A low-power integrated circuit for a wireless 100-electrode neural recording system receives @ 6.5 kb/s over 2.64 MHz inductive link transmits @ 330 kb/s over 433-MHz FSK transmitter 13.5 mW of power 4.7 by 5.9 mm^2 chip -- experimented with cat cortex "neural amplifier" Note that microelectrodes that measure the full 100 mV spike amplitude of a neuron, kill that neuron within a few minutes. Thus microelectrodes are usually a few microns away. "A typical neuron generates 10-100 spikes/second when active." [7] P. R. Kennedy, A. Hopper, C. Linker, and S. M. Sharpe, "A system for real time processing of neural signals for use as prosthetic controllers," in Proc. 1992 Int. Conf. IEEE Engineering in Medicine and Biology Soc. (EMBC 1992), pp. 1343-1344. [8] P. Irazoqui-Pastor, I. Mody, J. W. Judy, "In-vivo EEG recording using a wireless implantable neural transceiver," in Proc. 2003 Int. Conf. IEEE Engineering in Medicine and Biology Soc. (EMBC 2003), Capri Island, Italy, pp. 622-625. "Also see [9] for a thorough review of previous wireless biopotential recording systems" [9] P. Mohseni, K. Najafi, S. J. Eliades, and X. Wang, "Wireless multichannel biopotential recording using an integrated FM telemetry circuit," IEEE Trans. Neural Syst. Rehabil. Eng., vol. 13, pp. 263-271, Sep. 2005. 10 mW operation is desired because anything higher usually kills the surrounding tissue due to excess heat production The chip-package in this paper fits directly on the back of the 100-microelectrode Utah array. Coated in parylene and silicon carbide to protect it from internal body fluids. "Each neural signal generates data at a rate of 150 kb/s. A 100-electrode recording device would therefore need to transmit data at a rate of 15 Mb/s. This presents a significant technical hurdle given that extremely low power dissipation (~10 mW for the entire system) is necessary to avoid heating the surrounding tissue, small size requirements (<1 cm) prevents the use of an efficient transmitting antenna, and increased tissue absorption at high frequencies greatly favors telemetry operation below 1 GHz. (It should be noted that optical transcutaneous data transmission using an infrared emitter was demonstrated at data rates of 40 Mb/s, although the power dissipation of 120 mW would prevent it from being used in close proximity to brain tissue [11].)" [11] K. S. Guillory, A. K. Misener, and A. Pungor, "Hybrid RF/IR transcutaneous telemtry for power and high-bandwidth data," in Proc. 2004 Int. Conf. IEEE Engineering in Medicine and Biology Soc. (EMBC 2004), San Francisco, CA, pp. 4338-4340. Automatic spike-detection circuits [12] [13] Has a complete block diagram. Useful. Schmitt trigger class E amplifier Paper also discusses a finite state machine that listens for particular rises and drops and starts timers to determine whether or not input is a one or a zero from the wireless telemetry. Spike sorting and transmission of general properties of spike shape to distinguish multiple neurons at a single microelectrode interface/tip (platinum in the case of the Utah array)