MPS-TMS — Modular pulse synthesizer for TMS

A conventional TMS device is a capacitor discharging through a coil. That fixes the pulse to a damped sinusoid, fixes the energy per pulse at $\tfrac{1}{2}CV^2$, and puts nearly all of it into the coil as heat. Pulse shape was never a parameter anyone chose; it was what the circuit did.

MPS-TMS replaces the single discharge with many modules switching in sequence to approximate a commanded waveform. The comparison I find most useful: it is a digital-to-analog converter, operating at a few megawatts.

What that buys

Pulse shape, phase durations, inter-pulse interval, and burst structure all become numbers you set rather than consequences of component values. Concretely, that means a neuroscientist can ask which waveform recruits which population — a question that could not previously be asked with one device, because changing the waveform meant changing the hardware.

The two problems worth writing down

Monophasic pulses at repetition rates. Monophasic pulses are what most experiments want and the worst case for coil heating, which is why repetitive protocols were largely stuck with biphasic. Giving a sinusoidal current a long, low pre-phase opposite to a short main phase yields an equivalent electric field in the brain while shifting the current waveform down by nearly half. Loss goes with the square of RMS current, so this removes roughly three quarters of the coil loss — which is what makes rapid-rate monophasic stimulation possible at all.

Quantizing the command. A modular converter must repeatedly choose which output level to switch to next. Nearest-level modulation switches rarely but coarsely; carrier-based PWM is the reverse. Passing the nearest-level residue through a low-pass filter before a one-bit PWM stage keeps most of the fidelity at most of the low switching rate. The filter is the load-bearing part — one-bit PWM only helps when its input bandwidth is constrained, and without it the scheme degenerates into level-shifted-carrier PWM.

Where it falls short

Module count. Fidelity scales with how many voltage levels you can produce, and levels cost modules — each with its own capacitor, bridge, gate drive, and isolation. A system good enough for demanding waveforms becomes large, expensive, and hard to move out of a lab, which is the honest reason this is a research instrument and not clinical equipment.

That constraint is what AMPS attacks.