The settings of the MBS daq during the first data taking with fragments around 194Pb (W.Korten experiment) was not entirely correct:

1. both FEBEX nodes had the very same sub-sytem identifier 0x200. The timesorter did merge the two data streams into the lmd files. Now the events cannot be easily separated/identified, 
because both have the same WR header. The only option to identify which FEBEX was the source for an event might be to look for the actual channels receiving hits, because they are slightly 
different for both arrays/DAQs. This might only work for "good" events though...

2. The trigger signals transmitted to the scaler DAQs (VME & NTCAP) had a different input for both DAQs. The GAGG trigger (outside) was using the FEBEX self-trigger as input only, while 
the BGO trigger (inside) was using the FEBEX self-trigger OR the 10Hz sync trigger as input.

The runs with run number lower than 0016 are affected by the above points.

Now the setup has been changed, for run numbers 0017 and higher the following should be valid:

1. WR identifiers
The BGO DAQ on x86l-124 (inside) does use WR sub-system identifier 0x100
The GAGG DAQ on x86l-131 (outside) does use WR sub-system identifier 0x200
The VME DAQ on x86l-204 does use 0x400

2. transmitted triggers
Both DAQs now transmit only the FEBEX self-trigger to the scaler DAQs.

3. Gate Generator for PEXARIA input
We now use a Gate Generator to provide 200ns TTL signals for the PEXARIA, with the goal of avoiding double trigger/pulses (a known issue of the old EXPLODERs we are using)

4. SPARSIFYING
We turned off SPARSIFYING for both FEBEX DAQs. So 
in f_user.c 
#define DATA_FILT_CONTROL_DAT 0x80
was changed t0
#define DATA_FILT_CONTROL_DAT 0x82