Michael, Marc and Yuhang

DDS signals usually give an output of -6dBm, which is not enough for many mixers. Due to the lack of enough LO power, we had issues, such as CCFC error demodulation. To solve this problem, Matteo ordered several amplifiers. The idea is to put them inside the DDS board and connect the DDS output directly to them.

Yesterday, Aso-san kindly provided us an instruction before the implementation of these amplifiers. Today, we followed the design of Matteo and implemented part of those amplifiers (for DDS2 and DDS3).

Figure 1 shows the connection done for an amplifier (We did five in total for today).

Figure 2 shows the DDS2 board before putting amplifier (we found unfiltered CH1 output is giving signal).

Figure 3 shows the DDS2 board after putting the amplifier.

Then I took it to TAMA and did several tests. In the beginning, I found the signal was not present in CH1. Then I changed CH1 from unfiltered CH1 to filtered CH1(shown in attached figure 4). After this, I discovered that signal (shown in figure 5) increase from -8dBm to 9dBm after amplifier implementation. This signal is used as LO to demodulate the filter cavity length error signal for GR. Figures 6 and 7 show the check of PDH amplitude for these two cases. The PDH becomes a bit smaller with a larger LO. I compared TF and GR locking length noise with these two cases.

Figure 8 shows TFs. After implementing the amplifier, the unity gain frequency is smaller while the phase margin is better. The amplified case also shows a better phase for higher (compared with UGF) frequency region.

Figure 9 shows error signals. After implementing the amplifier, the integrated length noise becomes less. This error signal is not calibrated. Besides, it maybe better to compare them again when they have almost the same unity gain frequency.

All amplifiers are also installed inside DDS3. We will test it tomorrow.