To carry out measurements at various spots can be handled in two principal ways:
- Parallel spectrometer channels
- Sharing one spectrometer by implementation of switch/blocking devices
An implementation of a switch introduces additional losses, time delays, for only one channel after the other can be acquired, and – depending on the type of switch – the introduction of uncertainties due to moving parts.
Parallel spectrometers don’t have these drawbacks, can measure really in the shortest possible time, but may be an expensive solution. An electronic multiplexer allows to share one electronics with up to 8 spectrometer modules.
Sharing one spectrometer:
Two main principles are in place:
- Blocking of individual channels
- Switching/moving of optical elements to redirect light path
A 3rd principal is based on a 2D CCD (see below)
Independent of the used technique, the overall spectral range depends on the used fibers and can be as wide as 200nm to 1µm or 350nm to 2µm.
Blocking is realized by splitting up either the illumination or the detection path in multiple channels. Consecutively, all but one channel is blocked by a shutter. The splitting introduces losses in direct relationship to number of channels: the more channels, the lower the efficiency. At 4 channels, the efficiency is at about 20%, i.e. this technique is limited to few channels. The big advantage: the optical path is not moved/touched, which keeps reproducibility at 100%. Switching times are in the range of 0.1s.
Switches redirect the light path by moving fibers or optical elements such as prisms. A functional unit with fibers works as 1:4, cascading allows to increase the number of channels, but adds about 30% loss per step. Up to 32 channels have been realized, at an overall efficiency of about 30%. Drawback of switching is the movement of elements which introduces some variation. Reproducibility of 99.8% short term can be achieved. Switching from channel to channel with fibers can be done in about 10 ms. Depending on the switching element this time might be significantly higher. The switching times depend also on the fiber diameter.
A specialty for sharing one spectrometer for multiple measurement channels is the use of a 2D CCD detector array for readout. In case the optical entrance is split into multiple channels along the entrance slit – and a decent imaging quality of the spectrometer from slit to the detector – the channels can be read out independently. The usefulness depends on how good the individual channels can be separated. This is directly related to the imaging quality of the used spectrograph: Also the readout mechanism of the 2D chip is influencing the performance; there might be electronic cross-talk.