Re: [Discuss-gnuradio] Help needed regarding implementing spring mass system in GRC

[Michael Dickens]

Hi Afaq - Actually one could replicate this simulation in some capacity using async messages -- for all blocks, or just for the feedback portion. Async messages have no restriction on feedback as does streaming data -- see, e.g., the OFDM header/payload demux block in gr-digital & the related OFDM Tx / Rx examples. If you go for pure async messages, then you'll need to create a concept of "resolution" -- how many samples to process during each feedback iteration. If you go for the final feedback link being async, then GR will "define" the resolution for you, which may or not work for what you're trying to accomplish. The "resolution" defines the trade-off between CPU efficiency (measured, e.g., in terms of how many CPU cycles spent doing actual computations versus overhead tasks) and how quickly feedback is applied in the system: smaller values will be less CPU efficient but provide faster feedback, while larger values (up to a point) will be more CPU efficient but with more delayed feedback. I think either way you go, you'll need to write custom blocks (whether C++ or Python) to handle the very specific feedback mechanism. Hope this helps & good luck! - MLD

On Mon, Dec 4, 2017, at 11:57 AM, Dan CaJacob wrote:

Unfortunately, GNURadio doesn't allow for signal feedback the way Simulink does. You can do feedback *within* a block, but that probably defeats the flexibility you were seeking?

On Mon, Dec 4, 2017 at 9:33 AM Afaq Ahmed <address@hidden> wrote:

Hello,

I want to implement spring mass system with PID controller of example mentioned in following link 

[ http://ctms.engin.umich.edu/CTMS/index.php?example=Introduction&section=ControlPID ]

I want to create a flow graph exactly as given in the figure "Spring_mass_simulink" . Can you please help me how can i achieve this target? As i tried to implement it but unfortunately i always get an error of "Flow graph has loops"