Preamplifier Decoupling: Theory & Practice
Randy Duensing1 and Christian Findeklee2
1Philips GmbH Innovative Technologies, Germany, 2Philips GmbH Innovative Technologies, Hamburg, Germany

Synopsis

Since the introduction of RF coil arrays (Roemer, et al), the dominant approach to RF coil matching has been to use a degree of freedom to produce a large impedance mismatch between coil and preamplifier while maintaining an approximate noise optimal impedance for each channel individually. In reality, an array of RF coils has shared impedances which means that this approach is suboptimal at all locations in the FOV, and, in optimal reconstruction algorithms the impedance mismatch is independent of SNR. Knowledge of the coupling matrix allows for optimization of local SNR via a system tune/match approach.

250 Character Overview

RF coils are only fun and profitable if deep knowledge about the arcane secrets of preamplifiers and impedance matching is revealed. Throw off the out-dated popular traditions of preamplifier decoupling and individual coil tuning and enter the modern era of multi-coil matching science.

100 Word Summary

Since the introduction of RF coil arrays (Roemer, et al), the dominant approach to RF coil matching has been to use a degree of freedom to produce a large impedance mismatch between coil and preamplifier while maintaining an approximate noise optimal impedance for each channel individually. In reality, an array of RF coils has shared impedances which means that this approach is suboptimal at all locations in the FOV, and, in optimal reconstruction algorithms the impedance mismatch is independent of SNR. Knowledge of the coupling matrix allows for optimization of local SNR via a system tune/match approach.

Key Take-aways

1. Preamplifier decoupling is convenient but rarely has impact on SNR
2. Power matching is not optimal noise matching
3. The input-referred preamplifier noise models incorporate a current source
4. Noise couples from one preamplifier to another if there is a shared impedance (regardless of preamplifier decoupling) because of the noise current source
5. Excess noise (over the optimal noise figure of a preamplifier) is caused by any impedance mismatch from the optimal noise impedance
6. Share impedance implies that the noise optimal impedance at a port is affected by the relative signal distribution to the channels in a multi-element reconstruction
7. SNR can be optimized for minimum excess noise at one point in space (using different matching for each port) or for one eigen-mode of the coupled coil system (using identical matching for each port)
8. Preamplifier decoupling is preferred when not all RF elements are used, or other objects/devices couple to the coils or, noise whitening is not part of the reconstruction algorithm. It is also convenient in the traditional matching approach
9. The best mitigations to SNR loss are to have the lowest noise figure preamplifiers possible, know the actual optimal noise impedance and minimize coupling between coils from the start. High coupling means high relative loss of SNR regardless of preamplifier decoupling “masking.”

Acknowledgements

Thanks to Arne Reykowski

References

No reference found.
Proc. Intl. Soc. Mag. Reson. Med. 29 (2021)