Difference between revisions of "CDC 19 12 2007"

Revision as of 21:06, 12 December 2007

Interposer, preAmp, and signal cable

This is the continuation of CDC_12_12_2007.

Thanks to Fernando we have a new reference measurement; he measured the noise of the interposer+preAmp+signal cable: his measurement is shown in figure 1.

figure 1: interposer+preAMp+signal-cable (taken from Fernando's report) (x-y divisions: 20 ns - 20 mV).

The differential signal was measured with the cable between the CDC-straw-wire and the HV-distribution board disconnected on the CDC side. The measured signal is shown in figure 2.

figure 2: HV-board+preAmp+signalcable signal (pink) with cable between CDC and HV distribution board disconnected on the CDC side.

No oscillations are observed! If one compares to figure 1 the comparison is reasonable, but I think Fernando has a better oscilloscope!

For the sake of completeness, the cable between the CDC and the HV distribution board was also disconnected on the side of the HV-board, this signal can be seen in figure 3.

figure 3: HV-board+preAmp+signalcable signal (pink) with cable between CDC and HV distribution board disconnected on the HV distribution board side.

Again, no oscillations.

Conclusion: the oscillations are caused by the input signals on the preAmp. What follows are several measurements performed to acquire a better understanding of the situation.

Now all previous measurements were taken with HV power supply switched, off let's repeat the measurement showed in figure 3 but now with HV switched ON (the supply was always connected to the HV distribution board). The result is shown in figure 4.

figure 4: HV-board+preAmp+signalcable signal (pink) with cable between CDC and HV distribution board disconnected on the HV distribution board side - HV power supply switched on.

The shape of the noise looks different but the amplitude does not change.

How do the noise/signals look like directly from the CDC-wire? Let's start with the noise, we have to points where we can measure: on the cable going from the CDC-straw-wire to the HV distribution board and directly behind the capacitor that decouples the signal from the HV. In figure 5 the noise is shown measured directly on the cable, notice that the HV has to switched off in order to perform this measurement.

figure 5: CDC-noise on the cable between CDC-wire and HV distribution board.

This looks OK to me. Let's see how it looks after the capacitor on the HV distribution board: this is shown in figure 6.

figure 6: CDC-noise after the capacitor on the HV distribution board - no change if HV is switched on/off.

One can see that the amplitude increases - also there is a baseline offset. What can cause the offset? Well in figure 6 the preAmp is switched on, in figure 7 the same is measured but now with the preAmp switched off (LV set to zero).

figure 7: CDC-noise after the capacitor on the HV distribution board - LV is set to zero

Well, the base line moved closer to zero, mark that the vertical scale in figure 7 is different than figure 6.

Let's look at the signals. After putting 1800 V on the CDC and placing my probe after the capacitor that decouples the signal from the HV I was not able to distinguish a signal in the noise, i had to increase the HV till 2130 V till I was able to see the signal showed in figure 8.

figure 8: Signal after capacitor on the HV distribution board.

How could it be that the previous pre-Amp could see nice signals at 1800 V -> did it filter out the 100 MHz noise somehow? I am planning to try a suggestion made by Gerard to double the capacitance of the capacitor that decouples the signal from the HV, see how much it improves the situation. Anyway, figure 9 shows how this signal (showed in figure 8) looks like after the (new) preAmp.

figure 9: Signal after the preAmp (pink).

The signal is barely visible.