Quick check in order to understand the features at 70 kHz of the Slave error signal:
the guess is that it belongs to the fast piezo and changing the relative gain of both slow and fast path would change the shape, in order to give some clues to set the further tests

we change the gain of the PZT_SLOW path on the rampauto_ELEC_9801_01 of the slave injection loop.
At ~ 9/01/19_18H27MN LT
first trend is with G = 1/4 (reduced gain)
second trend is made with G = 1 (normal gain)
G is set by a blue potentiometer on the slow piezo channel

unfortunatelly we can't increase the gain since already found at the max position (potentiometer fully turned to the right)

Conclusion
- there is a light reduction of the 10 kHz bump with lower gain
- no change on the 70 kHz structures
- worth to increase the gain of the slow pzt to see whether slow piezo can eat the 70 kHz which could come from the fast piezo
- need to make a TF of the slow and fast piezo path
- we let the gain as before, set to 1 (fully turned right)