Are you sure you want ~40Hz? My device is more like 0.5Hz. If I recall correctly, a TENS machine can hit 40Hz. You can get a pretty nice one for ~$30. It does NOT supply a micro current though, and uses voltages over 100V. Definitely not for tongue use.

My circuit should work fine at 40Hz, and limit the current to 600µA or less, but you'll need different timing components on the 555 (or 556). If you want to set the current limit higher, that shouldn't be hard either but it would be a good idea to work it out in simulation to be sure. A 9V battery across a human head my not be able to drive more than a few mA anyway.

 

thanks for publishing this wayneh--nice work! interesting idea of using the 4017 the generate a reversing waveform. this is similar in function to several of my microcurrent devices including the acuscope. I can vary the frequency between a range of 0.5 - 380 hz which seems to make them useful to concentrate to local areas with the higher freqs or spread through the body with the lower freqs. what was the mod that you mentioned to do this? -regards

 

Thanks for posting this wayneh, it looks very interesting!

I'm trying to reproduce this build, but I run into some problems of getting the circuit to work. I'm confused by how to connect the dual gang pot which is connected to the 3K R10 and R11. The pot has 3 connections (1, 2 & 3). In your drawing 2 and 3 are bridged together, but there is no further connection from point 1. I suppose it should be connected to GND, is that correct? (I tried to connect to GND, but this still doesn't work unfortunately).

 

Thanks for posting this wayneh, it looks very interesting!

I'm trying to reproduce this build, but I run into some problems of getting the circuit to work. I'm confused by how to connect the dual gang pot which is connected to the 3K R10 and R11. The pot has 3 connections (1, 2 & 3). In your drawing 2 and 3 are bridged together, but there is no further connection from point 1. I suppose it should be connected to GND, is that correct? (I tried to connect to GND, but this still doesn't work unfortunately).

Right, the opposite pole is grounded. That’s show in one of the drawings but was omitted from the one that’s confusing you. The wiper is connected to the end opposite to ground to limit the impact of the wiper temporarily skipping off the resistive material as it moves. The pot goes th max ohms instead of infinite ohms.

Sorry it’s not working. How are you testing it?

 

Right, the opposite pole is grounded. That’s show in one of the drawings but was omitted from the one that’s confusing you. The wiper is connected to the end opposite to ground to limit the impact of the wiper temporarily skipping off the resistive material as it moves. The pot goes th max ohms instead of infinite ohms.

Sorry it’s not working. How are you testing it?

Thanks for the reply.

It's probably due to my incompetence with analog electronics. As a first step, I double-checked all traces and connections in the circuit for shortcircuits and proper connections. Now I'm using a scope to see if there is a waveform visible in any part of the circuit. However, it seems all the outputs are stable, no waves generated. To be honest, I'm not very certain how to start with properly debugging it.

 

Thanks for the reply.

It's probably due to my incompetence with analog electronics. As a first step, I double-checked all traces and connections in the circuit for shortcircuits and proper connections. Now I'm using a scope to see if there is a waveform visible in any part of the circuit. However, it seems all the outputs are stable, no waves generated. To be honest, I'm not very certain how to start with properly debugging it.

Divide and conquer. Check that you have power at each IC. Check the input and output voltages. You might want to break the connections between ICs to isolate the various stages. Make sure the timers are working and move forward from there.

 

Divide and conquer. Check that you have power at each IC. Check the input and output voltages. You might want to break the connections between ICs to isolate the various stages. Make sure the timers are working and move forward from there.

It took a while to report back, but I finally managed to get the build working! (and learned a whole lot about analog electronics)

One problem I ran into is that the diagrams of the PCB traces are a bit different than the actual circuit schematics, after changing these two traces my issues were mainly solved:
- The 1reset pin from NA556 should go to + rail, the 1Cont shouldn't be connected to +rail
- The 2reset pin from NA556 should also go to + rail, the 2Out shouldn't be connected to +rail

If I'd want to change the frequency of the pulses, would you recommend to change the 45 µF capacitor (U5)? For example, for a higher frequency, use a smaller capacitor?

 

One problem I ran into is that the diagrams of the PCB traces are a bit different than the actual circuit schematics, after changing these two traces my issues were mainly solved:
- The 1reset pin from NA556 should go to + rail, the 1Cont shouldn't be connected to +rail
- The 2reset pin from NA556 should also go to + rail, the 2Out shouldn't be connected to +rail

The reset pin of the 555 timer is internally pulled up, so the lack of those connections doesn't matter. The simulation shows direct connections of the reset pins to V+, but it's not really necessary.

You're right about the other errors and I think both were caused during the drawing process. Those connections to V+ may well have been the ones meant for the reset pins but then I shifted the IC position without fixing those. I need to check my builds to see if these errors made it into them. They likely did! Thanks for taking the time to find and report the problem.

If I'd want to change the frequency of the pulses, would you recommend to change the 45 µF capacitor (U5)? For example, for a higher frequency, use a smaller capacitor?

U5 is an opamp, so I'm not sure what you are asking. C2, R1 and R2 control the main timing. You could get quite a bit of frequency range by using a pot there, and if fact my earlier version used this. For this later build, I decided the user doesn't really need control of the frequency.

Here are corrected pictures of the backside solder traces and the frontside views. [Update] The wiring drawing is updated to incorporate corrections identified in post #61.

 

So FSM stands for Frequency Specific Microcurrent - Machine.

And I was using the tongue as an example because the resistances just about everywhere else are so high I'd need to boost the voltage first to produce even microcurrent; I think. Also, there is an amazing book called The Brain's Way of Healing where they use microcurrent stimulation on the tongue to cause the brain to self correct after traumatic brain injuries, stroke, parkinsons, and all sorts of movement disorders.

Is it possible to send private messages here? My posts are still being moderated, and I'd love to respond to you more quickly, but it will likely take a bit before they go live in the thread.

hello mate
i am very interested to know how your project went? did you manage to build the machine. there is a lot more info on the successful wave forms and methodology now becoming public. for some reason it seems the original creators of this tech have been very open about the inner workings and the tweaks that gave the best results. i will post those links in a reply to wayneh's comment at the beggining of this thread. please let me know where you got with this mate. thanks

 

The first design of this I made, I included a frequency control and also TENS capability. With usage it became clear that neither is really useful. TENS units are cheap and have lots of features, and address a completely different need. There was really no need to adjust the cranial frequency of this device because no one can be sure what setting to use. It works fine at a fixed (but varying) frequency.

i found these latest results for the most effective frequencys used in neurostimulation

https://jneuroengrehab.biomedcentral.com/articles/10.1186/1743-0003-11-96
The Portable Neuromodulation Stimulator (PoNS)™ is a small electrode array (3 × 3 × 0.1 cm, and 100 g) that is held in place on the tongue’s surface with light pressure to the roof of the mouth (Figure 1). The stimulation consists of 19-V pulses delivered at a rate of 200 Hz with every fourth pulse removed[29]. The 143 electrodes are pulsed sequentially in groups of nine. Subjects were instructed to increase stimulation to a moderate-high level (pulse width adjustable from 0.4 to 60 μs) that was tolerable and not painful.

https://www.resna.org/sites/default/files/legacy/conference/proceedings/2013/Outcomes/Liegl.html
PoNS™ device uses an unbalanced biphasic waveform designed to ensure net zero current (Kaczmarek, 2011) to reduce the chance of tissue irritation and has 19 V max and 6mA operational limits. Pulses are delivered to the tongue in triplets of pulses at 5 ms intervals every 20 ms. The subject can control the pulse-width (0.4-0.6 µs) by adjusting the intensity buttons on the device. The buttons on the device allow the stimulation to be turned on and off and increased or decreased in intensity. Each time the device is turned off, the intensity resets to the lowest level

and most tellingly a patent for the device states....

In practice, stimulation which delivers repeating pulses in the nature of a simple waveform, e.g., a square wave or series of regularly-spaced pulses, has been found to be effective. Also preferred are “bursts” of pulses which repeat at some frequency, with the pulses within the bursts repeating at a higher frequency (for example, as if the peaks or “on” periods of a square wave or other waveform were themselves formed of a series of pulses). The fundamental or harmonic frequency underlying the stimulation waveform may be chosen in dependence on the particular application. As examples, testing and/or inference suggest that the appropriate frequency for sensory dysfunction might be around 40 Hz; for Parkinson's disease, around 30-50 Hz; for involuntary movement/tremor, around 1-40 Hz; for voluntary movement coordination, around 50-100 Hz; for cognitive processes, around 100-300 Hz; for bradykinesia, around 150 Hz; for sleep or anesthesia, around 80 Hz; and for relaxation or wakefulness, around 0.5-2.0 Hz. In some applications, it may be most effective to have the delivery frequencies of certain electrodes (or other actuators/elements for delivering stimulation) differ in accordance with their location, e.g., electrodes in one area may deliver stimulation at one frequency and electrodes at another area deliver stimulation at another frequency (wherein the frequencies need not necessarily have a harmonic relationship).

https://www.google.com/patents/US9020612

 

please help

 

1) Nathan: Help with WHAT???
2) By the way, I see here a rather close topic as those RIFE software (freeware) generating the microampere currents of various frequencies for healing aims. Last time I had see it at https://www.dropbox.com/s/oxq0pjailnap6eq/rife.zip?dl=0 . Feel free to read the text files where You see the ideology behind it, and frequency listings with specified target illnesses (however I never had read any work certifying that frequency list is not `sucked out from finger`). But that theory sounds quite logic.
3) I have read the various articles on topic of active and reactive components at conductivity of milk, blood and other animal body liquids, and surprizingly, there are obvious a certain quazi-resonance type of impedance graph by a frequency. So, it is clear that some frequencies makes a larger impact on biology than other frequencies.

 

Hi,
This post is very interesting to me. I design consumer electronics products (mostly audio stuff) but have strayed into the "TENS" unit's in the past. I worked with a company in China to produce a hair removal unit based on a xenon flash lamp next to the skin. Yes, if you put a flash lamp from a camera (the "real kind", not the one built into disposable's) and put it on your skin it will blast the hair away. I was tasked with getting the thing working, as well as the algorithm for skin tones. I digress...

Though the circuit is very simple, I can't help but ask, why not use a microcontroller for all the timing stuff? If you're not already using one or familiar with them, a project like this is a great way to start; all the "programming" is really just setting up timers, which most uC's have many of. My personal favorites are from Microchip, and they have a new little "dev board" with a built in programmer; you just plug into a USB from your computer and begin coding. Once you get the hang of it, you can make these things do all of the stuff mentioned throughout this post, plus (eventually) adding feedback, temperature measurements - whatever you can think of. Just food for thought...

My services for design and layout of circuits and PCB's is always available. Drop me a line to know more.

Mike Tripoli

 

...why not use a microcontroller for all the timing stuff?

I would ask the converse: Why use a µC when you don't need to? I admit that, if I was already familiar with one of them, I likely would have chosen to go with that. But I knew it could be done with simple 555 and opamp circuits, so the incentive to climb the learning curve wasn't there.

 

I would ask the converse: Why use a µC when you don't need to? I admit that, if I was already familiar with one of them, I likely would have chosen to go with that. But I knew it could be done with simple 555 and opamp circuits, so the incentive to climb the learning curve wasn't there.

Hi Wayne,

You ask a great question. Normally, when I see any project (admittedly, because I've been using uC's for about a thousand years I think of ways to implement it with a uC (please keep in mind, I make my living designing and manufacturing ANALOG circuits, preamps, amplifiers, guitar effects and amps, "tube" designs through to Class D,G, H and Z) simply because they offer just about everything one could want or need; built in A/D, D/A, opamps, comparators, PWM, multiple timers, etc.etc.

When doing anything where timing is at the core of the design, a uC, IMO, is the go to device. Whereas in a 555 design, if you want to change timing, you sub-out components (I'm going to completely skip over the issues with 555's; they change freq. with voltage, getting a "real" square wave from one requires diode tricks, parasitic capacitance makes the things jumpy at higher freq's, etc.). With a uC, you literally change a couple of values in the timer registers and, just like that, perfect PWM. Want to make it variable, add a pot going to the A/D, that value updates the timer register. Once you try them, you can get hooked very fast; the number of projects becomes infinite.

Admittedly, there is a learning curve, but that is the same with anything new, no? You had to learn the topology of the device your currently using, say, the 555. If you start with (I'm not a shill, I'm just mentioning the tools I use) something like PicBasic from meLabs (www.melabs.com) you can "inch" your way into using them. There are so many development boards available it's crazy. They have an absolute ton of reference programs as well, that in many cases, one can cut-and-paste to get something going. I always say, make your first project for a uC exactly the same as you would have when you started with discrete components; blink an LED. Just about everyone at one point or another has done that with a couple of transistors, resistors and caps (the old astable osc). Then you most likely blinked an LED with a 555. Doing the same thing on a uC, you quickly realize that you've set up a timer, toggling the output pin as intervals. That is the heart of the OP's project.

The first TENS unit I built was in the 80's for my Dad. It used a 556, a MOSFET, a handful of passives and a transformer. Now, I can do the same thing with one 8 pin PIC. There are also 6-pin (SOT-23 if you do SMT) uC's that are perfect as little "everything" devices. I use them like jelly-beans; I never worry if I can "find the right IC" for something, I drop in a PIC and it's whatever I want it to be.

Sorry for the long reply; whether its this or another project, I encourage you to look at what's available. My advice (for what it's worth) is to ignore things like "Arduino's" at first, and do some BASIC on a PIC. You will never regret it...

B'rgds,
Mike Tripoli

 

Just one more thing...

I have a friend, a mechanical designer that I met though work, that designs large conversion machines; now his company manufactures "toppers" for the food industry. When we met, his machines all used PLC's to do everything. I was involved to modify their machines for use in manufacturing my product. As I went through the machine, there was a bunch of very simple stuff all being done with PLC's. I'm not a PLC guy, (and never will be, frankly) and reduced a bunch of the PLC's down to, at the time get this, the first BasicStamps from Parallax. The functions were things like "look for a switch closure then assert a data line" which is a couple of lines of code in BASIC. At first, he thought I was nuts, and we went toe-to-toe a few times about design decisions. Ultimately, he saw that these things were useful... That was about 15 years ago; today, he has surpassed me, and programs embedded uC's (PIC 18 and 24 devices in "C") which I don't do. When he "retires" he wants to start a business doing nothing but embedded controls on uC's.

Have a look at this: https://www.parallax.com/product/555-28188

B'rgds,
Mike Tripoli

 

Just one more thing...

I have a friend, a mechanical designer that I met though work, that designs large conversion machines; now his company manufactures "toppers" for the food industry. When we met, his machines all used PLC's to do everything. I was involved to modify their machines for use in manufacturing my product. As I went through the machine, there was a bunch of very simple stuff all being done with PLC's. I'm not a PLC guy, (and never will be, frankly) and reduced a bunch of the PLC's down to, at the time get this, the first BasicStamps from Parallax. The functions were things like "look for a switch closure then assert a data line" which is a couple of lines of code in BASIC. At first, he thought I was nuts, and we went toe-to-toe a few times about design decisions. Ultimately, he saw that these things were useful... That was about 15 years ago; today, he has surpassed me, and programs embedded uC's (PIC 18 and 24 devices in "C") which I don't do. When he "retires" he wants to start a business doing nothing but embedded controls on uC's.

Have a look at this: https://www.parallax.com/product/555-28188

B'rgds,
Mike Tripoli

aaaahhhh what to do! i was hoping to learn arduino a bit better ( i made the knight rider light so far....) but a pic seems a little more compact? a good use for wearable products it would seem. food for thought

1) Nathan: Help with WHAT???
2) By the way, I see here a rather close topic as those RIFE software (freeware) generating the microampere currents of various frequencies for healing aims. Last time I had see it at https://www.dropbox.com/s/oxq0pjailnap6eq/rife.zip?dl=0 . Feel free to read the text files where You see the ideology behind it, and frequency listings with specified target illnesses (however I never had read any work certifying that frequency list is not `sucked out from finger`). But that theory sounds quite logic.
3) I have read the various articles on topic of active and reactive components at conductivity of milk, blood and other animal body liquids, and surprizingly, there are obvious a certain quazi-resonance type of impedance graph by a frequency. So, it is clear that some frequencies makes a larger impact on biology than other frequencies.

huh.. all that writing and i never did clarify what i needed help with , how odd now i look back over it. thank you for pointing that out. i was asking for help to build the device in another thread and i was linked here and forgot to recap my mission for the new thread users but in hindsight i should keep my topic in my thread so as not to dilute this one. the conversation is here:
https://forum.allaboutcircuits.com/...tongue-stimulator-device.140859/#post-1188221

 

Hi,
Nothing wrong with an Arduino, a BeagleBone, various "pi's" - all of them have a place depending on what you want to do. I've been using PIC's since the first "C54's"; I'm very familiar with the PIC paradigm, so I go with with I know. That being said...

I took a quick glance at the other link, stopped when I saw the word "biphasic". At one point in my career, I worked at a company designing wearable heart-defibrillators. These things (defibrillators) are more than just a "big electrical charge"; the waveform is "shaped" as to do the least amount of damage to the heart muscle; a "biphasic" waveform being one method (there are many). I was in charge of working on the output section, IGBT's were just coming onto the scene, and I did many tests using them. I needed to be able to adjust the waveform during test, so I used, you guessed it, a PIC. The actual unit used a DSP to do all the waveform generation; these were "canned" routines and I needed to change them for different tests. Anyway...

I would completely abandon any kind of "discrete" solution (like 555's) and jump on a µC of some kind. "Granularity" is going to be the keyword here; take whatever your fastest pulse is; that becomes the minimum for anything you do from there (after all, a 10µS pulse from a timer is just a 1µS delay repeated 10 times). Once you've decided on this, then decide the best choice for your platform. You may find that an Arduino is fine, or, you might wind up with a FPGA being clocked at 100MHz...

Good luck.

Mike Tripoli

 

Introduction
This is a project to build a micro-current electrical stimulation device. These are also called tDCS devices and are similar to MENS devices. They are not much like the common TENS devices which use much higher voltage pulses.

I have posted about this project before here. Since then I've built a couple of these devices for friends and they've been very pleased. So I thought I should update the circuit and post the complete information here.

A tDCS device provides a constant current from zero to 600µA applied the the earlobes of the user. That alone can produce a therapeutic effect, but it's more effective if

1) the current is pulsed at ~0.5Hz,
2) reverses direction (this is an AC device), and
3) the pulse timing varies "randomly".​

The changing frequency feels somewhat random and prevents a "burn in" in the patient. It's more effective than a steady beat. This is a probably a small refinement, but I wanted to include it.

I based my project off this device, described in patents US2010145410, US2010047834A1 and others. I also added TENS capability in my first unit since most of the circuitry was already in place to easily add this capability. I have since removed the TENS circuit. It's a completely different treatment and there is no reason that someone using tDCS will necessarily care about using TENS. And, you can buy very elaborate and nice TENS units for ~$30. No need to build your own.

Circuit Overview
The strategy for my circuit was to use a dual timer 556, with the first timer slowly changing the voltage on the second timer's control pin to change the timing. The clock pulses from the second timer go to a 4017 counter to create the proper waveform. Finally, the waveform drives a dual op-amp configured to provide a sort of H-bridge. An adjustable constant-current is applied to the output earclips.

Schematic and Description
View attachment 110482
The voltage on the timing capacitor C5 of the first timer U2 feeds an op-amp voltage follower U6, and then the voltage is amplified by a 2nd op-amp U7 to make a ~30-second triangle wave from rail-to-rail. It's a crude triangle, since it's really just the middle of the timing capacitor's RC curve . This is applied to the control pin of the second timer U1, which would otherwise be running at ~0.5Hz. Pulling the control pin high and low changes the frequency over a ~4-fold range from 0.25Hz to 2hz. (It also changes the duty cycle, but that's not an issue in this circuit.)

The 2nd timer's output is fed to a 4017 counter U3 and the "0" and "2" outputs are compared. This provides a way to alternately remove and reverse the direction of current flow. Using the "2" pin against the "0" pin as outputs allows the current to reverse direction with a dead space in between when the "1" pin is hot and the 0 and 2 are both at ground. The pulse on the 3 lights an indicator LED and the counter resets on 4.

Output from the counter controls the state of a dual op-amp controlled-current circuit. The circuit works with an LM358 for U4 and U5, but the simulation looks better using a TLV2721. The LM358 can't get as close to the ground rail, and the simulation shows the "dead space" intervals as reduced-current shoulders. But the current never goes to zero as it does (in simulation) with the TLV2721. Again, I built two successful units using the LM358. I've switched to the TLV2721 for later builds.

The current intensity into the load R8 is adjusted using a dual gang pot, since there are essentially two channels, plus and minus. R10 and R11 in the simulation are actually each a 3KΩ fixed resistor and a 10K variable resistor in the pot. The minimum of 3K results in an output intensity of ~0µA. At the maximum of 13K total, the output reaches 800µA, a little higher than the target of 600µA but still quite safe.

View attachment 110485

Parts List
Following is a recent Mouser order for most of the required parts
View attachment 110486
View attachment 110487

I had a few parts, like the LED, on hand already and so there may be a few parts missing from the order. Let me know if you find one missing and I'll add it.

Here is where I ordered some cheap ear clips on e-bay http://www.ebay.com/itm/331870106853
I've also ordered from Lhasa Ohms in the past:
Ear clips: http://www.lhasaoms.com/Ear-Clip-Electrodes.html
Wire leads: http://www.lhasaoms.com/Lead-Wires-with-3-5-mm-Jack.html

Build
I use these 4x6 cm prototyping boards to build the circuit on. Standard fiberglass with 0.1" hole spacing. It's a little tight on the 4x6. Use a 5x7 if you want extra space.

My circuit design is done is drawing software. I use Intaglio. I can supply the original Intaglio file to anyone that asks but I have not attached it here because I suspect there are few other users of that software here. Instead I'll post the various layers.
The board:
View attachment 110493

The board plus solder traces
View attachment 110492

Plus components
View attachment 110491
Plus wiring
View attachment 110489

Detail of the output and adjustment pot
View attachment 110488

The backside solder traces, viewed from the back
View attachment 110494

See the project linked in the introduction for a photo of my first complete build. I'll be making more of these soon and will post more detailed photos.

Hi, as I am very interested in tDCS research, would you be able to point me to more information/studies looking at the most effective stimulus. I am very interested in research supporting 0.5Hz pulsing, randomisation of the pulses (certainly this would be supported by the theory of homeostasis) - but I know no study supporting this, and AC versus DC. Your help would be greatly appreciated.
Kind regards
Michael

 

Hi, as I am very interested in tDCS research, would you be able to point me to more information/studies looking at the most effective stimulus. I am very interested in research supporting 0.5Hz pulsing, randomisation of the pulses (certainly this would be supported by the theory of homeostasis) - but I know no study supporting this, and AC versus DC. Your help would be greatly appreciated.
Kind regards
Michael

Start at the Alpha-Stim website. They have a bibliography there.