# A wearable electric stimulation watch

Discussion in 'The Projects Forum' started by kevin0228ca, Jun 10, 2015.

1. ### kevin0228ca Thread Starter Member

Jun 5, 2015
30
0
greetings.
I am building a wearable device which will electric stimulate a person's arm.
Requirements are
current should be constant less than 10 mA
current should be able go both direction
supply voltage should be 50+ in order to be able to stimulate human skin
wearable so I guess should be battery powered
stimulation signal should be biphasic square wave so current does not accumulate
I am quite new to electric engineering, so please verify if my logic about building block of my system is correct.

I have a few questions.
For constant current source, i am using a circuit

Vcc should be 50+.
then in order to induce least voltage drop across R3,
V+ going to to opamp, which was originally 0-5V square wave, should be amplified to 0-40+V, then go to opamp?
can opamp amplify to 40+V? because I think most opamp have Vcc limit around 30V?

Also, is possible to convert 0-10mA square wave to +- 10mA square wave?

for charge pump I am thinking about using IC from linear technology or maxim.
is possible to boost from 9V to 50+V?

Can anyone verify if my logic is correct?
Thank you.

2. ### blah2222 Well-Known Member

May 3, 2010
565
33
Hi, I have built something similar for a school project.

First off, it is a fun project but not something you should just jump into if you are still fresh to circuit theory or electrical engineering. In this case, you are dealing with high voltage and direct contact with the stimulation circuitry which is not to play around with.

When I built this, I made sure it was battery-powered and any data transmission was wireless, to make sure there were no connections to mains power.

Is this going to be controlled by a physical switch or software? What kind of protection circuitry are you going to imply as to not shock the person intermittently as soon as they are hooked up?

Hooking a patient up and having the stimulation occur right away at the end-level is quite uncomfortable, normally the amplitude has a gradual ramp-up period, is that part of the plan?

The biphasic waveform will require a switchable H-bridge to change the polarity of the waveform.

Are you going to isolate the stimulation power source from the analog/digital circuitry? Planning to use opto-couplers?

The op-amp in your circuit above is rated for +/-18V, but based on the configuration you have using a PNP transistor and your +50V design, that will not meet spec. Why that configuration and why 50V?

Every person's conductance will be different and vary depending on activity level at that moment, body composition, and a slew of other things.

These are only a few of the things to think about, but I would definitely recommend getting comfortable with the basics before trying to build this and strap it to your body unless you are comfortable with what everything is doing.

3. ### kevin0228ca Thread Starter Member

Jun 5, 2015
30
0

I think I will use software to handle on - off of circuit.

I thought if I use a constant current source, by varying R3 I could vary Load current to be minimal so person would not get shocked as soon as they wear?
What kind of circuit can I build to prevent danger in case constant current source malfunction?

The 50V supply is because I read that maybe 50+V is needed in order to stimulate human because of human skin resistance.
Also commercial TENS unit is usually 50+V.
Is 50V necessary or less than 50V is sufficient?

so I need to convert 0-10mA to +-10mA, then use H-bridge to convert so current goes both direction?

4. ### wayneh Expert

Sep 9, 2010
12,382
3,241
Are you planning to use leads? I can't picture many people wanting to stimulate just their wrist.

My wife just bought herself a nice TENS unit with leads for pads, another set for ear clamps, and a pair of foot "sandals" with embedded electrodes. The device has a nice display and is programmable, with many options. It's the size of a small iPod and the rechargeable battery lasts 10 hours or something like that. All this for <\$40.

Add a wrist strap, and you're done.

Or, if you must build one, take a look at mine.

5. ### kevin0228ca Thread Starter Member

Jun 5, 2015
30
0
though for my level looks quite complicated.

I developed a schematic which is able to produce 50V across VLoad, with constant current of 10mA at max 5000 ohm.

since AD822 is limited by 30V, I use opamp1 produce 25V and opamp2 produce -25V to give 50V.

Is solution correct?
And now is +10mA, how do I get -10mA?
Also, can current flow in both direction across VLoad? because I thought current must flow away from collector of transistor.

Thank you.

6. ### Bordodynov Active Member

May 20, 2015
671
194
Here'sone possibleelectronic circuit:

7. ### DickCappels Moderator

Aug 21, 2008
2,772
669
blah2222 has some good points. Your output is getting close to 60 volts. That sounds high. How do you know that voltage that high cannot cause damage, even when current limited to 10 ma?

How much voltage and what current limiges do commercially available stimulators use?

I'm glad you are thinking about how to protect the user in the case of failure of the current source. A good solution might to be to use enough resistance between the voltage source and the electrodes so that an unsafe current cannot be produced. Alternatively, you could have a current monitor that shuts down the circuit when a specific threshold is crossed.

8. ### blah2222 Well-Known Member

May 3, 2010
565
33
Another thing to keep in mind is that the load is not purely resistive as you have assumed (see http://soundlab.cs.princeton.edu/learning/tutorials/sensors/img29.gif). Your design will not work as you expect with that assumption.

When you apply a square wave pulse of current, the voltage will ramp linearly due to the capacitive element of the load. A simple approximation would be a resistor (few hundred ohms) in series with a parallel resistor-capacitor combination.

Stimulation frequency and duty cycle are also very critical here. Your on-time shouldn't be more than a few hundred milliseconds as that can get quite uncomfortable for the user and increasing the frequency past 40 Hz starts to get uncomfortable as well.

Based on your current design you have no mention of pulse duration or frequency of stimulation at all, which besides current amplitude are pretty darn important. You also don't seem to be aware of what kind of load you are dealing with. I would recommend that you sit down and read up on electrical stimulation and bioelectricity if you are planning on moving forward with this.

Final thoughts, why are you wanting to stimulate the wrist? At most that will create an awkward movement/tingling of the fingers. I am interested in your goal here...

Last edited: Jun 11, 2015
9. ### wayneh Expert

Sep 9, 2010
12,382
3,241
Or, to reiterate, just go buy one.

10. ### blah2222 Well-Known Member

May 3, 2010
565
33
Depends on what they are wanting to get out of the project. If this was a research/design project then a DIY makes sense, I am getting the sense that this is not the case, and I would have to agree with you.

11. ### kevin0228ca Thread Starter Member

Jun 5, 2015
30
0
I must have missed your response.

We are doing a school project and that's why we are developing ourselves.
I am a software engineer student instead of electric engineer therefore I have questions about circuits.

I also find resistor capacitor representation of skin.
But I am confused.
What do you mean voltage will ramp linearly due to the capacitive element of the load.
You mean I should use a resistor capacitor load instead of a simple varying resistor to simulate as skin resistance to provide constant current?

Thanks for reminding about frequency and pulsewidth.
In fact, I know these variables and already have a program that change these.
I did not mention because I thought irrelevant to circuit.

Aug 21, 2008
2,772
669