Make a Pair of Audiostrobe-compatible LED Glasses for Use With Brainwave Entrainment Software




Introduction: Make a Pair of Audiostrobe-compatible LED Glasses for Use With Brainwave Entrainment Software

Audiostrobe glasses are used in conjunction with light and sound machines, as well as some brainwave entrainment software. The glasses have LEDs built in which pulse in synchronization with sound, usually either binaural beats or isochronic tones. The light and sound work together to entrain the brain's brainwaves. Essentially, your brain adjusts its dominant operating frequency to match the stimulus.

There are many claims made regarding brainwave entrainment; some have been scientifically proven, some haven't been proven, but make sense and are theoretically possible, and some sound like the deluded ramblings of a government conspiracy nut. Entrainment can relax or stimulate the brain; various frequencies have been shown to increase attention span, reduce stress, and aid in memory retention. In addition, entrainment can be used very effectively for hypnosis and meditation. There is some evidence that entrainment can, to a limited extent, replace sleep, although the only study I've seen used magnetic pulses to entrain the brain. Some people use it to achieve lucid dreaming, in which you realize that you're dreaming and take control. There is also, of course, a bunch of baloney floating around as well; people claiming everything from astral projection to telekinesis to regrowth of hair. Some even claim to have achieved enlarged genitalia through the use of brainwave entrainment. YMMV. If it works for you, email me and I'll let you know where to send the check.

This instructable will detail the construction of a modular pair of Audiostrobe-compatible LED glasses. I use a couple of entrainment programs, all of which can output an Audiostrobe signal. I've heard that photic entrainment, when combined with auditory entrainment, can be much more successful than the auditory stimulus alone. Upon pricing these glasses, I found that they range from $20-$60, and you had to buy a separate decoder device which starts around $115. Not only am I poor, but the thought of spending $60 on a pair of safety glasses with 12 LEDs glued onto them doesn't really appeal to me. So here's a set of plans for a modular Audiostrobe decoder. The decoder unit can switch between two outputs, so you can place two sets of LEDs on a pair of glasses and switch between them at will. You can either have two different colors of LED, different numbers of LEDs, or a different arrangement of LEDs on the glasses. The glasses themselves can be swapped out for a different pair, allowing you to keep several different styles for different needs or users. The decoder is built into an Altoids mint tin, because what Instructable would be complete without one of those?

The cost for the project will vary, depending upon what you have on hand and how good you are at scrounging. I already had everything except the audio jacks and the LM567, so I'm not sure exactly how much everything cost. The LM567 costs $1.80 at Digikey; I got mine 3/$1 at Electronics Goldmine. The jacks were $.53 apiece, and the 7805 costs $1.95 $.49 at Digikey. The electrolytic caps are fairly cheap; you only really need one of them if you're building this on the cheap. See step 1 for a list of what can be left out and what can't. The smaller ceramic caps and resistors are dirt cheap, but you either have to buy an assortment or buy a minimum amount (10 at Digikey). The switches I ripped out of a broken playstation controller (had an LED, too, though I didn't use it for this).

Teacher Notes

Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.

Step 1: Some Background Info

After digging around on the Internet, I found a schematic detailing a circuit to decode Audiostrobe signals. I won't post it here because I was unable to contact MagicJim, the author, to get permission. I based my circuit upon that one; however, since he simply used the example circuit included in the LM567 datasheet, I'm fairly comfortable that I'm not infringing upon any copyright by posting this.

Audiostrobe works by encoding a high frequency (19200hz, apparently) signal into the audio stream. Most humans can't hear it, but a tone decoder can detect the signal and apply power to the LEDs when it is present. I'm not sure if the commercial decoders are this simple; maybe there's additional data sent with the stream, perhaps like a serial data line, only in audio. I was pleasantly surprised to find out that some software has brightness control built in, and it works with this circuit. After reading the tone decoder's datasheet, I'm clueless as to why. The decoder is a simple switch; if the tone is present, it connects the LEDs to ground, with no current or voltage regulation. The only thing I can think of is that the software uses PWM to drive the LEDs. In any case, this circuit works.

MagicJim commented that you need two circuits, one for each eye. Although dissimilar auditory stimuli work well, as in binaural beats and other techniques dealing with the two hemispheres of the brain, I've seen no studies showing any benefit to separate stimuli for the left and right eye. In any case, I was more interested in having selectable LED banks than in seeing how massive a migraine I could give myself by stimulating each eye separately, so both eyes are synced.

The decoder unit runs off of a 9V battery. Due to the inefficiency of linear voltage regulators, combined with the low capacity of 9V batteries, I don't expect battery life to be phenomenal. I originally planned to use the MintyBoost circuit to power it from two AA batteries. I didn't have the parts on hand, however, and I didn't feel like waiting for parts to arrive. I haven't been using the device long enough to get any real battery life data, and at any rate, I've been using old batteries that I had after replacing the ones in my smoke detectors, so it wouldn't be accurate anyway. If it turns out that battery life sucks, I may redesign the circuit to include the MintyBoost.

I've made a few changes to the original circuit. I added some ripple protection caps to the 7805. They're probably unnecessary since we're using a battery. I had planned on having a DC jack with contacts that broke when a plug was inserted. That way it'd run on batteries until you plugged in a source, at which point it'd automatically switch to the supply line. I scrapped that because there were too many cables going into the unit as it was. If you don't happen to have them on hand, you should be able to leave C6 and C7 out without any issues.

I also added a diode to protect against inserting the battery incorrectly. I placed a SPST switch for power, and a SPDT switch to choose between output lines. I also added a power indicator LED and accompanying resistor. If you don't want the indicator, LED1 and R3 may be safely left out.

The sound and signal input is supplied by a male to male audio cable. This cable can be hooked up to you computer's sound card if you're using entrainment software, your cd player if you're using commercial cds, or the output of a light and sound machine. I've used it with several different entrainment software packages, and it worked with all of them, but I haven't tried it with cds or l&s machines. I'd be willing to bet it works with cds, Hemisync for example, but I can't promise anything. L&S machines, on the other hand, are iffy. If they output a bog standard Audiostrobe signal, it should work fine. But if they utilize any special features, they may not work.

The cable for the glasses is a standard audio cable with a 3.5mm stereo plug. You can either cut one off of a pair of headphones or cut one end off of a male to male patch cable. I'd recommend the patch cable, as they tend to be a bit beefier, and can handle a bit more abuse.

There is a headphone jack included in the unit. Most entrainment software and cds require that you wear headphones, rather than speakers, while using them. I was originally going to just use an external cable splitter, but decided that the likelihood of me losing it and being unable to use the unit was great enough to justify the extra $.50 cost of an additional stereo jack.

Instead of using a 9V battery snap connector, I added some spring contacts. I've always hated snap connectors, and I've broken quite a few over the years. Unfortunately, I couldn't find any sacrificial electronics that were willing to give up their contacts, so I had to improvise. See step 4 for more info.

There are numerous design flaws in my unit. The actual glasses, for example, were built while I had MagicJim's circuit breadboarded, before I started making changes. I placed one set of LEDs on the glasses, wired them and expoxied everything into place, all the while thinking how great it would be if I could stick a second set of LEDs on the glasses. No, I thought, I don't want an extra wire dangling. It wasn't until roughly twelve seconds after the epoxy finished curing that it finally occurred to me that I could've used the second signal wire to power a second set of LEDs and use a SPDT switch to choose which one to use. That change made it into the circuit, but I haven't built a new pair of glasses yet.

There are several similar changes that could be made to make this unit better, but I'm going to detail the construction of the project as I made it. I know the unit I built works, but I can't guarantee any changes I suggest won't break it. In addition, by pointing out problems that occurred for me I might save somebody else some grief. At the end of the instructable, I'll suggest any changes I feel are necessary.

I'm not going to go into excruciating detail in certain areas that have already been covered well by other instructables. I'm not going to talk about etching the PCB, for example, because there are numerous instructables detailing exactly that. Nor am I going to go into extreme detail on things like drilling the holes for your jacks and whatnot. This is basically just the outline of what I did, what went wrong, and what I would change so that somebody else can do a better job their first time than I did.

I should point out that I am not an electrical engineer, nor do I know anything about electronics (a fact that will likely become painfully clear shortly). I don't know anything about audio, either, so I make no claims about the safety of your high-end sound card while using this. There's no amplifier in this circuit, and I've no idea how much of a drain it places on the signal. It's possible it can cause an impedance mismatch, and if so, your equipment might be toast. Additionally, there's no line isolation between the circuit and the signal input. You're placing an electrical circuit into a metal box and plugging it into your computer. If you were sloppy, you're going to have a bad day. You've been warned. If something bad happens, it's not my problem. Do not build this circuit if you can't take responsibility for what might happen. That dire warning aside, I've been using it for several weeks now, with two computers and a cd player, and haven't had any problems.

Oh, and for God's sake, if you've got epilepsy or some similar light-induced disorder, don't build a freaking machine designed to pulse light and then try to sue me, okay?

Step 2: Parts and Tools


Resistors (use 1/4W, to be safe):
  • R1 - 470 Ohm
  • R2 - 47 Ohm (read notes at end of step 8; you might prefer to just use a bus wire)
  • R3 - 47 Ohm (if you use a diode other than green with 2.2v drop, you might need to change this; probably not critical)

  • C1 - 0.1uf
  • C2 - 0.005uf
  • C3 - 0.02uf
  • C4 - 0.01uf
  • C5 - 1uf
  • C6 - 100uf
  • C7 - 10uf
Note: C1-C4 can be any type; I used ceramic. C5-C7 should be electrolytic

  • D1 - 1N4004 (any small-signal diode should be fine here)
  • LED1 - standard 3mm green, 20mA with 2.2v drop

  • 3 x 3.5mm stereo jack (if you get different jacks, the leads may not align with the traces on the board; check the datasheets)
  • SW1 - SPST switch
  • SW2 - SPDT switch
  • 9V Battery connector (see note)
  • 9V Battery
  • PCB or protoboard
  • Wire (around 24-28 gauge)
  • Altoids mint tin
  • Various nuts, bolts, and washers.
Note: I chose to go another route for the battery connection. My way makes changing batteries much easier (and less likely to have the connector break while changing batteries), but it's significantly more time consuming to build than just soldering in connector wires. Your choice.

  • Safety glasses
  • LEDs, color and number of your choosing
  • Resistors, dependent upon your choice of LEDs
  • Either a headphone cable or an audio patch cable

Tools and Supplies:

  • Soldering Iron
  • Solder
  • Wire cutters
  • Drill (preferably drill press, but hand drill will do)
  • Screwdrivers, etc.
  • Epoxy (get the two hour stuff; 5 minute sets too fast to be useful)
  • Super glue (for when epoxy just doesn't cut it)
  • Electrical tape
  • Sandpaper
  • Rubbing alcohol

Nice to have:
  • Flux
  • Dremel
  • Belt sander
  • Needle files
  • Center Punch
  • Helping Hands or vise
  • Magnifying glass

Step 3: Create the PCB

Create the PCB using whichever method you prefer. I used the toner transfer method. The edges of the PCB must be curved to fit into the Altoids tin. If you're using the toner transfer method, transfer the toner to the board and etch, then use sandpaper or a belt sander to remove the excess PCB outside the curves. I waited until after etching to sand for ease of aligning the silkscreen; as long as the audio jacks you're using don't have metal cases that are connected to ground, the copper outline won't hurt anything.

I've included the bottom trace layer as well as a component layout "silkscreen" layer. After etching, I go back and apply the silkscreen. Makes placing the parts so much easier, plus you've got a nifty battery polarity indicator (admittedly only useful if you use my method of battery connection).

The layouts are in GIF format, 600dpi instead of the standard 300dpi. Make sure you download the image in its original size. When you print, the boards should end up 3.5'" x 2.2". The image with both boards should be printed as 7.5" x 2.2". Optionally, I've included pdf files with the images formatted correctly for standard 8.5" x 11" paper. Just open and print. Be sure you don't have any headers, footers, etc. added in, or the boards won't be the right size. Get out the ruler and check the output, just to be sure.

The masks I've uploaded have been altered slightly based on my results. The battery pads are larger, as are the pads for the jacks. The power lines have been made thicker, as well. And the text is slightly larger.

If I remember correctly, I used a #8 drill bit for all of the holes except the power and jacks. The jacks are thin, elongated leads, so I drilled two holes side by side, and I did something similar with the power pads. I don't remember the bit size I used for those, however.

If you decide to go a similar route to my own for battery connections, you'll want solder pads on the top side as well for reinforcement. I didn't feel like making a toner mask just for two pads, so I just drew them on with a Sharpie permanent marker. Draw them roughly the same size as the bottom side pads. Let the ink dry for a few minutes, then go over it again. What the hell, go over it again once more for good luck.

Note: there is one area that might give you a little trouble, depending on your printer and the method you use. One of the pads on the Audiostrobe output jack is fairly close to the ground lead. I don't ever have problems with bridging; my issues are always with overetching/opens, so I doubt it'll be a problem. Just check it before you etch to make sure it's not bridging. There's a note on the Traces gif showing you where I'm talking about. You'll also want to pay attention to the pads on the 7805.

Step 4: (Optional) Make Battery Spring Contacts

If you'd prefer to spend a little extra time now in exchange for a much easier life later on, I'd suggest you build the battery connectors I did, or something along these lines. If you can scavenge spring connectors from another device, so much the better. I couldn't find any, so I had to improvise. By dint of heroic effort, by which of course I mean I looked around the room for something suitable, I found the perfect solution. Well, not perfect, because the perfect solution wouldn't require half an hour with a dremel to get working.

I got a couple of cheap picture hangers and cut them to size. By cheap I mean really thin, flimsy metal. The one which inspired the idea was too thick; there was no "spring" to it. I got some at Walmart that are satisfactorily shoddy. Note that mine are brass plated, and solder loves brass. This is an important point. If yours are a silver color, they're probably galvanized. Good luck soldering those.

I used a cutoff wheel with the Dremel to cut the tabs shown in the first picture. I soldered them in place, with much cursing of unstable parts that like to move before the solder's solidified completely. I cleaned the board scrupulously with rubbing alcohol (note that I populated my board before doing this; I didn't have any picture hangers on hand and couldn't wait to start soldering). I placed the battery on the board for a fit check. It was then that I discovered the value of getting off of my fat butt and walking across the room to get the battery rather than guesstimating. As you can see in the second picture, the contacts are a bit short. They make only the most incidental contact with the battery's terminals, and my bet would be that the slightest movement of the unit would break contact, albeit only momentarily. Not good enough.

In the third picture, we see a more satisfactory set of contacts. Although they don't look it, they touch both the top and bottom of the terminals, and the better angle of the contact area gives a much more secure hold on the battery when it's wedged into the case. The angle I added at the back really seemed to help.

Step 5: Drill Board Mounting Holes, Populate and Solder Board

I have to apologize here for the lack of photos. I thought I was doing a really good job of taking pictures as I went, but apparently there were several rather significant lapses. Before I populated the board, I chose some locations for mounting holes and drilled them. Obviously I avoided areas with traces, but I wasn't too terribly careful because I planned to use nylon washers for electrical isolation. The locations shown worked out great for me, with the exception of the one shown in the top right of the back of the board. The line select switch takes up that space. Even so, there doesn't appear to be any wobble when you plug or unplug into that jack.

After drilling the holes, I epoxied some nylon washers to the bottom side of the board. Not only do these add the height I need to keep the leads from touching the case, they keep the nuts from contacting the traces. I don't remember the exact size I got; it'll vary depending upon the screw you use. I used M3 screws, 1/4" I think. Be careful not to get epoxy in the hole. I smeared a very thin amount of epoxy around the holes, pressed the washer flat to the board, and then placed a fillet of epoxy around the rim of the washer using a syringe. Be sure to get the 30 minute epoxy and not the 5 minute stuff. Once everything's situated, you can cure the epoxy faster if you apply heat. Just set the board on a heating pad and place a cardboard box, like a shoe box, over it. You could use super glue here, but I wouldn't recommend it because once you put the washer down, that's it. You can't move it around if you didn't get it perfect.

Now might be a good time to figure out where to place your mounting holes in the Altoids tin. I wish I'd thought of doing it at this step when I built mine. Place a washer on one of the mounting screws, place it through the hole in the top of the board, and screw on a nut. You should have screw, washer, board, nylon washer, nut. If the screw pokes out through the bottom of the nut, you either need another washer between the screw and the board, or you need a shorter screw. Repeat for all of the other mounting holes. Now place the PCB on top of three 3x5 cards beside the tin. The 3x5 cards give a reasonable approximation of the thickness of the tin's bottom. With the nuts resting on the cards, the board's height should be very close to where it will be when mounted. Mark a line where it contacts the side of the tin. If you used the stereo jacks I suggested in the parts list, you can view the datasheet and see that the plug is 0.236" or 6mm in diameter. So the center of your mounting hole needs to be 0.118" or 3mm above the board line. You can either measure it, preferably with a pair of calipers, or you can eyeball it. I chose the latter method, so I can heartily attest to the fact that it sucks.

Make sure that the hole aligns with the center of the silkscreened jacks horizontally and mark its location. Sharpies work well. You will, of course, need to make certain that the board is aligned with its eventual position inside the tin. Try to be as accurate as possible here, but don't break out the micrometer. Later we'll discover that we have to drill the holes a bit oversized anyway, so there'll be some built-in slop.

Using some fairly rough (320-grit?) sandpaper, scuff the inside of the tin's bottom. Clean the tin using alcohol. Unscrew the screws and place the board into the tin. The nylon washers will be resting on the bottom of the tin. Using a Sharpie, mark the center of each hole, then trace around the inside of the hole.

Now you need to place your parts. Again, I didn't get a shot of the unmounted board with the parts placed. The best I've got is with the board mounted in the tin. There are a few parts with polarity here; IC1 and IC2, C5-C7, D1 and LED1. The electrolytics have a stripe running down one side with a "-" symbol. The polarity is marked on the silkscreen with a "+" symbol; obviously the negative side of the capacitor goes opposite this. If you didn't do the silkscreen or it isn't legible, just refer to the silkscreen gif from step 3. The silkscreen has been mirrored to allow for toner transfer, so flip it horizontally in your favorite image editor. IC1 has a notch on one side; align that with the notch on the silkscreen. IC2 should be mounted with the leads bent so that the metal tab along the back is flat on the board. D1 has a black line (assuming you've got a standard red glass diode; if you've got a larger black plastic diode, it'll be a white or silver line) around one end; align that with the black line on the silkscreen. LED1 might be tricky. There should be a raised rim around the base. One side should be flat. Hold the LED so that the flat side is aligned with the flat side on the silkscreen. If the leads are aligned with the holes, you're in luck. Otherwise, look into the LED. There should be two metallic pieces discernible. One should be vaguely triangular. That one should be oriented towards the flat side on the silkscreen. Note that you'll want to protect the LED's leads from shorting somehow. I stripped the insulation off of some wire and placed it on the LED leads. Very small diameter heat shrink tubing will also work. In a pinch, electrical tape will do as well.

The resistors and ceramic caps can be oriented either way. Clip the leads short, and bend them along their traces on the back side. If you view the pic of the back of the board at full size, you can just make out the bent leads. Be careful not to bridge any leads. Make sure before you solder that the audio jacks are flush to the board. If they're raised at all, they'll be much less sturdy as you're plugging into them. Plus, they'll be a pain to drill mounting holes for. Solder away. Flux is your friend. Until it's time to clean the board, when you'll start to hate it with a passion.

You can see I mounted my switches to long wires. I laid out the wires to the power switch to see how long I needed them, as I didn't want any excess interfering with the battery. The line selection switch I simply guesstimated how much I would need. As I said, I took these switches from a playstation controller. They're actually both SPDT switches, but I'm only using one side for the power switch. I left them on the bit of PCB they came on, as it makes a handy mounting platform. Actually, it occurred to me that it would make a handy mounting platform just after I finished removing them both from the PCB, so I got to solder them right back on. Amazing, how I always come up with good ideas immediately after finishing the work they would have saved me. Anyway, I epoxied some nuts to the PCB, measured the placement, and drilled some mounting holes in the Altoids tin that don't even remotely match the nut placement. I really need to stop that eyeballing habit.

Step 6: Secure Mounting Nuts, Drill Holes in Tin

At this point, you need to drill the holes for your jacks, LED, and switches. You may have marked the holes for the jacks in the last step, or you may have waited until the jacks were mounted in the board so you could just trace around them, you crafty devil, you. I'd prefer to mark them before so I can get a more accurate centerline, then check for fit once the jacks are in place, but whatever works for you. Either way, double check the placement of your holes, and break out the drill.

I'd really recommend using a drill press to drill the holes if you have one. The bit likes to wander a bit if you don't. Regardless, use a spring-loaded center punch on the center of the holes. Chuck up a small drill bit. If you just jump to the bit you want for the final size, you're likely to have it catch in the thin metal of the tin and either damage it or worse, damage you. You'll want to start small, maybe 1/8", then work up to a 1/4" bit. "But wait," I hear you say, "the jack's more than 1/4" in diameter, and you said we'd need to make the hole oversize!" That's what the needle files are for. Once you've reached 1/4", break out the needle files and enlarge the hole slightly. How slightly?

The reason for the oversized holes is likely obvious to you by now, but in case, like me, you're a little slow on the uptake, allow me to explain my stupidity in detail. The metal barrel on the jack into which we place our plug is the ground connection for the jack. On the line input and headphone jacks, it doesn't matter if the ground contacts the metal tin, because both jacks are tied to a common ground and we went to great pains to ensure that none of the other components contacts the tin. The audiostrobe output, on the other hand, must be isolated from ground. The ground from the audiostrobe line goes to the tone decoder; the circuit remains open until the audiostrobe signal is detected, when the decoder ties the line to ground. Current flows through the LEDs and they light up until the signal stops, at which point the decoder breaks the connection. Short the audiostrobe's ground connection to ground, and the LEDs come on and stay on, regardless of signal.

I hadn't realized until I had soldered the jacks in and was poking around with the multimeter that the external metal barrel was the ground connection. This really made placing the board into a metal tin sound like a pretty crappy idea. Nevertheless, it was one I was loath to give up, especially after I went to the trouble of sanding pretty rounded corners on my PCB. So, after an exhaustive search for the best solution, during which I must have looked through at least two (2!!) drawers, I settled on paint. White, oil based enamel paint, to be exact.

Flash forward. Paint sucks. Scrapes right off smooth metal. This, by the way, is going to be a recurring theme in this little saga. Tried a thin layer of epoxy. I'll spare you the details. Heat shrink tubing would probably work great, but I didn't feel like performance testing the heat resistance of my thin, cheap plastic stereo jacks, so I opted for my favorite, electrical tape. I cut a piece of electrical tape about 1" long, then cut it into 3 strips along the length. Started the wrap just on one side of the bottom, wrapped 360 degrees to overlap the start, then cut the tape about midway up the side. The idea is to get just one layer thick along most of it, but two layers along the bottom. You'll be levering the jacks along that bottom edge, so you'll want the extra layer. Don't trim the tape yet! Pinch it in neatly in several places and try to turn it into the barrel of the jack. If you trim the tape now, it will catch on the edge of the hole when you're inserting the board and get skinned back. If you've gone up from 1/4" to the next size drill bit (my next size is 5/16"), this probably won't be an issue. If, on the other hand, you've used needle files to enlarge the holes for a cleaner look, you may be in for a tight squeeze.

Continue with the drilling. Mark out the locations of your switches and any mounting hardware, if necessary, and drill. Don't forget the hole for the LED. You're on your own as far as placement goes, as your switches won't be the same as mine. Plus mine look like crap, so you don't want to copy me. Once you're done drilling, be absolutely certain that you've removed all the metal dust, chips, and scrapings from the interior before you continue.

Hopefully in the last step you scuffed and cleaned the inside bottom of the Altoids tin. I didn't, so listen carefully when I tell you that you need to scuff and clean the metal like an obsessive compulsive. I can assure you that, should you go to the trouble of levering the PCB into position, giving yourself a blood blister and nearly killing the dog in the process, only to find, as you triumphantly screw your mounting screws down, that the nuts you so carefully epoxied into place are prone to popping loose with reckless abandon, you will not be a happy camper. So I once again strongly encourage you to scuff and clean the metal before you epoxy the nuts.

Should you be one of those carefree souls who blithely forges ahead without fear or caution, regardless of the warnings of others: have hope. I have discovered, to my surprise and delight, that humble super glue will serve to rebond the nuts to the metal. I don't know how well it would work to bond bare nuts to the tin, but it bonds nuts with a fillet of hardened epoxy around them to the tin exceedingly well. The problem being, of course, that you have one shot, and one shot only, to get the nut in the right place. Once it's down, it's down for the count.

At any rate, mix up your epoxy, spread a thin layer around the holes you traced, and press the nuts down, making sure they're aligned with the holes. Then either lay a fillet using a syringe or just use a thin stick to place epoxy. You want the epoxy fillet to be even with the top of the nut, and the diameter of the fillet should be at least twice the diameter of the nut. Be very careful not to get epoxy in the nut! This is very easy if you're not using a syringe. If you do get some in there, just chuck the nut and start over fresh. You'll likely need to remove all of the epoxy from the first nut before laying its replacement, lest the epoxy ooze up into the center of the nut.

Wait for the epoxy to dry. This would be a great time to skip ahead to step 8 and build your LED glasses. Never fear, you get to play with epoxy there as well. Once the epoxy has hardened, start laying electrical tape. Try to cover every exposed piece of metal along the bottom. If you placed your parts properly when populating the board, the sides of the tin will not be an issue.

Step 7: Test, Mount Board

You may want to test the board before you mount it. I assume you've already got a source of Audiostrobe signals; hence your only reason to build this project. If not, look in the last step for a list of software that can generate these signals for you. For now, just use the test file I've provided. It's m4a, not mp3, but Winamp and Media Player should both play it without a hitch. I'd assume most other media players would as well. Encoding to mp3 apparently cuts the carrier signal out. I haven't experimented with that, so it may just be my settings. I'd recommend burning the file to cd and using an old portable cd player for testing and the first couple of times using the unit, just to be safe.

  • First, with the power disconnected, check with a multimeter to make sure you've got a connection from each pin on the line input jack to the corresponding pin on the headphone jack.
  • Check the resistance from the barrel of the line input jack to the barrel of the audiostrobe jack. It should be around 1k Ohms. If it's open, there's a problem; if it's a dead short, there's a problem.
  • Double-check your polarities, especially on the electrolytics and the 7805. C5 is closest to the edge of the board; its negative stripe faces the edge of the board. C6 and C7 butt up against each other; their negative stripes face each other. The 7805's metal tab should be flat against the board.
    • With the power still disconnected, connect the line input to an audiostrobe source. Plug some headphones into the headphone jack. You should hear the audio. If not, well, I don't know what to tell you. You already checked the connections to the jacks. Check 'em again.

Now connect the power and turn the switch on.
If the LED doesn't light:
  • Check for voltage between the battery connectors (not terminals)
  • Check LED1's polarity.
  • Check for voltage between the negative battery terminal and the LED lead that's farther away from the battery. If there's at least 2.2v, the LED's polarity is wrong or it's bad. If there's voltage, but less than 2.2v, either you've got the wrong resistor value for R3, the battery's almost dead, or you've got a connection problem (likely a shorted component). If the resistor is correct and the battery's fresh, check all of your connections for bridges.
  • If there's no voltage, check for voltage between the negative battery terminal and the lead of D1 that's facing away from the battery (the one with the polarity symbol). If there's voltage (around 7.5-8V), you've likely got an open somewhere in the circuit. Start looking for unsoldered connections or broken traces.
  • If there's no voltage at D1 check the polarity. If the polarity's correct, check for voltage on the side of D1 that's facing the battery. If there's voltage there, D1 is bad. Replace it with a diode that can handle more current.
*If there's no voltage at the near side of D1, your last hope is SW1. Check for voltage at the SW1 lead that's farther from the edge of the board. If there's no voltage there, and there is voltage at the side closer to the edge, SW1 is likely faulty. Try shorting the leads of SW1 and see if the LED lights.

If the LED lights, you're ready to try the unit out. If you've already built the LED glasses that accompany this unit, turn the power off and plug them in. Otherwise, you'll need either an analog multimeter or an oscilloscope to check the operation. Digital multimeters will not work. Play the audiostrobe file I've provided. If you're listening through the headphones, don't be alarmed by the highly obnoxious nature of the tones. The actual audiostrobe signal is inaudible to most humans; I made the tone sound this way so I could see how the audiostrobe pulse was syncing with the audio tones. It doesn't usually sound like that :)

If you've got the glasses plugged in, the LEDs should flash briefly when you first power on the circuit; if they don't, check the line selection switch. If you're using an analog multimeter, the needle should jump upon power up. If you're using an o-scope, I'm going to assume you know enough to not need instructions from here on out.

The file starts out with a very quick pulse that ramps down to around 1hz within a few seconds. You'll hear it in the audio. The rate then smoothly begins to ramp up, about 1hz every six seconds, if I remember correctly. So the LEDs should be flashing about once per second immediately after the ramp down, twice per second six seconds after that, three times per second six seconds after that, etc. The file runs about two and a half minutes. Whether you're using the glasses, the multimeter, or the scope, you should be seeing the pulses. If you're not, and you can hear the tones, try turning up the volume. There's a minimum threshold for the circuit, and in this particular file, the audio tones are extremely loud, even uncomfortable. If that works, don't worry; you won't have to listen to all audio at that volume.

If it doesn't work, and you've passed all of the above tests, you're in for some basic electronic troubleshooting, which unfortunately is beyond the scope of this instructable. The only quick suggestion I have would be to check the voltage between the negative battery terminal and pin 4 of IC1. The voltage should be about 5 volts. If it's not 4.5-5.5V, the problem is likely with IC2, the voltage regulator. Otherwise, the most likely culprits are the electrolytics, especially if you bought them surplus or removed them from old equipment, or the tone decoder itself.

Hopefully the unit passed the above tests. So now you get to shoehorn the board into the tin. It won't be easy (unless you drilled your holes waaay oversize). The easiest way I've found is to insert the end with the two jacks first, pointing them into their holes. Bend the wall of the tin above the audiostrobe jack hole outwards to allow the jack to pass into the tin. Start maneuvering the PCB into position, alternating between trying to slide the left end of the PCB (with the two jacks) into position and lowering the right end of the board. Once the barrels of the two jacks are in their holes and aligned correctly, the board will usually just pop into place.

Before you secure your switches, screw down the board. I cannot stress enough the need to be gentle when tightening your screws. Finger tight, no more. The washers shouldn't move when you poke them with tweezers, but neither should there be indents in the board. Then repeat the above tests. If the board fails, either you've got a short somewhere or you broke something installing the board. If the LED glasses light up and stay lit, your audiostrobe jack barrel is almost certainly shorting to the tin. You'll have to remove the board and reinstall your electrical tape.

Once everything's working correctly, mount your switches and position your LED. Pop in the battery and close the case. Use an X-Acto knife to trim the electrical tape flush with the ends of the jack barrels. You'll need something to brace the battery against the case. I originally had visions of foam rubber strategically glued to the wall, but in the end, folded paper won me over with its simple, yet sleek and modern appeal.

Step 8: Build the LED Glasses

Okay, this is the easiest part, so of course I buggered it all up. I'll walk through what I did while pointing out the numerous mistakes as we go. Hopefully, you'll then be able to avoid them, or at least feel really dumb if you don't.

I already pointed out that these are single channel glasses, not taking advantage of that high tech switch thingy we installed in our decoder unit. Only one set of LEDs. And only two LEDs per eye, right next to each other. When experimenting with the breadboarded circuit, I checked to see if two LEDs would be sufficient to see through closed eyes. They were. It wasn't until after I had built the glasses that I realized two LEDs were bright enough if they were facing my retinas and my retinas were facing them. Change the angle of either, however slightly, and all you get is a faint suggestion of light. My next pair of glasses is going to have at least three LEDs per line for each eye, spaced about three quarters of an inch apart, in a star pattern (well, triangular, but if there's two lines...)

Anyway, I took a pair of safety glasses and marked on the lenses what I thought was the center of my field of view when my eyes were relaxed. Then I placed the LEDs slightly above that. I did this purposely. The idea was that my eyes would automatically look upwards towards the flashing light. Studies have shown that closing your eyes, consciously relaxing, and rolling the eyes upward causes the brain to produce alpha waves. I figured I'd get a free boost to the efficacy of the glasses this way. There were a couple of things wrong with this reasoning:

  • Contrary to what I expected, the eyes don't naturally gravitate towards the flashing light while using the machine. They gravitate towards the at rest position, which I had carefully marked on the glasses before placing my LEDs elsewhere.
  • If the LEDs were in the center of my field of view, I'd likely be able to see the pulsing no matter the position of my eyes. As it is now, I can only see it if I consciously look up. Since the entire point of brainwave entrainment is to reduce the activity of the conscious mind, this method is not entirely reliable.
  • It only occurred to me after I built the glasses that I don't always want to produce Alpha waves. If using the machine for meditation, for example, I'm likely trying to reach Theta. If it had worked like I planned, the placement of the LEDs would actually be hindering my progress rather than aiding it.

To place the LEDs, I made holes for the leads in the plastic lenses. I wanted the LEDs mounted inside the eyepiece so they would be closer to my eyes. Were they mounted on the outside of the lens, further from my eyes, the area they'd illuminate would be greater, but the intensity would be less. To make the holes, I straightened out a paperclip and heated it with a blowtorch. Once it was red hot, I poked it through the lens. After making all of the holes, I scraped off the piled up melted plastic from around the holes with an X-Acto knife. I mounted the LEDs with epoxy.

The LEDs for each eye are wired in parallel, and the two eyes are wired in series. So in my glasses, the positive (signal) wire runs to the anodes of the LEDs in the right eye, which are connected to each other. The cathodes (marked by the flat side, remember) of those LEDs run to the anodes of the LEDs in the left eye. The cathodes of the LEDs in the right eye are connected to the ground of the audiostrobe jack. I cut one end off of an audio patch cable rather than cut up a pair of headphones, so I've only got one ground wire. If you use headphone cables, you should tie the two ground wires together.

I just used the LED leads to connect to each other, as you can see in the first picture. I used a scrap piece of wire to connect the cathodes of the right eye to the anodes of the left. The ground wire is encased in an extremely ill-fitting piece of heat-shrink tubing. The cable is epoxied to the front of the glasses; the connection stops at the hinge for the stem. You can see the unused channel's signal wire waving forlornly in the wind. If I hadn't epoxied the LEDs, I could go back and install another channel, but with my LED placement problems, I'd rather just build a new pair from scratch.

I used high brightness red LEDs. My next pair will likely have red and blue. Which brings up the problem of R2 that I mentioned back in the parts list. In theory, each LED array needs a different value current limiting resistor. This resistor (R2) ensures that the LEDs don't draw so much current they blow themselves out.

This pair of glasses has a 2x2 array; 2 parallel arrays of 2 LEDs in series. The LEDs probably have a forward voltage of 1.9v and a typical 20mA current requirement. If we changed to, for example, blue LEDs, which have a higher voltage drop, the value of R2 needs to change. Similarly, adding resistors to the array changes the value needed for R2. The easiest way around this would be to remove R2 and replace it with a bus wire. Then wire the correct value resistor inline with the LED array on the glasses. This way, you can swap out glasses without worrying about the resistor value.

On the other hand, it probably isn't a huge deal unless you make glasses with wildly differing resistor needs. I suspect Audiostrobe uses PWM to vary the brightness of the LEDs. PWM can be used to overdrive LEDs far beyond their maximum current, increasing their perceived brightness greatly. It works by rapidly switching the LED on and off. The brightness can be modulated without changing the power applied by altering the amount of time the LED is on and off. Because the LED is off a significant portion of the time, it doesn't overheat. But it flickers on and off so quickly we see it as being continuously on. If that's the case, you likely won't have any problems even if you don't change R2. But when I rebuild my glasses, I'll probably swap out R2 for a bus wire and integrate the resistors into the glasses.

I know some people suggest blacking out the lenses to decrease ambient light. I myself prefer not to walk into walls while moving from the computer to the La-Z-Boy, and instead opt for the more old-fashioned method of turning the room lights out upon getting situated. I would suggest that blacking the lenses might be a good idea if, for example, you're using the glasses to reduce stress on a plane before a flight, but I highly doubt you'd get this thing past those highly trained TSA employees.

Step 9: Wrapping Up

So there are a few things that in retrospect could have been done better.

  • R2 should have been integrated into the glasses
  • There's no brightness control for the LEDs. This isn't a problem if your software allows you to adjust brightness, but if your software doesn't have the option or you're using commercial cds, you're out of luck. It would have been easy to integrate a variable resistor, but I didn't have any of a low enough value on hand and was too impatient to wait.
  • Uses 9V batteries. Not much capacity in these. I don't know the current drain of the circuit; it would vary anyway depending upon your LED array. But considering each entrainment sessions lasts anywhere from 30-60 minutes, I don't see you getting many uses out of each battery. Real world testing has yet to be done.
  • I would have liked a DC input jack, but space was tight and I didn't like the idea of yet another cable dangling off of the unit.
  • There's no isolation between the circuit and your input line. Should there be a short or some catastrophic failure, your valuable equipment might go away on you.
  • I don't know enough about audio to know if it's a problem, but it's possible that this could cause an impedance mismatch issue. I'm not entirely clear on impedance, but my understanding is that a source is designed to drive a load with a specific resistance (i.e., 8Ohm vs. 4Ohm speakers) and driving the wrong type can cause serious problems. Whether this might be an issue or not, I can't say. So far, nothing's exploded on me. Cross your fingers (or better, somebody who knows what they're doing tell me if I should be praying).

Software using Audiostrobe (only software I've personally tried):


SBaGen - rather non user-friendly, uses text-mode configuration files to generate binaural beats. If you like running Linux without a GUI, you might like this. Otherwise, I'd go with an easier to use app.

Gnaural - Open source, multi-platform binaural beat generator. Has an easy to use graphical interface similar to Brainwave Generator, a commercial program. In fact, I'd say the two were almost identical in features and design, except that Brainwave Generator has a massive library of presets, while I can't find any for Gnaural. If I'm looking for something interesting to try, Gnaural comes up short, but if I know what I want to do, it's the one I use.

Edit: I just went to create a session using Gnaural and discovered that it actually doesn't have Audiostrobe support. I did all of my testing with Neuro-Programmer 2 and Brainwave Generator; for some reason I thought I had tried the glasses out with Gnaural, but I guess I was mistaken. I checked Gnaural, and found that it, too, lacked Audiostrobe support. So I guess there's no freeware (that I'm aware of) that uses Audiostrobe. Sorry about that. I'm so ashamed <hangs head>.


Neuro-Programmer 2 $45-$60, 15 day trial - Far and away the best program I've found. Has a large array of presets, and has by far the largest number of entrainment techniques. In addition to binaural beats and audiostrobe, it uses monotones and isochronic tones (which I found to be far more effective than binaural beats), dual induction techniques, recording modulation, hemispheric synchronization (fancy term for asynchronous tones) and more. Has built-in tools to record scripts and suggestions (and powerful tools for manipulating those during sessions), and can generate an entire session, including inductions, visualizations, suggestions, etc. without external tools. I know this sounds like common sense, but most of the other software I've tried simply generates the tones; it's then up to you to figure out when inductions, recordings, etc. should start and insert them at the proper time with an audio editor. This is much faster, easier, and more flexible. There are many more features, but you get the idea. Included with the trial download is an ebook filled with information about entrainment. Even if you know you're not interested in the software, download it for the ebook. I know I sound like I work for the company, but honestly, this is a fantastic piece of software. I haven't purchased it; I don't have the spare cash right now, but as soon as I do I'm buying it. The only negative thing I can think of is the 15 day limit. The trial went by way too fast. If you're just starting to try entrainment, you're not going to know whether it's making any difference for you in 15 days. A standard one month trial would have been nice. Another thing; from reading through the forums, it appears that they're in the process of coding Neuro-Programmer 3. I don't know if registered users of version 2 can upgrade free to 3 or not; I know users of version 1 weren't able to upgrade free to 2. So you may want to hold off until they release 3. Apparently version 3 is on hold for now. No reason to wait.

Brainwave Generator $40, one month trial period - The oldest piece of entrainment software that I know of. Once upon a time, many moons ago, I purchased this program. I've used it on and off over the years with limited success. Has a gigantic library of user submitted presets, most of which, unfortunately, are complete garbage. Still, there are a few jewels to be found. Hasn't been updated in two and a half years, and people have slowly stopped submitting to the library. Only uses binaural beats and audiostrobe.

Be the First to Share


    • Backyard Contest

      Backyard Contest
    • Silly Hats Speed Challenge

      Silly Hats Speed Challenge
    • Finish It Already Speed Challenge

      Finish It Already Speed Challenge

    41 Discussions


    4 years ago on Step 1


    I made the glasses-only part of that some time ago, but
    am now trying to add the audio bit too, so that I can get a
    binaural/monaural beats track and the blinking LEDs from a single

    The LEDs work flawlessly as long as NO headphones are
    connected. Seems like the headphones drain most of the power and there
    is not much of it left for the decoder. I connect the headphones with
    a simple splitter (well, a prototype of it) plugged in before the
    decoder (just like in the author's circuit).

    Since I am still a
    newb I don't really know how to solve this. Do you guys think a simple
    audio amplifier, say something like this
    would help? Or am I going about this the wrong way entirely? There is
    no mention of an amplifier in this instructable, so I am guessing that
    the problem exists between the chair and the soldering iron.
    I power my circuit with a 9V battery passed through the L78 regulator.
    Would appreciate any advise.


    4 years ago on Introduction

    I'm having some problems getting the LED's to fade properly. They're jerky, unless I set the maximum brightness down so low that they don't turn off or on all the way.
    Here's their best behavior in NP 3, in the maximum brightness configuration.

    I did some tests with frequency sweeps, and found that the LEDs stayed on from ~400-19300 hz, with a complete drop off between 18250 and 18350 hz. I connected an oscilliscope to the lead between the transistor and resistor that goes to pin 8 (the signal output if I'm not mistaken), and played a signal sweep from 10000 hz to 19300 hz.

    Here's what the goggles did during that interval.

    Any possible diagnoses would be much appreciated, thank you!


    Reply 4 years ago on Introduction

    Correction: The oscilloscope measurement was between the *capacitor* and resistor, not transistor. Apologies, I'm new to electronics, and have been getting help with production so far.


    5 years ago on Step 1

    we have made low field transcranial magnetic stimulation helmet for all neurological problems with signal generator With 3 years warranty and support


    Frequency 0-30 Hz you can choose any frequency you like inbetween 0-30 HZ

    Intensity upto 25 microtesla

    Power 9 volts

    Wave form square wave



    5 years ago

    All Requencies i have found

    19.2 khz Audiostrobe Monocolor

    Spektrastrobe (Multicolor)

    18.7 khz Red

    19.2 khz Greenn

    19.7 khz Blue

    Please make a Spectra Strobe circuit. thanks


    Would the design work the same with 7-color LEDs for the glasses, and if not as is, what modifications would be necessary to drive the color shifting. I ask because unified glasses for AVS (L&S machines) can reproduce any color by encoded tone frequency, and operate each side independently


    6 years ago on Introduction

    @chantling, many thanks for your response. I am still trying to solve the puzzle.

    There is 5.06V between pin 4 and pin 7.

    1. I checked the frequency of the signal coming to the breadboard (yes, I am breadbording it) and it's correct, so there is no problem with the jack.

    2. I tried multiple things (sox, audacity) to generate simple sine waves. I tested the frequency using a multimeter and it's right.

    3. I have just checked them now. R1 and C1 seem close enough:

    R1 = 462 ohm.

    C1 = 105.5 nF;

    C2 = 7.4 nF. For some reasons all the capacitors from the 5 nF batch I got are around 7 nF. Will try to get some even smaller ones to increase the bandwidth.

    4. The exact chip is LM567CN. I thought it may be faulty, but my guess is that the problem is with my circuit. I have only recently started playing around with electronics, so I must be making some obvious mistake... I will keep digging...

    Thanks, really appreciate your help.



    Reply 6 years ago on Introduction

    Those are the right values. You can decrease capacitance for C2 by adding another cap in series, opposite to what happens with resistance. Two 7.4nF caps in series should yield 3.7nF, 3 in series will net you 2.467nF, etc. I skimmed over the LM567 datasheet to see what the formula is for determining C2 and bandwidth, but it uses input signal strength as a variable, so it's hard to tell exactly how much the extra 2.4nF affect the bandwidth. It's possible, given that higher capacitance there decreases the bandwidth gap, that your circuit is too picky.

    The LM567CN should work fine. I used an LM567N in my build. The letters at the end of a part number are typically a package indicator. They reference the difference between an 8-pin DIP, an 8-pin SOIC, etc. Note that this isn't always the case, so always check the datasheet in your other projects. Sometimes a different package has different input voltages or outputs. Usually a chip (LM567) family all reference the same datasheet and there's a table in the datasheet showing the difference between the various child products. In the case of the LM567, for example, there are a few (very) small differences in power usage, performance at min and max voltage ranges, etc.

    If you drop the capacitance and it still doesn't work, my best guess is a faulty LM567 or possibly low input signal. I can hear the Audiostrobe tone (although as a 35 year old male I shouldn't be able to, so maybe my sound cards are always generating noisy signal and I'm hearing the noise) when it's playing by itself; if you hook up a pair of headphones to the audio out, can you hear the signal? I can tell you that there's a huge volume difference required to drive the circuit depending on the source and what headphones you've got hooked up. You calibrate the brainwave generator with the circuit plugged in to the computer so the audiostrobe circuit works correctly at the volume you're planning to use it. Then you record the session, export it to an MP3 player and it doesn't work without turning the sound up to uncomfortable levels, so you've got to go back in and recalibrate it. Bit of a pain, really. But anyway, I'd try increasing the volume as high as you can before calling it a bad chip.


    Reply 6 years ago on Introduction

    You were right, it was the volume, I turned it up and it works now! The chip seems to only respond to frequencies between about 17 and 20kHz, which is good enough for me.

    Looks like I can go ahead with the glasses now.

    Thanks a lot!


    6 years ago on Step 1

    alikasundara - you shouldn't need the 7805 if you're powering it via usb. That's just so I can use a wall wart that isn't 5V. The LM567 datasheet says it'll work from 4.75V to 9V, so it's not terribly picky. Use a multimeter to verify a 5V differential between pin 4 and pin 7; if so, the chip should be powered. If it's got power, the things I can think of that would cause an issue are:

    1. Bad connections, especially at the stereo jack you're using for an input. Assuming you're breadboarding the circuit, I'd bypass any jack you may be using and wire directly to the pins until you've verified the circuit is working as intended.
    2. Bad input signal. What are you using to generate the audiostrobe
      signal? If you generate an audiostrobe test signal and record it using a
      loopback, you can use Audacity's (free sound software) spectrum
      analyzer on the sample. You should clearly see the 19.2khz signal. For
      that matter, if it's just the Audiostrobe signal, you should be able to
      see the audiostrobe pulses in the waveform in the normal view.
    3. Incorrect values for R1, C1, or C2. R1 and C1 set the center frequency, C2 sets the bandwidth. None of them are super critical, but if you got one of the units mixed up (say, 5uF for C2 instead of 5nF) it's not going to work correctly.
    4. Bad LM567. If everything else is good, this is a possibility. Taking a quick look at the datasheet, it looks like the smaller C2 is, the greater the bandwidth. So you could try replacing the 5nF with, say, 1nF or even smaller. That should open up the bandwidth to the point where pretty much any audio frequency should trigger it. If you've checked all of the above and you're sure you're getting good signal to the chip, then it may be bad.

    Good luck!


    6 years ago on Step 1

    Hi there,

    Just wondering how important it is to use the 7805? I am powering this from my USB (5.22V) and getting the low output no matter what input frequency. Something is clearly wrong with my circuit and I am trying to figure out what...



    8 years ago on Step 9

    Hi im using gnaural opensource software, i add a new isochronic tone voice but in the base freq i set to 19200 hz, i can see the beat in the spectro with adobe audition,this work with your audiostrobe?


    12 years ago on Introduction

    AaaaaaHA! Curse you evil brainwave entrainment glasses!! You are going to force me to post an Instructable for aluminum foil hats and reflective mylar glasses to reflect your evil LEDs and brainwave entrainments, things. Take THAT you pesky hypnosis-izers, aaaaaahahahahahahaha. I also have a 12-step Instructable ready for recycling an Altoids tin into a crush proof portable candy or cuff link holder. Otherwise, the technology in this is pretty cool, and you have paid very good attention to the small details.


    Reply 10 years ago on Introduction

    Do you have filter-glasses for safely watching TV?


    Reply 12 years ago on Introduction

    Ah... But wide bandwidth hats? I wouldn't be satisfied with an aluminum foil hat that only filters out the AM band... ;)