tDCS .. electric brain stimulation

this is an open-source tDCS .. Kickstarter rejected them, but they shared the design and schematic so we can build our own.


  • interesting! i'm wondering if the result is the same as nootropics.
  • What would you define as the result of nootropics in the intended context? Different interactions with different parts of the brain cause different effects on overall cognition. This holds true for anything that interfaces with the brain be it nootropics, sensory substitution, mental exercises, or even just learning new skills. So if this tDCS indeed has an effect (and it does look intriguing), it could be compared to nootropics as much as wiring your neurons to an artificial machine or studying a field of knowledge.

    Now, what I'd be more interested to know is--what is the resulting cognitive effects of stacking this *with* various nootropics? Could a synergy be found there which provides even greater gains in various factors of cognitive performance? Definitely something to tinker with.
  • The results I got from zapping myself with the prototype in that video are somewhat inconclusive.  It seems that the stimulation doesn't really induce the flow state, but just makes it easier to get into it.  Anyway, I have plans to actually, fully test this thing, with a control method and everything, so I'll let you know how that goes.
  • edited May 2012
    @Ian ok thanks...

    @cypan i use nootropics for better concentration for reading new knowledge and for skills improvement.. actually i use modafinil and piracetam
  • rdbrdb
    edited May 2012
    The linked circuit seems very underdesigned.  I can't believe they want to charge $99 for these devices.  I wouldn't put it on my head even if I got paid $99.  If the battery is removed, the capacitor could discharge and damage the LM334.  It has no failsafe current protection in case the current regulator dies - the 5 mA fuse isn't adequate, it would take too long for it to trip.  It has no ESD protection whatsoever.  It has very little control over the stimulation and has no feedback mechanism to ensure that the device is operating properly.

    Any solution that uses a 9V battery suffers from a problem: the impedance of the electrodes and head can easily be up to 5 kΩ, at which you can only reach a maximum of about 1.8 mA.

    I also don't think using PWM to ramp up the current (like ian described in the comment section of his YouTube video) is a good idea.  Unless the PWM is filtered appropriately, it would inherently result in an alternating current to go through the brain, which could have completely unanticipated effects.

    So, I sat down this morning and designed a new, more reliable and more robust circuit.  It uses commonly and cheaply available parts, and is driven by two 9V batteries.  It uses a microcontroller to program the amount of current; it uses a lowpassed PWM signal as DAC, which is fed into an op-amp circuit that converts the voltage into a current.  By adjusting the duty cycle in the microcontroller firmware, the output current can be adjusted - so the output current can be ramped up programmatically, and a timer could be responsible for automatically ramping it down after 20 minutes of use.
    To be sure, this follows up with a simple FET-based constant current limiter, just to be sure.  I put a varistor between each electrode and gnd to provide ESD protection (varistors are faster to act than diodes).  (I don't have a lot of experience with ESD protection circuitry, so other suggestions are appreciated.)

    Various ADCs on the microcontroller measure the impedance of the head, the actual current sent, and detect possible saturation of the op-amp (which is an indication that the electrode impedance is too high, for instance if the saline solution dries up.)  This could be indicated using LEDs, a 7-segment display, or bar-graph LEDs that are hooked up to the microcontroller.

    I haven't written the firmware yet, but this should be relatively straightforward.  With all this, it should have similar functionality to the commercially available solutions.

    Anyway, here's the circuit.  (Thanks go to ThomasEgi for some tips he provided.) Click to enlarge.
    The microcontroller shown is an ATmega8, but I will probably end up replacing it with something simpler like an ATtiny45, or perhaps ATtiny44 if necessary.  These have much lower power consumption.

    One improvement I've considered putting in is an H-bridge that would allow for the current to be reversed programmatically.  I'm not convinced that it is strictly necessary to be able to do so, though, so this design simply has a toggle switch.

    Anyway, I'd love to hear your thoughts and comments.  (If there's something you don't understand about the circuit, please ask, I don't mind explaining.)  Once this circuit is improved and thoroughly tested, and if there are people interested, I would love to get these manufactured and distribute kits among the people who are interested.

  • i'm not an engineer, but i will ask a friend to explain me the terminology etc..

    also, if you manufacture it, i'm interested...
  • rdbrdb
    edited June 2012
    Does anyone else have any comments or suggestions about the design?
  • Looks good to me, I'd be interested.

    I tried taking a DC nerve stimulator with sponges to attempt this. Either my placement was bad or it just won't work as needed. It went up to DC 1.87mA when using 100hz... Still just made my face pull/twitch when I had sponges above left eye and one on right temple.
  • rdbrdb
    edited June 2012
    @Shishou: what do you mean by "100 Hz"?  How can a DC signal have a frequency?  Do you mean you used AC stimulation?

    I've been thinking about AC stimulation for a while for the purpose of brainwave entrainment, but it doesn't seem to be as useful as tDCS (which has the purpose of decreasing or increasing neural activity in a certain region), which also has a lot more studies backing it.
    However, my design would be relatively easy to adapt for AC stimulation; it would probably require a higher PWM frequency, changing the lowpass filter cutoff, and perhaps putting a second FET current limiter in (or bypassing it entirely).
  • @rdb, I'd be interested in a kit. You definitely have expertise, and should you go ahead with producing tDCS kits, you'll have my business to support further developments and advances down the road--if the price is reasonable, of course.
  • rdbrdb
    edited June 2012
    I was planning on doing some extensive testing right when I come back from vacation - stress testing, trying to see what kind of stuff I need to screw up for it to become dangerous, testing some ESD shocks, making common mistakes like inverting battery polarity or uploading the wrong program, measuring how it responds to impedance changes etc.  I want it to be as reliable as possible.

    Then, l was planning on having the PCBs manufactured and I can distribute the parts kits to the people who are interested.  I don't imagine that the final price would be higher than $30 per kit, and that's even a pessimistic worst-case estimate; the final price will probably end up being quite a bit lower than that.
  • Do it!!
  • I just stumbled upon this resource, which contains very useful information about electrode placement, current density, and stuff:

    I wouldn't recommend the hacky device suggested in that page, but besides the horrible spelling, the page does contain a lot of useful research on current density, which will help in preparing electrodes.

    Besides that, JoVE provides very useful information on how to use a tDCS device properly, too:
  • rbd, I am very interested in you proposal, and I would be willing to buy several units even if they were 2 or 3 times that cost. You have a very well thought out design and I would love to get involved. Perhaps you could even make a business out of it! What is the current status of your project?

    As for suggestions for the circuit design, are you familiar with HD-tDCS? It uses smaller electrodes (with a correspondingly lower current) to have a more focused effect on the cortex, and it uses multiple electrodes to stimulate different brain regions simultaneously. I don't know what it would entail engineering-wise, but perhaps you could look into it.
  • Sorry, I've been really busy with various things.  I'll probably have some time to work on it some more this weekend, and I'll publish a cleaned-up design.

    So far, I've been prototyping and testing the design some more, and I've also been experimenting with various electrode materials.  I've gotten promising results, but I need to do more testing.

    I'm familiar with HD-tDCS, and I've tried to find information on the actual hardware design.  All I was able to find was some mention of "passive current divider", which sounds to me like they just hooked up the electrode pathways in parallel.  This could easily be achieved with the existing design, by just tying the multiple cathodes or anodes together.
    I can't be certain that that's what they do, though; I'm guessing that the information is protected by stupid intellectual property laws or something.
  • Take your time rdb, you're a gentleman and a scholar. In the meantime I'm doing my cognitive enhancement experiments with my newly acquired Shiva neural stimulation device (woohoo!).

    As far as electrode materials go, I managed to find some conductive polyurethane foam sheets at Fry's. I had to drive to a distant location to get it, online may be better. But be selective, most websites I came across sell anti-static foam, which I believe is a bit different from conductive foam. 

    I haven't experimented with it, as I don't have a legit tDCS setup yet (9v battery + resistor = not legit). But the very unscientific "tongue test" seems to prove that it has much higher resistance than saline-soaked sponges, as the conductive foam did noticeably less tongue-zapping. I'm sure a well-designed circuit will compensate for whatever resistance the foam has. I have no idea how well this foam will send a current through a head full of hair.

    Keep in mind, resistance increases with distance across the foam. I'm guessing uniform current distribution will best be achieved by attaching a metal plate to the back of the foam, with the lead wire attached to the metal plate. This would ensure the current travels only through the thickness of the foam and not across its surface (right?). 
  • rdbrdb
    edited August 2012
    Antistatic foams have resistances in the megaohm range.  They aren't suitable.  I'm not sure how suitable or safe conductive foams are in general; personally I think I'll stick to the tried and proven method of using a saline-soaked sponge.  That said, if you find that conductive foams work and are harmless to human skin, go for it.

    The resistance between the backplate and the skin should be under a kiloohm (preferably much less), otherwise you end up needing a much higher supply voltage to compensate for the extra voltage drop over the electrodes.

    I was planning on attaching a stainless steel sheet to the back of the sponge, and riveting that to a wire terminal, and coating the backside of the sheet in silicone.  I'm not 100% sure that stainless steel won't leach chromium when used as electrode material though, I need to do a bit more research on that - I'd prefer to electroplate something with silver chloride of course, but I don't currently have a setup for that.  More experimentation is needed.
  • "More experimentation is needed"

    Pretty much grinding, right there. Well played, sir!
  • Good thing it didn't cost too much, then ; )

  • edited August 2012
    Does anyone have information on oscillating-tDCS or tACS (transcranial alternating current stimulation)? Both pass a frequency of your choice through the brain, but the difference between them is that tACS signals reverse polarity while o-tDCS signals don't. So far I have only seen them mentioned in a few studies: 
    (most helpful)

    I wish there was as much info out there as there is for standard tDCS, as I am extremely intrigued by this approach. It seems you could combine the neuromodulating properties of tDCS with the myriad effects of brainwave entrainment. You could even use complex EEG-derived signals to target specific deep brain structures in addition to the cortex. I would love to produce an all-in-one stimulation device with multiple anodes/cathodes ran by multiple channels, monitored by neurofeedback software that delivers customized signals to various brain regions simultaneously. Such a versatile device would keep me busy for YEARS with my research!

    All of the features of my "dream device" are already built into a magnetic neurostimulation device I own (except for the neurofeedback software--that's a pretty tall order requiring EEG, programming and lots more research). This device simply uses a computer sound card that plays sound files into electromagnetic coils (essentially headphones with the magnets taken out). But the magnetic stimulation doesn't seem to pack as much punch as tDCS, so I'm curious about creating the tDCS equivalent of already existing magnetic stimulation devices. Any ideas on what a that might look like? Obviously the coils are far safer than live electrodes; a significant problem for tDCS that has to be accounted for. 

    I'm a pro with Neuroscience/Cognitive Psychology, but my Electrical Engineering/Programming knowledge is weak at best. However, it seems tACS and o-tDCS could easily be achieved with the right circuit design. Any insight from you techno-buffs would be greatly appreciated!
  • rdb's circuit can be easily modified to produce whatever waveform you want. ramps, dc, ac
  • rdbrdb
    edited August 2012
    Yeah, oscillating tDCS can be achieved easily using my design; it may require a higher PWM rate and tweaking the cutoff frequency of the lowpass filter (right now it's at a sub-hertz frequency) to allow for higher frequency oscillation.  Which frequencies are you targeting - the common brainwave bands, such as theta as described in the article?

    I don't have access to the tACS literature (stupid paywalls).  If tACS does what I think it does, my design may need some adaptation to support negative currents.  For instance, a bipolar supply or virtual ground may be needed (easy enough) and the FET-based current limiter needs to be replaced by something else (perhaps two of those in series, one of them in reverse would work).
  • Yes, I'm targeting the brainwave frequencies, 1 thru 40 hertz-ish.

    When I'm back at my university next week I can access any studies I want without paying for them, so I'll see what I can find concerning tACS. It's been a long summer... I have over 200 studies I want to get my hands on, most of them involving various types of neurostimulation. I'm going to be busy!
  • @rdb - on the post about the circuit you created on may 28th - have you developed it further, and are you interested in selling it yet?
  • rdbrdb
    edited October 2012
    I have changed a few minor things in the meantime.  I am still interested in continuing with it, but due to various circumstances, I barely have any time to work on it.

    EDIT: I thought I had posted this before, but apparently I haven't; a slightly more organised schematic is available here.
  • @rollsroyce Well in short most of those studies are going to conflict with each other, so much work needs doing to sort out both where the best electrode placement is for tDCS and also what stimulation frequencies actually yield and form of enhancement. Most studies are concerned with MEPs (motor evoked potentials) which is a very general indicator of enhancement. Although there are two papers I am quite fond of you might want to take a look at. Both are TMS studies:

  • Thanks Sherazon, interesting studies. And yes, I've also found that you can talk yourself into circles based on the findings of these sometimes conflicting studies.

    For anyone who's interested, here's what I've learned lately: For those too squeamish to build their own tDCS device, the next cheapest option is using an Iontophoresis device, the one linked below is $245. This is the same exact thing as tDCS, in fact they use these devices in many tDCS studies.

    tACS does not have excitory/inhibitory effects because the alternating current essentially makes both electrodes an anode and cathode simultaneously (rdb alluded to this in an earlier post). Oscillating tDCS combines the effects of tDCS and tACS--it excites/inhibits and has an entrainment effect as well. That means you can both stimulate a brain region and determine its dominant EEG frequency (theta, alpha, etc.). My new love is transcranial ultrasound stimulation (TUS), but this is virtually brand new, with only a few studies and no exploration by biohackers (yet). Can't wait to see what happens as this technique develops!

    Also, many thanks to the members of this forum for inspiring me. I've put off studying Electrical Engineering for years, but my frustrations in learning about neurotechnology have finally made me take the plunge. If any of you other novices want to do the same, I recommend UC Berkeley's free online courses (links below). MIT offers a few free courses as well.

  • rdb, just out of curiosity, could your circuit be constructed on an arduino board? Arduinos already have a microcontroller, voltage regulator, LM358, etc. I'm aware of the OpenStim design, but I'm curious if it can be done without the digital potentiometer
Sign In or Register to comment.