Author Options:

How To Create Frequency Specific Microcurrent? Answered

Curious if some of the electronic gurus might have the answer to this.

I am looking to build a two channel (A/B) microcurrent (allowing from 20 - 600 microamps) with selectable frequencies ranging from 3 to 970 Hz). This is based on a medical device that I'm looking to replicate. I've figured out some of this but again, considering the wealth of knowledge here I'm sure some would know the best way to build this. Oh yeah, I figure using a rechargeable battery which could be whatever voltage is best recommended.

So what do you think???



2 months ago

I know this is an old post, but has anyone built a FSM machine successfully since?

I don't know of any (microamp) FSM having been built.

Currently I'm back tracing a quad multi program Tens (milliamp) like machine in my spare time.

You just joined what is your skill set and application (only curious :-)

Oh I’m only a hobbyist. I’m not the most experienced either, but I’ve built a few things in the past (stereo speakers, radio guided toy car, a bob beck machine (which is a much simplified version of FSM I guess)... I like to open my broken appliances for fun and see if there’s anything I can do to make them work again)

As for my interest in FSM, I have a condition, nothing life threatening but some days the symptoms are so bad that I just want to die. I’m at the point where I just try whatever to see if I can feel better, even for just a while.

I too have a condition.. Mine is peripheral neuropathy and lately i have tried a Capsaicin slave which due to the nature of of the hot pepper overwhelms the nerves that cause pain to the point of serene non-feeling for hours of time.. Hope you can find relief..

Attached is a very simple to make Low Power "Tens" circuit for you...


Thanks! I'll have a look


8 months ago

Here are illustrative pictures, first four are Clivite Spark Pump..

p16.jpgP000.bmpp2.jpgp7.jpgBlock charge.jpgpiezo power-2-25-2016 007.jpg

9 months ago

3) I think I understand what you meant: do you mean that since you have the RLim resistor in series with the load, any variation in skin resistance will be negligible? If so, do you think 75 K is enough? My understanding is that skin resistance can be up to 1M.

Be creative measure your own skin resistance..

I did threshold of feeling AC current between two fingers..

Most people in my office could detect 1/2 to 2 ma current between adjacent fingers..

With an electrician who had to wet his finger to stick it in a lamp socket to feel anything to my Japanese friend who could detect 50 uA AC.


8 months ago

2) I am designing a PCB to deliver FSM and I have a 9VDC supply, so in order to drive 45VDC I thought I would need an SMPS from 9 to 45

3) Just to check that I understood you correctly, do you mean that Rlim in series with the high output resistance of two BJTs is high enough that makes any change in the load negligible?


9 months ago

1) The circuit is being changed from 5V to your 9V. The CMOS inverter has a lower drive current, putting 3 gates in parallel will guarantee you drive the isolator LEDs properly...

2) This is a microcurrent device and I realty don't know what your SMPS is designed to do..

3) You understand the application correctly... The NPN Vce is limiting the high voltage and high resistance to properly make all skin resistance values compatible..

if you go and drive a mosfet bridge you could reach 300V and a comparably higher series resistor...

BTW the sticky electrode pads do reduce you finger high skin resistance substancially.


9 months ago

Thank you iceng for posting!

I have a couple of questions

1) you wrote “Today I would change the 150 ohm to 330 ohm using 3+3 parallel CMOS gates at 9VDC....” Why would you do that? What is your reasoning?

2) Assuming I use a boost SMPS to get my 45 volts, how much current can I draw from it safely

3) I’m no expert, but I would think that one problem with applying FSM to the human body is that the skin resistance has a very large range. Can we engineer some kind of feedback into the driver so that at the very least, a consistent DC is always running through the load, between the load being say, 15 K and 1 M? (It would also be nice to make sure the AC also has a constant amplitude). Although, I think you meant to clarify this by writing “Remember the high voltage and high resistance acts as a constant current regulator for skin resistance variations”

Thank you so much!


9 months ago

DavidG asked me by PM for a driver...

Here are two drawings (you need to click to see the whole image) of the same schematic, whichever is easier to visualize.. The LOAD is your body which gets controlled low current AC and uses a single Quad opto-isolator.. Today I would change the 150 ohm to 330 ohm using 3+3 parallel CMOS gates at 9VDC....

Remember the high voltage and high resistance acts as a constant current regulator for skin resistance variations.. I don't remember the NPN emitter collector withstand voltage which you want to use... And the FQ was low under a 1000 cps

Too bad no one is left to give a best answer :-/


1 year ago

Hi there, I'm wondering what happened on this thread?

Did pletchman ever complete his project!?

I want to build one now too!

I picked up a book "Build The Micro-Stim" by Reginald W. Davis which actually gives the schematics to build one of these things. One thing -- he uses a 9volt battery to power the LED display and a 45-volt battery to power the unit. I'm curious why he chose or needed to use a relatively high voltage. Also, if I were to use a 9-volt as the power source how would that impact the generation of the microamps and the pulsating frequency??? Any thoughts?


Feeling guilty about not providing a circuit for you. Now I recall the old design knowledge.

A high voltage allows a designer to make a simple constant current by requiring a high series resistance compared to the small irrelevant skin resistance variation. This also works as there are no microamp regulators that I could find and looking at using an opamp current fractional gain structure was unstable.

No problem. I appreciate and respect your time and expertise. This is more of a pet project so I've got no deadlines or anything. So I think I understand the high voltage appeal but I assume this isn't necessary since the commercial variant of this device only uses a 9-volt battery.

Also, in Davis' book he gave an updated version which allowed variable frequency via a potentiometer. In the original design he used four IL-74 optical couplers to reverse current flow but now opts for a 2506 quad opto-coupler. Below are the images of his complete circuit layout

So I'm thinking about building a variant of his design and I'd like to use a li-ion battery or something small and portable. If you could help me through integrating his design with a lower voltage battery that would be awesome.

Screen Shot 2015-02-18 at 12.50.58 AM.pngScreen Shot 2015-02-18 at 12.50.18 AM.pngScreen Shot 2015-02-18 at 12.50.40 AM.png

Yet as I understand the 9v commercial still has high voltages presumably from some inductive kick-up inverter ?

Keep in mind I could not find a satisfactory semiconductor microamp regulator for skin resistance variations..

Hmmm -- the photos I provided for the commercial unit show all the inner workings. Do you see what they use to invert to higher voltages?

Found two inductors L1 & L2, one of which could be used to program an older flash memory the other could be your skin HV.


I see those. This unit does have flash memory (very poorly designed and still in use) so you're likely correct on one. So I assume you are saying they implemented an inductor to compensate for any changes due to skin resistance. So you're saying with a higher voltage like Davis' used in his version (45-volt) there isn't a need for this "buffer"

So what if I wanted to use a very efficient 3.7v Li-Ion battery for size and portability and the ability to keep it charged for when needed. What additional components would I need to use to make sure it worked?

What I was offering is any battery can use an inductive element in a kickback inverter circuit to make a higher 40 to 90 voltage that is then regulated stable enough to make your constant skin current feasible.

Say 50 VDC 100 uA needs a 1/2 megohm (500,000 ohms)

Skin resistance is known to vary from 1000 to 20,000 ohms while an individual may only range 2000 to 10,000 ohms or +-4000 ohms

A change of 4 of 500 would be 0.8% variability all because of a higher voltage.

Ok. I've been making progress on working this out. I'm looking at either an oscillator circuit using the venerable 555 timer chip or possibly the quad optocoupler suggested in the Reginald Davis book. However, I've also come across using a signal generator to achieve the desired frequency and this seems much more flexible and (equally important) programmable to work with. I'm a little unclear if there are any reasons I can/can't integrate a signal generator in lieu of the oscillator to generate the desired frequency as needed. Please impart your thoughts and wisdom.

Thanks in advance as always.

The 555 can easily deliver twice the frequency 5 to 2000 hz divided by a flip-flop deliver a 3 to 970 hz square wave to your opto current H Bridge.

Only the first 555 period will be a little longer as the timing capacitor first charges from near zero volts.

The only problem with the 555 is I've got to use a slew of capacitor/resistors to get each of the combinations. I don't want to use a rotary potentiometer. I've looked into the 256-bit digital ones but, again, if you look at the AD9850 or AD9851 signal generator can get all frequency ranges precisely and pretty cost effective and since its software controlled I can set it up to run through an Arduino type board. What do you think? One concern or question is if the 35-volt input will be an issue and also is there any functional difference in the frequency created by the 555 timer oscillator circuit verses the DDS signal generator?


Fine get your frequency using what you want. I only mentioned the 555 as a simple fq gen.

The 40 VDC is only applied to the optoisolators.

how do you apply the 45-volts to the frequency after the 555 timer or signal generator? The output on the ad9850 signal generator is only about half a volt and I know 555 timers don't allow higher current through them either. So what's the next step after I create a frequency? Thanks

So would the boost converter by applied to the current AFTER the frequency is created/established. Does it get changed going through the boost circuitry? Also, your skin tolerance information was very useful. Would you suggest adjusting the anticipated microamp output to take into account the typical resistance value for skin? In this case I'd be using conductive gel and leads so I assume the resistance would be minimal. But say the conductive gel method has a typical 3,000 ohm resistance (might be less not sure). Would I assume that additional resistance into all my microamperage calculations. What do you think?

Also, I am not married to a signal generator vs the 555 timer. I just thought from an ease of use standpoint it seems much easier than calculating every R & C amount for 1000 frequency settings. Please let me know if you feel I'm missing something or have a reason to doubt a signal generator in this application.

As always thank you for your wisdom and input

got it. I worked in a boost converter to get 35v to work with.

In Reginald Davis' version his complete parts lists was the following:

Screen Shot 2015-02-17 at 10.13.54 PM.png

I posted this a while ago but not seeing it here. Below are images of the existing circuitry on a commercial FSM unit. Not sure if it gives any insight but here it is...


By the way, this commercial unit uses only a 9-volt power source and my understanding is a single battery lasts literally weeks and weeks since the output is so minimal.


3 years ago

OK here is the first part, a sine wave phase-locked-loop that can operate on a 1000 to 1 frequency range.


All that remains is to add an adjustable current limited 25V DC source and a push-pull output, most likely complimentary mosfets because bipolar devices would inject base-emitter currents in excess of 20 ua.

Will see if I can get a design for you.

hello ICENG

Glad to have someone with your wealth of knowledge taking a peek at this "challenge"

1 OP-AMP no skill but I am familiar with them

2 RMS or peak I presume RMS but don't think it's overly critical

3 Yes, skin contact via gel-based conductor pads at two points for channel A and two points for channel B

TENS units operate on milliamperes and a more narrow frequency range. frequency specific microcurrent runs on microamps and a broader frequency range with two channels instead of one as well. Thx


3 years ago

HI ! ... Long time no hear and were a couple years older.

How skilled are you with op-amps ?

The microamps are RMS or peak ?

What is the load the current flows into ? ... Skin ? .... resistance ?