, 2016 at 7:55 AM
GoFlow should be a game-changer - congratulations, foc.us!
Respy January 6, 2016 at 2:10 PM
But how long can you stay in business overestimating the market size/volume that is required to cover all costs?
George Phillips February 11, 2016 at 10:48 PM
To Respy:
I'm not a business analyst so please forgive my ignorance, but in terms of interest in the scientific community TDCS continues to rise in number of and variety of research projects.
The number of doctors and psychologists participating in the research, or investigating it is rising as well. I can provide you with figures but I encourage you to investigate this.
Not counting the "bio-hack" crowd or individual enthusiasts, (I want to carefully note that anyone doing this research should educate themselves very carefully, thoroughly, and constantly). I think Foc.us will more probably have issues with keeping up with demand rather than overloading stock.
Other than early design issues that were superbly acted upon and improved, Foc.us has brought to the table a well thought out product both from a professional and educational viewpoint while keeping the original idea of providing a safe and affordable product to people. Other firms (not ALL firms) "jumping on the bandwagon" continue to sell very expensive produducts that come nowhere near the flexibility that Foc.us has. I am not paid by Foc.us or rewarded by them for writing. I have done almost three years of research in TDCS and am continuing my analysis and reviews of research, so I have a little bit of experience in it.