Sub Menu contents

The CFHT delay line integration


An unexpected internship

Everybody knows that going to the US is quite administratively hard and complicated. Even more when you want to work there and earn money. You need to know that the intern status does not exist : to go there you need a J1 visa which allows you to be paid.

To get this J1 visa, you need of course to fill several papers and you need an official paper from where you are to work which ensure that you are indeed going to be paid. This paper is the famous DS-2019. At the beginning, I was supposed to leave at the beginning of March. But it took a very long time to send this fuc*** paper which postponed my departure to the beginning of April.

As I was keeping Guy P. posted about my departure to Hawai'i, as soon as I told him that I postponed my departure of one month that he immediatly asked me if I wanted to work with him and his team on the integration of the delayline of the CFHT. I accepted at once knowing that it would be worth working with Guy and his guys !

Just so you know, the CFHT delay line is part of a big project which is 'OHANA. This project consists in using the biggest telescopes atop the mauna Kea in Hawai'i together. Basicaly, the delay line will be used to cancel the optical path between two beams coming from Gemini North and the CFHT.

Location of INSU Gemini North CFHT Mauna Kea summit

The CFHT delay line

The delay line will be used to create an interferometer between Gemini North and the CFHT. The principle of an interferometer is to recombine beams which come from different telescopes or antennas. This recombination is effective when the optical path of the beams are all the same, which means that all the beams have to cover the same disctance.

The delay line is used to cancel the optical path difference between two beams. The beams are injected in the delay line at its two extremities via an otpic-fibre cable and through off-axis parabola. The beams travel in the delay line and they leave it being injected in optic-fibre cable. Once they left the delay line, the two beams traveled the same distance and the can be recombined.

But before using the delay line, you have to be sure that each beam exits the delay line and that is roperly injected in its optic-fibre cable, otherwise you will lose intensity, power and information.

This is were the integration of the delay line has its importance.

My Work

We did the integration step by step. The first one was to set the off-axis parabola in order to have a collimated beam : the beam needs to have the same diameter wherever you look at it.

For that, you need to put the end of the optic-fibre cable in the off-axis parabola focus. To do that, you project the beam on a wall which you supposed to be at an "infinite" distance (around 20 meters).Then you move the cable to reach the parabola focus.

This setting is very hard because every mouvement of the cable needs to be understand to know wether this mouvement allows you to get closer to the focus or not. And you have to move the cable in the three directions of space which multiplies the wrong mouvement. It is also very hard for the eyes because we work with a laser projected on a white wall which rapidly is exhausting. You need to be well focused to estimate the tiny changes du to the cable shifts. Finaly, we faced the problem that when we left the parabola well set, the day after, the settings were not good anymore.

And we had to set four parabola : the two parabola that inject the beams in the delay line, and the two parabola that inject the beams in the exit optic-fibre cable after the beams travelled in the line.

Once the parabola were set, we had to move on the next element to be set which were the mirrors. The beam which exits the fibre, reflects itself on the parabola which sends it on a mirror before going in the delay line.

We had to set each mirror so that the beam reaches the middle of the mirror in front of the former one. This setting was a bit easier than the one for the parabola, but we still had to face the fact that the settings changed during the night.

The next step which was for me the last one, was to put the continuous carriage on the rails. The carriage is made of four dihedron and he can move on the rail to cancel the optical path difference.

The setting was to check that one beam can exit a fibre, travels through the delay line and then can be injected in a fibre. When I left, the settings were still on going.

Finally I can say that this month working at the INSU was a really nice experience, and I liked working with the integration team.