Milovana.com

I have decided to analyze 2B Estim power Box output using an Oscilloscope, so here are the results for the Geeks here.

In Future I'll get a good USB stick that works with the oscilloscope so i can take screenshots instead of pictures.
Personally I'm very surprised by the results... Does this looks legit?

Blue = PC Output; Amplitude of 2 [V]; File = "100Hz square vibe 50pct"
Yellow = 2B output;

2B output - Over 1 [kOhm] resistor; "Adjust" = 2

2B output on 30% (the output wont rise any higher); over 1 [kOhm] resistor;
Amplitude is ~ 100[V] pulses every 7 [ms]

2B output, over 1 [kOhm] resistor; Single pulse zoomed in.
Amplitude is approximately 100[V], Pulse Time ~ 120 [us]

100 [Ohm] resistor, 2B at 8%
Amplitude ~ 8 [V], Width 120 [us]

100 [Ohm] resistor, 2B at 30% (Maxed out)
Amplitude ~ 17 [V], Width 120 [us]


On 2B changing the "Adjust" Knob will move the pulses closer the higher the adjust go;

No change in Pulse Shape was observed.
Slight variation of signal Dip was observed when input Amplitude changed.
I did not observe comprehensible connection between input file and 2B output, except in slight pulse height variation.
I've tried several different files to the same result.
On "Continuous" output of 2B, the same signal were observed.

What do you think?

It's certainly neat & interesting, but what are we comparing it to? I'd be curious to see the waveform differences between the 2B, a standard TENS device and one of the DIY devices to get a sense of the thing. Otherwise, it's just interesting, but not meaningful

I have a clone of Erostek 312, I could try to look at it's output. If my SO's old oscilloscope still works... If you're gonna live with someone, pick a person who has a working oscilloscope, more electronics tools than you and 24/7 access to the local hacker workspace

It is known that commercial devices (quite for sure all battery powered ones) work like this. In order to trigger a neuron just a pulse is required. (Neurons can only fired at specific rates, Neurons don't "feel" the shape of a pulse, e.g. rectangular, sinusoidal or Mickey Mouse shaped, and neurons can't "hear" whether an audio file sounds nice ... )

These coil generated pulses are an efficient and pragmatic way to produce a current controlled estim signal. (Devices based on audio amplifiers use serial resistor for that purpose. The energy required for the stimulation is less than 1W, the remnant is transformed into heat.)

Theoretically it is even possible to produce proper multi-phase signals by exact timing of the pulses ...

Nevertheless, there are two surprises

1. Pulse width is quite small and should only trigger most sensitive neurons. IMHO 0.3..0.5 ms would be better.
2. The device only fires pulses of single polarity. Different polarities would trigger twice as much neurons because the polarity on which they react depends on their orientation, i.e. is random.

It wouldn't surprise me if the 2B was averaging the power over a longer window than the amplitude changes in that audio file.

Would you be willing to take a measurement which has both silence and sound, with maybe a 200 ms long capture?

Ideally something like silence, 1V 200Hz sine, 2V 200Hz sine. Then we could see how long the 2B takes to ramp.

It would be really interesting to see how the 2B behaves with Adjust at 2 and at 99 with that setup, too.