Project D: Stimulation, studied properly
Design a charge-balanced current-source stimulator, drive electrodes in a saline bath, measure the voltage waveform and the electrochemistry, and understand the Shannon limit. Everything about stimulating tissue except the tissue.
You are skimming: the title, the first figure, and the short version. Switch to Read in the header for the full page, or Deep to open every deep dive.
The safety charter forbids stimulating a nervous system outside a lab. It does not forbid understanding stimulation completely, and the way to do that is to build a stimulator and study what it does to an electrode in a saline bath, where every part of the physics and chemistry is measurable and nobody is at risk. By the end you will know why stimulators are current sources, why every pulse has an opposite twin, why the electrode voltage has the shape it has, and where the limits come from that every implanted stimulator obeys.
Why do neural stimulators deliver current pulses rather than voltage pulses?
Charge, and control of it. Neurons are excited by the charge injected near them, which is current times time. A voltage pulse delivers a charge that depends on the electrode’s impedance, which varies between electrodes, changes as tissue encapsulates them, and changes within a pulse. A current source delivers a known charge every time. It also makes charge balance exact: equal and opposite currents for equal times leave zero net charge, which is the condition for not running irreversible chemistry at the electrode.
The stimulator
A current source. The Howland current pump, or an op-amp driving a MOSFET with a sense resistor in feedback, produces an output current proportional to a control voltage, independent of load. Range 0 to 1 mA, compliance voltage of ±10 V or so (the maximum voltage the source can develop across the electrode before it runs out of supply). Two of them, or one bipolar one, to source and sink.
A waveform generator. The microcontroller produces the control voltage through a DAC: a cathodic (negative) pulse of set amplitude and width, an interphase gap, an anodic pulse of the same charge, and a rest. Typical values from the clinical literature: 100 µs pulses, 100 to 500 µA, at 50 to 200 pulses per second. Make every parameter settable.
Charge balance. Equal and opposite pulses is passive balance in intent; component tolerances make it imperfect. Add a shorting phase after each pulse pair that connects the electrode to the return through a resistor, discharging any residual, and a DC-blocking capacitor in series as a hard guarantee that no net current can flow.
Monitoring. Measure the current with the sense resistor and the electrode voltage with a differential probe, both to the scope and, decimated, to the plotter.
The bath
A beaker of 0.9 percent saline. Your microelectrodes from the electrode project, or a stainless steel wire for a start, and a large stainless return electrode. An Ag/AgCl reference for measuring the working electrode’s potential separately from the return.
What to measure
The voltage waveform. A current pulse into an electrode produces a voltage with an instant step (the solution resistance times the current) followed by a ramp (the double-layer capacitor charging). At the end of the pulse the step reverses and the capacitor discharges. Measure the step to get the access resistance; measure the ramp’s slope to get the capacitance. Compare to your EIS from the electrode project. They should agree.
The safe window. Increase the pulse amplitude. At some point the voltage ramp reaches a level where it flattens: the electrode has started to electrolyze water (hydrogen at the cathode, oxygen at the anode) and current is going into chemistry rather than charging the capacitor. Bubbles appear at higher currents. The charge per pulse at which this begins, divided by the electrode’s area, is the electrode’s safe charge injection limit. Compare to the literature values for your material: platinum around 50 to 150 µC/cm², iridium oxide and PEDOT several times higher.
Charge imbalance. Deliberately mismatch the anodic and cathodic pulses by 10 percent and run for an hour. Measure the electrode’s DC potential drift against the reference. Then enable the shorting phase and repeat. Then the DC-blocking capacitor. Each mitigation’s effect is a figure.
Long-term. Run a million balanced pulses at a safe charge density on a coated electrode and re-measure its EIS. Then at an unsafe density. Photograph the electrodes.
- Build the current source. Verify with a resistive load that current is independent of load and matches the control voltage.
- Add the waveform generator, shorting phase, DC-blocking capacitor, hardware current limit, and monitoring.
- Drive a stainless wire in saline. Capture the voltage waveform. Extract access resistance and capacitance; compare to EIS.
- Find the water window on stainless, then on platinum or coated electrodes. Compute charge density limits.
- Charge-imbalance experiments with each mitigation.
- Million-pulse durability on two electrodes. Report.