Simple tDCS Device: a Start

Disclaimer: don’t build one of these, but if you do, this design is probably better than everything thus far on the internet, but maybe not. Ultimately I am not responsible for anything you do or build.

Transcranial direct current stimulation (tDCS) is a technology used to treat a variety of ailments, especially anxiety and depression. Read more about it on wikipedia or in one of the 1000+ studies investigating the technology. This article showcases a DIY tDCS device that I have built.

The implementation is so absolutely simple. The only choice was picking the LM334 current regulator, since after that, TI nicely provides you with a schematic to follow. This is literally the most simple device that one can create. I added an LED just so I could stand out a little bit.

Why this circuit?

Despite tDCS devices be so simple, I have found a few devices I am not fond of online. I would not recommend any circuit that uses the LM317 for any current regulation since the minimum recommended output current of that IC is 10mA, about 5-10x higher than our design specification. This is why the LM334 is the better option here. Also, many point out the temperature dependence of the LM334, which other online designs do not take measures to prevent. But this can be easily rectified with the design on figure 15 of TI’s data sheet, which is pictured below. Though, even without this precaution, the change in current is only a predicted 7uA / K , or less than 1%/K, so it isn’t really a huge deal anyways. Just keep your tDCS device away from the fireplace.

Image from Texas Instruments “LM134/LM234/LM334 3-Terminal Adjustable Current Sources”

I also decided to use a DC-DC voltage converter for my circuit, but this is really not necessary if you have a fresh 9v battery. For the F3 -> FP2 montage and my DIY sponge electrodes (below), I have observed that only 5-7 volts are required to maintain a nice 2mA. The DC-DC regulator might be required for higher impedance montages or combination tACS+tDCS in lieu of an extra 9v battery.

Figure 15 in breadboard form. R2 and R1 values were generated through resistors in series.


There are two downsides to this circuit. Firstly, because the current level is dependent on two resistor values, it is more slightly more cumbersome to design a circuit with a 1mA/2mA switch. I wasn’t interested in doing this, however it could be done with some resistor/transistor/switch fun. If I wanted a whole spectrum of currents, I might as well design something with tACS capabilities (spoiler alert).

Secondly (and this is actually a downside), because the LM334 really wants to push that 2mA out there, when the power is initially connected or electrodes are reapplied, one gets a fun shock. Simple solutions include a series potentiometer to slowly ramp up the current, or a very large inductor which will provide a nice L/R time constant. None of these are ideal since one is a bit impossible and the other requires me to manually move a knob every time I want to adjust my headgear. A well designed circuit would be able to sense this huge voltage spike and limit it accordingly.

Lastly, some sort of voltmeter would be nice. I found it quite useful to observe the electrode voltage so that one can infer the quality of their electrode placements.