Open Source Hardware TDCS
I’ve been having a lot of fun experimenting with Transcranial Direct Current Stimulation (TDCS) in the last two weeks. There is a lot of information available on the Internet about TDCS. There are many hobbyist circuit diagrams available (e.g. 1 2 3) and at least two commercial designs that have been published by Focus Headband v1 and GoFlow.
There’s not much missing from the hobbyist TDCS discussion, except perhaps collaboration and a bit of polish. I am not particularly qualified to help with either, but I’ll try anyway.
TDCS: Brain Zapping for Fun and Profit
I’ve created a simple TDCS circuit diagram around the LM334Z. The initial design is compact, features trimmer control of current from 0.5mA to 2mA and easy electrode connections using standard pin connectors. The pin connector also makes an easy spot to connect your ammeter during testing.
You’ll notice that it is very similar to the GoFlow schematic linked above. I’ve also provided a pcb layout:
Practically speaking, this means that anyone interested in tdcs may (at their own risk): Download the designs, edit or modify them, fork their own design, and sell a product based on this design all without any proprietary (expensive) software.
Those who aren’t inclined to make changes, but want to experiment, can use the pcb design and the bill of materials to make small batches of devices for testing or personal use.
Want to learn about circuit design or make changes to the board?
If you’ve never used gEDA before, start with this simple tutorial. Once you’re familiar with the gschem and pcb UI feel free to clone the repo in github with the fork button or clone the repo using the CLI:
git clone https://github.com/nocko/tdcs
Once you’ve got a copy, go to town. There will be a bit more development information in the README file in the repo (when I get around to it). If you make changes, send them to me! I’d be interested to see what people are working on and if your are good, I’ll merge them and credit you.
Want to build a tdcs pcb for home experimentation?
Great! The gerber files are included in the repo under the gerber directory. You can compress this directory into a zipfile and upload it to oshpark and they’ll send you three boards for $7.55 (or you can send the gerber files to your favorite pcb prototyping firm).
Once you have the boards ordered, buy the components listed in the bill of materials. The per unit price is about $10 (including the board).
To actually use the device you will need some accessories most of which could be improvised (warning: affiliate links):
- Pin Connectors:
- To connect the electrode cables to the pcb. You could just solder, as an alternative.
- Banana Plugs:
- To connect the electrode cable to the electrode (if using the commercial electrodes above). Soldering is a cheap alternative.
- Ordinary speaker cable or lamp cord. I bought mine from the hardware store.
- Power Supply:
- To power the unit via the built-in barrel connector. As an alternative, you could connect a 9V battery here for currents <= 1mA.
Is this dangerous?
You should always be worried about passing electric current through your brain. However, it’s a primary design goal of this project to keep it safe.
Testing in rats established that tthe current density needed to cause brain lesions in rats was 142.9 A/m^2 for duration greater than 10min (Liebetanz et al., 2009).
In the initial design, even if the current regulator fails short circuit (allowing the maximum current to flow), the current through the brain is 3mA. The electrodes described above are 25cm^2 making the maximum current density 0.8A/m^2 nearly 180 times less than that which causes damage in rats.
That being said, I wouldn’t recommend that anyone attempt experimentation without first making themselves comfortable with the literature on TDCS and the schematics provided. Under no circumstances should you interpret this information as a recommendation to do anything.
Is this real?
Mostly. The design has been prototyped on a bread board, and regulates current well across range of the trimmer pot (0.5mA - 2.0mA) at input voltages greater than 10V for load (head + electrode) resistances of 5k (~1k greater than my brain+electrode setup).
I haven’t received the initial batch of PCBs from oshpark yet, the pcb graphic above is the rendering from their site. So there is a chance that the pcb will need further modifications.
Expect a follow-up with pictures of real devices in a few weeks with descriptions of tests on the actual design. If everything looks good (and there is any interest), I may setup a way to order boards directly.
Where to go in the future?
I plan to support bugfixes to this initial design and perhaps limited enhancement focusing on safety and reducing cost.
The next step is to work on a fork to add a microcontroller. Ideally, we’d have software control of the current, automatic shutoff and logging to a computer.
A digital ammeter display or other visual feedback system would make the design a lot more user friendly and a few bells and whistles for researchers (sham mode additional anode channels) are on the road map.
I am also interested in collaborating on a 3d printable case.