The invention discloses a multi-
machine collaboration framework based fast recorder and a self-calibration and multi-
machine collaboration method. The fast recorder comprises four
signal conditioning modules, two slave computer
microprocessor control unit (MCU) modules, a dual-port
random access memory (RAM) module, a host computer
microprocessor control unit (MCU) module, a
synchronous dynamic random access memory (SDRAM) module, a reference
voltage source module, a universal serial
bus (
USB) interface and a display module. The two host computer MCU modules, the dual-port RAM module and the host computer MCU module constitutes a multi-
machine collaboration framework of the recorder, two analog-digital
converters are integrated in each slave computer MCU
chip, and AD synchronous sampling
clock signals are supplied to the four ADCs of two slave computers by an idle counter /
timer of a host computer; by means of the multi-machine collaboration framework, the sequential serial operations of sampling, caching, transfer, storage, correction, display, upload or unloading and the like is converted into synchronous efficient parallel operations; sampling data is corrected on line according to a self-calibration polynomial generated off line, and the precision of the recorder is improved; and an input terminal of the recorder adopts analog and digital dual-filtering technology and then is combined with photoelectric isolation
USB communication of an output terminal, so that the fast recorder has an excellent anti-jamming performance.