Previously on e744:
After a chat with Charles Houarner we decided to have configuration with a Muvi card per VXI crate. This is supposed to reduce the dead time (DT) by parallelising the crate readout. This have a few implication:
- The slave crate need a GMT, because the MUVI read out was never tested using a GAMER
- The GMT is not designed to act as centrum receiver, so a centrum slave module is installed in the slave crate
- The action file compilation is not designed to have a centrum receiver, so the compilation has to be done by Fred. He is working on fixing the issue, in the mean time, we SHOULD NOT try to change ADC/QDC/TDC label and recompile throught DAS
- The slave GMT needs to be block correctly, Charles setup the VIS2/BLK signal with a logic FIFO.
- The slave GMT is triggered by the master GMT VAL
- The MUVI from the Slave VXI is still triggering the MASTER GMT, as usual.
After all this setup, we change the NARVAL topology to have two crate and a EBYEDAT merger. This merger prove to be no longer functional. Clotilde had a look but have no clue so far why it is not working.
We switch to MFM data format with a EBY2MFM merger. Because of this change of format we had to change both Online and Offline program in order to extract the EBYE buffer from the MFM frame.
This is no tested and working, for some reasons it worked only using ROOT6. |