2results about How to "Facilitates temperature compensation" patented technology

A ring oscillator (50) comprises a chain of inverters (52) coupled between an oscillator supply node (54) and a reference node (56), and a current generator (58) coupled between the oscillator supply node and a system supply node (60) and configured to inject a current (IOSC) into the oscillator supply node. Each inverter comprises a first (NF) and a second (NL) low-side transistors coupled in series between the reference node and an output node of the inverter, and a first high-side transistor (PF1) coupled between the oscillator supply node and the output node of the inverter. The first low-side transistor and the first high-side transistor have respective control terminals coupled to an input node of the inverter to receive a respective inverter control signal (CK). The second low-side transistor has a control terminal coupled to the oscillator supply node. The ring oscillator circuit further comprises a biasing circuit (500) including a first bias transistor (NFZ) and a second bias transistor (NLZ) coupled in series between the reference node and the oscillator supply node. The first bias transistor has a control terminal configured to receive an oscillator control signal (StartP) indicative of whether the ring oscillator is in an active or inactive operation state. The second bias transistor has a control terminal coupled to the oscillator supply node. The first bias transistor is configured to selectively couple the reference node and the oscillator supply node in response to the ring oscillator being in an inactive operation state.

A temperature compensation method for MEMS accelerometers based on fuzzy neural networks is presented, belonging to the field of MEMS accelerometer temperature compensation technology. This method addresses the problems of existing MEMS accelerometer temperature compensation methods, such as missing optimization of algorithm antecedent parameters, insufficient compensation accuracy, and poor real-time performance. The method includes: constructing an FCM-T-S fuzzy neural network module; optimizing the network antecedent parameters using the FCM algorithm; updating the network consequent parameters using gradient descent; and then deploying the trained FCM-T-S fuzzy neural network module into the MEMS accelerometer temperature compensation system. A data acquisition module synchronously acquires the temperature output voltage and acceleration output voltage of the MEMS accelerometer. The FCM-T-S fuzzy neural network module outputs the compensated true acceleration based on the temperature and acceleration output voltages. Results: High-precision temperature compensation for MEMS accelerometers is achieved, improving their output accuracy and temperature stability.