Characterization of a CMOS sensing core for ultra-miniature wireless implantable temperature sensors with application to cryomedicine

Abstract

In effort to improve thermal control in minimally invasive cryosurgery, the concept of a miniature, wireless, implantable sensing unit has been developed recently. The sensing unit integrates a wireless power delivery mechanism, wireless communication means, and a sensing core-the subject matter of the current study. The current study presents a CMOS ultra-miniature PTAT temperature sensing core and focuses on design principles, fabrication of a proof-of-concept, and characterization in a cryogenic environment. For this purpose, a 100 μm × 400 μm sensing core prototype has been fabricated using a 130 nm CMOS process. The senor has shown to operate between -180°C and room temperature, to consume power of less than 1 μW, and to have an uncertainty range of 1.4°C and non-linearity of 1.1%. Results of this study suggest that the sensing core is ready to be integrated in the sensing unit, where system integration is the subject matter of a parallel effort.

Keywords: All-CMOS; Cryomedicine; Cryosurgery; Implant; Miniature; Temperature sensor; Ultra low-power.

Figure 1

Schematic illustration of the proposed proportional-to-absolute-temperature (PTAT) sensing core

Figure 2

Schematic illustration of the circuit model of the BJT’s in the PTAT including non-ideal parasitic resistances R_B1, R_B2, R_E1 and R_E2.

Figure 3

Photograph of the fabricated sensor chip (a) and the packaged chip mounted on PCB (b)

Figure 4

Experimental setup: (a) a schematic illustration of the system, and (b) a photograph of the experimental stage

Figure 5

Results obtained from three representative chips in a self-biased mode (#1 and #2) and an externally biased mode (#2): (a) temperature data as a function of voltage output, and (b) average nonlinear offset, calculated as the temperature difference between the experimental data shown in Fig. (a) and a linear curve connecting the boundary temperature values for each dataset.