Harvey P R, Mansfield P
Department of Physics, University of Nottingham, United Kingdom.
Magn Reson Imaging. 1994;12(1):93-100. doi: 10.1016/0730-725x(94)92356-6.
Two passive filter circuit designs are compared which enable series resonant trapezoidal drive of the modulated read gradients used in both echo-planar and echo-volumar imaging. Using simple circuit simulation software, details of the expected power requirements and performance have been evaluated for peak operating currents of up to 1000 A. The circuits evaluated were of a 7th order design incorporating a 110 microH gradient coil operating at fundamental frequencies between 800 Hz and 2000 Hz. Results indicate that the simpler circuit design is not necessarily the most efficient. In addition, by suitable modification of the input waveform algorithm, it can be shown that the shape of the resonant gradient waveform may be varied entirely under software control from triangular to trapezoidal, the minimum ramptime being specified by the highest harmonic accommodated in the circuit. The energy storage, non-dissipative nature of the circuit design also means that the gradient ramptime is effectively independent of amplifier power capabilities, making a fast rise trapezoidal waveform as easy to implement as either triangular or sinusoidal waveforms. Using the approach outlined above, trapezoidal gradient modulation at peak current amplitudes in excess of 320 A and risetimes less than 60 microseconds have been easily produced in this laboratory for use in echo-planar imaging and for investigations into ways of optimising gradient performance before the onset of peripheral nerve stimulation.
比较了两种无源滤波器电路设计,它们可实现用于回波平面成像和回波容积成像的调制读取梯度的串联谐振梯形驱动。使用简单的电路仿真软件,针对高达1000 A的峰值工作电流,评估了预期功率要求和性能的细节。所评估的电路为七阶设计,包含一个110微亨的梯度线圈,工作在800 Hz至2000 Hz的基频之间。结果表明,较简单的电路设计不一定是最有效的。此外,通过对输入波形算法进行适当修改,可以证明谐振梯度波形的形状可以完全在软件控制下从三角形变为梯形,最小斜坡时间由电路中容纳的最高谐波指定。电路设计的能量存储、非耗散特性还意味着梯度斜坡时间实际上与放大器功率能力无关,使得快速上升的梯形波形与三角形或正弦波形一样易于实现。使用上述方法,在本实验室中已轻松产生了峰值电流幅度超过320 A且上升时间小于60微秒的梯形梯度调制,用于回波平面成像以及在周围神经刺激开始之前研究优化梯度性能的方法。