Attached is my schematic for the Beck Brain Tuner. I've been using it somewhat successfully for insomnia problems related to my other health issues. It also helps a little with my suffocation migraines but doesn't fully eliminate them. Beck's claims are listed below.
I don't anticipate much in the way of follow up posts, but I'm going to have to insist that posts stay on topic (which this site is fairly good at). If you want a long discussion, please start a new post and post that link in this one. If industry trolls find this post, I'll ask for your accounts to be locked. I don't want this turning into overunity.com with a hundred pages of arguing nonsense. If you're new to soldering and need help, this isn't the post to ask. There are literally thousands of soldering videos on youtube that will show you far more than I can here. Otherwise, enjoy.
Digi-Key didn't have any transformers I liked, so I got the TY-144P (mentioned below) from Mouser. A signal isolation transformer with similar specs should work just fine. It shouldn't cost much more than $5 (in 2023 USD). Since this is intended to be a medical device plugged into your brain, you shouldn't be buying your electronics parts from fleabay or amazon. 555 timers can get into the low RF range. You don't want RF plugged into your skull, so check your parts carefully.
Most of my prototype unit (pictured below) was built from parts I had leftover from other projects. I don't know the exact cost, but it won't be very high. I haven't mounted it in a project box yet, but I may do another Frankenstein bottle since I have plenty in the recycle bin. I did the math for the timers (shown on the PDF) and used fixed resistors instead of trimmers (which is why you don't see any). When you measure each part value for a timer, I've found the math equations to be pretty accurate. Make sure you clean off any leftover flux residue from the timers. Flux is mildly conductive and will change the frequency settings. For critical parts, I prefer to flux, tin, then wipe them off before soldering them to the protoboard.
You can run this circuit higher than the 9v shown, but it won't get you too much more. If you exceed 15v, you'll blow the timer chips. If you use an unregulated wall wart, you should add a voltage regulator to the power block and not use batteries. An LM317 or LM7809 (or 7812) are easy to find and use and should work OK. Below 7v (a dead 9v battery), this circuit seems to lose its effectiveness.

This is copied from my write up to make it searchable:
I don't anticipate much in the way of follow up posts, but I'm going to have to insist that posts stay on topic (which this site is fairly good at). If you want a long discussion, please start a new post and post that link in this one. If industry trolls find this post, I'll ask for your accounts to be locked. I don't want this turning into overunity.com with a hundred pages of arguing nonsense. If you're new to soldering and need help, this isn't the post to ask. There are literally thousands of soldering videos on youtube that will show you far more than I can here. Otherwise, enjoy.
Digi-Key didn't have any transformers I liked, so I got the TY-144P (mentioned below) from Mouser. A signal isolation transformer with similar specs should work just fine. It shouldn't cost much more than $5 (in 2023 USD). Since this is intended to be a medical device plugged into your brain, you shouldn't be buying your electronics parts from fleabay or amazon. 555 timers can get into the low RF range. You don't want RF plugged into your skull, so check your parts carefully.
Most of my prototype unit (pictured below) was built from parts I had leftover from other projects. I don't know the exact cost, but it won't be very high. I haven't mounted it in a project box yet, but I may do another Frankenstein bottle since I have plenty in the recycle bin. I did the math for the timers (shown on the PDF) and used fixed resistors instead of trimmers (which is why you don't see any). When you measure each part value for a timer, I've found the math equations to be pretty accurate. Make sure you clean off any leftover flux residue from the timers. Flux is mildly conductive and will change the frequency settings. For critical parts, I prefer to flux, tin, then wipe them off before soldering them to the protoboard.
You can run this circuit higher than the 9v shown, but it won't get you too much more. If you exceed 15v, you'll blow the timer chips. If you use an unregulated wall wart, you should add a voltage regulator to the power block and not use batteries. An LM317 or LM7809 (or 7812) are easy to find and use and should work OK. Below 7v (a dead 9v battery), this circuit seems to lose its effectiveness.

This is copied from my write up to make it searchable:
This schematic is a modification of the Brain Tuner tear down schematic floating around Internet and fixes some of Dr. Beck's design problems. Beck claims it will help with addiction, insomnia (sleep related), stress, anxiety, and depression (there might be a benefit to other mental problems). People with epilepsy probably shouldn't use this. License is free for non-commercial use. No warranties expressed or implied.
The original parts list came to a little over $50usd (not including tax and shipping), yet I've seen these things sell for $500 (for a simple multiple frequency circuit). This design is very similar to a power inverter and not too far off from a TENS unit. Each part in the timers should be metered to know its true value. The resistors and capacitor forming the oscillator should have tight tolerances if you want accurate frequencies. If you have an oscilloscope or frequency counter, you can make up for the remaining tolerances with the trimmers. The rest of the parts should be close to listed values but aren't critical. Remember that resistors add up in series. If the capacitor is too small, capacitors add up in parallel. A larger capacitor or resistor will slow the 555 timer. A number of independent studies have shown that frequencies aren't that magical or important, so close may be good enough, and 555 timers (RC oscillators in general) are NOT known for their accuracy. They will drift with temperature. The recommended use time is 20-40min, and beyond 45min diminishes the effects.
The timer variable resistors should be sealed multi-turn trimmers and board mounted. The trimmers can be replaced with fixed resistors, thus increasing reliability a little. RVct is Output Intensity and should be a wire wound pot that is case mounted. Tiso has some sharp inductive spikes that some people with sensitive skin will find uncomfortable. RVct can be increased up to 4k7 in this case.
All the unmarked diodes are 1n4148's. Do1-3 are used to turn off the Qp1 high side switch and prevent it from being on at the same time as Qp2. Dp1-2 are needed for the status LEDs and to help protect the transistors from sharp inductive ringing. Dpwr can be a 1n4001 to a 1n4007 or similar.
U2's frequency is fixed at 4Hz, and I removed unnecessary parts to lower the costs. If you want this variable, add a 50k pot in series with Rb3 and lower Rb3's value a little. Frequency range of U2 is 2-6Hz.
I tested 4 isolation transformers for Tiso. The original one in the tear down schematic lowered the voltage and worked poorly. I eventually settled on a TY-144P from china (about $5 at the time of writing). Beck wanted a lot of harmonics for "extra" frequencies, and china transformers are noisy. Beck gets these "extra" frequencies from Tiso also being a big inductor that is "rung" by the square wave pulses, thus Tiso should NOT have just a few wraps of wire on the primary and secondary. If that model isn't available in the future, get an equivalent center tapped with 15k:15k impedance, roughly 1k:1k DC resistance (going down to 500:500ohms would probably work). The transformer should be close to 20mm wide. I tried winding my own transformer, but it didn't work out. There should be about 2000 windings per side, and this is difficult to do at home with less than hair thin magnet wire. Unfortunately, this many windings will unbalance the output waveform (the primaries aren't wound in parallel), but it still works.
I generally prefer not to use 555's in my designs because of the pin locations making a wiring mess, but this is what was common and available 40-something years ago. If I were to design this on my own and keep the various frequencies, my next unit will be MCU based. Beck has mentioned that the pulse duty cycle should be less than 10%, so this is probably an early version of his design. Duty and frequencies are much easier to change on the fly with an MCU than a 555, not to mention that an MCU has far more accurate timings.
Nothing in this design should run hot. My initial test prototype would use 20-30mA depending on 9v NiMH charge. I increased some of the resistor values to help lower the current more and be battery friendly.