On 17/03/15 15:08, Marcus Müller wrote: > Hi Daniele, > >> Also, I don't see why you cannot combine other blocks to realize >> your own PLL implementation. Where did you read that you "can’t use >> companion to add and drop the current available blocks to construct a >> phase locked loop"? > Wai Yee has a point here -- you can't have a circle in your (sample > stream) flow graph -- GNU Radio doesn't allow that for reasons of > causality. > > You can have a "loose" loop with feedback done with message ports, but > that's not the same as a control loop (since there's no defined delay > between the sample for which a error was estimated and the sample where > the error correction is applied when using the asynchronous messages). > > Thus, you can't drag'n'drop together a "proper" control loop from > discrete things like adders and integrators that you could design and > analyze theoretically.
Hello Marcus, I realized this while having my coffee this morning. I should avoid to reply to emails before assuming the right amount of caffeine :) > In effect, GNU Radio's control loops are monolithic blocks -- that's > somewhat sad from a modularity perspective, but then again, there's the > control loop base class that allows you to implement arbitrary control > loops rather easily Last time I looked into it, this base class is not that generic: it can be easily used to implement phase-lock loops, but it did not seem designed to control other quantities, like amplitude for example. But I may have had a too quick loot at it. Cheers, Daniele _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
