Runtime of Freemaster recorder is too much!

[kiddyang]

I put the FMSTR_Recorder in PWM reload interrupt service routine, and tested its runtime. When triggered, its runtime could be more than 50us at most!

PWM frequency is 10kHz and period is 100us, if Recorder function takes too much time, then it will infuence algorithm running in PWM reload interrupt very badly.

How could this problem be solved?

P.S. If I use only scope, its runtime still could be more than 25us, which is also not acceptable.

[sagardhavali]

Hello Kidd,

We also have the same problem. We cannot use Freemaser at every 100us but yet we want to capture the data to see what is happening in the algorithm.

We are evaluating other options:

1. Use P&E micro's Multilink Universal debugger to connect to JTAG and connect the FreeMaster to it via JTAG for the above purpose.
2.  Use Vector's VX1000 tool box to connect to JTAG and use CANape to capture the data.

With 2nd option, we have been able to capture 32bytes at every 100us successfully at worst case execution!

Hope it helps.

Sagar