Rolf and Dave

Report on ISI driving of ITMY March 1st 2012

The problem is introduced by the addition of a fourth ADC into the hepi/isi models. There is a hard limit set in the code for the decimation filtering history arrays to only support up to 3 ADC modules. The addition of a fourth caused the ADC decimation filtering history to write into the space used for DAC upsampling filer history data, thereby causing some form of this ADC input to be output to the first DAC module. I believe this has been in the code since the early blue box days. While other arrays used by ADC/DAC data have been updated to use MAX_ADC_MODULES and MAX_DAC_MODULES some time ago to set array sizes, as defined in cdsHardware.h (presently 12 ea), these history arrays apparently got overlooked. This hard limit would not effect IOP tasks, or any models running at 64K, as they do not perform up/down sampling.

a new rcg tag 2.4.2 was created as a copy of 2.4.1 with the MAX ADC/DAC substitution installed. All LHO models were recompiled and restarted against 2.4.2. with the exception of:

h2susb6: h2sustmsy model would not compile, missing links. Running 2.4.1 on this front end for now

h2seib6: would not reboot, needs further investigation

h2psll0: hold off until friday, running 2.4.1 on this front end for now