The mounting and cabling proceeded in the following order, which was pretty much defined on the fly since there was no written plan of how to do it:

1. Connect AOH at position TOP/ICC1/OP1, run fibres through opening in rod frame to outside
2. Connect AOH at position TOP/ICC1/OP2, run fibres through opening in rod frame to outside (photo 1)
3. Connect AOH at position TOP/ICC2/OP1, run fibres through opening in rod frame to outside
4. Connect AOH at position TOP/ICC2/OP4, run fibres through opening in rod frame to outside (photo 2)
5. Connect AOH at position TOP/ICC2/OP3, run fibres through opening in rod frame to outside - fibres not yet in final position (photo 3)
6. Connect AOH at position TOP/ICC2/OP2, run fibres through opening in rod frame to outside - fibres routed in final positions, those from OP2/3 running under those from OP1/4 (photo 4, photo 5)
7. Fibres routed along ROD frame, held in position with foam retainers, and back into centre of ROD at CCUM
   (photo 6, photo 7, photo 8, photo 9)
8. ROD turned over, steps 1-7 repeated on bottom side. Potential problem with fibre routing on ICC2 seen again (photo 10)
9. Bottom-side fibres must exit to TOP-side by passing over the cooling pipe - care must be taken to avoid the sharp corners of the cooling blocks (photo 11)
10. Fibres are coiled above CCUM to take up excess length. Note that the prototype AOHs used here all have the same 2m long pigtail, which is too long in all cases. Connections are made by ROD side, starting with fibre 1 being connected to the AOH furthest from the CCUM. Connections are made with the adaptor blocks free to move and not yet mechanically attached to the ROD (photo 12)
11. Once all connections are complete, the adaptor blocks are fixed to the ROD mechanically (it would have been better if the tapped part of the aluminium retainer had been facing the CCU and not the end of the ROD). All fibres are fixed into position to stop them moving about (photo 13)
12. The fibre & cable clamp is put in position. We tried the bare carbon fibre piece but found that some foam was required to take up the large difference in diameter between the CCU ring cable (3-4mm) and the fibres (0.9mm). It also seemed better to us to route the CCU ring cable in the centre of the ROD, but this may not be compatible with the destination of this cable in the final system (cable may end up being too short). (photo 14, photo 15)
13. ROD fibre cabling finished. It would have take us approximately 3hrs to complete had all the pieces been to hand immediately.