Login

DESY discussion on MTCA RTM and SIS8300KU

→ Europe/Zurich
CERN

CERN
Gregoire Hagmann (CERN)
Description

Discussion for the DWC8VM1LF modification and on the SIS8300KU noise measurement

 

https://cern.zoom.us/j/68905346121?pwd=ZFvWsacb3JmqrcR364UGFh0vVHuZ7S.1

Hide

Participants

Frank Ludwig, Uros Mavric (DESY), Diego Alamon, Vincent Ripaud, Grégoire Hagmann (CERN)

Summary of the mains points discussed:

1) The modification required by CERN would be the « 2nd generation » of the DWC8VM1 . DESY is fine to provide the support to apply the modifications. DESY is also working on a « 3rd generation » with even higher noise requirement. The 3rd generation timescale does not seem to match CERN needs

2) Frank is in favour of keeping an amplifier in the IF chain. To be studied in  details as the current amplifier is not good enough.

3) The source of the 1/f noise of the new demodulator seems to come from the mixer itself. DESY proposed to evaluate the mixer proposed (to be confirmed). 

4) DESY did not observed big  noise/spurs performance spread between SIS8300 module (L2 version). Discrepancies between channel was observed but not at the same level as presented by CERN. (asynchronous or synchronous sampling at DESY?)

5) Uros suggested to evaluate spurs amplitude vs RF input power

6) SIS8300KU noise floor measured by CERN consistent with DESY measurement at first glance. Amplitude and phase noise are measured after IQ demodulation. Post processing (FFT) was done with the objective recreate the single-sided band PM/AM noise equivalent to a signal source analyser. CERN needs to validate the measurement in close loop with a phase noise analyser (FSWP)

7) New ADC DESY selected is 16 clock cycle latency  (160ns at 100MSPS). Higher than AD9268 from the SIS8300KU, an issue for CERN as the latency must be minimised

8) Ratio of 10 between IF and CLK seems a good choice for DESY. A ratio of 8 is not recommended as twice the factor of 4 which brings many alias with IQ demodulation.

9) the LHC RF frequency is sweeping by 5kHz. Spurs at 10*deltaF might or might not be armful for the system. Hard to predict. It is armful (emittance growth) only if the noise is excessive at discrete frequency points in the PM/AM with are the betatron frequency sidebands (+/-3kHz offset wrt to RF and RF+/-revolution frequency of 11.2kHz). 

 

Following steps

The following steps regarding the DWC8VM1LF modifications were discussed:

- DESY will study the proposal on the DWC8VM1LF modifications in the following three weeks

- Next meeting with the tentative date on July 3rd at 13:00 (to be confirmed)

- Goal is to have a prototype board before end 2026 at the latest. 

- CERN will study the CSI recommended by DESY. Reference paper https://journals.aps.org/prab/pdf/10.1103/jhc3-dtzw

There are minutes attached to this event. Show them.