AnsweredAssumed Answered

How is float implemented in ADS

Question asked by ali_i on Jun 3, 2007
Latest reply on Jun 5, 2007 by ali_i
I believe there is an issue with the precision of float in ADS.  As part of my System simulations it is necessary to write AGC controllers in C++, which are then called from ADS using wrappers.  I have used this technique successfully for some time.  

However I have found a problem recently.  I am monitoring the level of a digital signal and calculating the RMS value.  The commanded gain based on the RMS value seemed to vary considerably from one simulation run to the next with very little change in the input signal.  However this problem would not show itself when debugging the code using .Net and ADS.  The most common cause of this sort of an issue is initialisation of variables, however this was not the case.  I placed tracer code to debug in run-time mode and found that if the signal being monitored was large in value then the squared value was correct.  Initially I started suspecting the “pow† 

Outcomes