Length Sensing & Control Designs




This page has some description of a preliminary LSC design. The first step is just making a DARM loop design
followed by MICH, PRC, & SRC loop designs.

DARM Loop

This is a 2nd version of a DARM loop for Advanced LIGO. I've added the optical
spring by putting a frequency dependent spring into the quad suspension model
to simulate the real dynamics. I then use Optickle to generate the low power
plant response assuming the nominal SRC detuning (optimized for DNS).
The overall plant transfer function comes out right and this allows modeling
the effect of the optical spring on the quad's dynamics.

This is stop-gap method which I think is mostly correct. Eventually we'll include
the pendulum model into Optickle.

The optical spring and the optical resonance are both compensated digitally. I've
left a little bit of mismatch in the compensation because that's the way it probably
will be and also just to remind us that these resonances are there.

The SUS violin modes are not included in the mechanical model, but they are
notched out in any case to simulate the real loss of phase margin. Also
included in the model are anti-aliasing on the ADCs & DACs.
To get some more phase margin I've modified the AA filter to be less aggressive
than the LIGO-I electronics. The (post-DAC) AI filter is still the same.

L- Noise

The plot below shows 3 different noise curves for AdvLIGO. The black trace and the red trace
are the NS/NS tuned and BH/BH tuned curves from this page.

The green trace is a first guess at the noise on the 'first day'. This is quite a bit
noisier than the initial LIGO displacement noise but probably not crazy for how things
might look before we get the entire system commissioned.

I propose we use these 3 curves as our 3 baseline sensitivity modes and should expect to
sit in the green 'Noisy' mode for awhile.
After some further optical modeling the hope is to post here the SNR for each of the 5 DOFs and then also
the THD for each so as to predict the upconversion due to fringe non-linearity.