According to entry of telescope change of p-pol EOM, the smallest beam diameter is 0.6mm. And the datasheet tells us the power density threshold is 1000W/cm2.  Then I calculated the maximum power I can send is 1.4W. Then I found the AUX2 laser clamp current is 1.554A, which corresponds to 516mW of infrared beam. This is well lower than the damage laser power. However, the change of the AUX2 laser power also affect the laser power sent to fiber detector(DET01CFC/M). So we should measure the power transmitted by fiber. It should be less than 5.5mW(if the wavelength is 1550nm). The damage power of it is 70mW.

The Green measurement before GRMC is shown in attached figure 1. It is 424mV. The transmission while scanning is shown in attached figure 2. The dark noise is 133mV and TEM00 is 302mV. Actually, I checked the manual of thorlab biased PD, the dark noise comes from the dark current. So in principle, this dark current should give only an offset. If this is correct, we can just remove from all the detection value this offset. That means the green we detected before GRMC should be 293(424-133)mV. And the TEM00 transmitted by GRMC should be 169(302-133)mV. So the ratio is 169/293 = 58%. This result agrees with the measurement we did before of GRMC.

Besides, if the dark current just give an offset. The higher order mode we saw on the camera maybe really low. And this means the mode matching now should be fine enough. Anyway, we checked in the case we set 70dB gain of PDA10CS and highest resolution of oscilloscope. And then look at only higher order modes part of GRMC scanning spectrum. This is shown in attached figure 3. We can see the higher order mode becomes a little bit clearer. We can use this to improve matching better. Also a video is attached here. https://drive.google.com/open?id=1sDFy5q8tSbb1VEuDanusD3TIvpaR2KVj

Finally, I found the unknown peak appeared in last entry is because of yaw misalignment.