Control Engineering: Block Diagram

[wantan7671]

Hi guys, this is quite an inbetween question , not really homework but still theory.

http://img850.imageshack.us/img850/1290/81696653.jpg

Came across this diagram and was wondering if the arrow (in blue box) can be ignored ? Quite irritated by it now since I can't be sure how I should create the Transfer Function or obtain the Control Canonical Form.

Any help would be fantastic !

 

[Floid]

There is no way of knowing without more context. There might be some unknown gain term there... but without the problem statement for which that block diagram describes who knows.

 

[wantan7671]

the question is just to find the transfer function for that block diagram. I drew the blue box myself. Just wondering if it's actually valid to have a drawing like this with "loose ends"

 

[jim hardy]

"Just wondering if it's actually valid to have a drawing like this with "loose ends" "

Is it really a loose end? Looks to me like input to 2nd integrator stage.
in my day it would have been fine.

Were one building an analog computer that'd be a place one could hook a recorder.

I recognize that as a bear to analyze
would be lots quicker to build with opamps.

old jim

 

[wantan7671]

Thanks for the insight Jim. So for now, if nothing is connected there, can I assume the entire block only has U (input) and Y (output) points ? And obtain the Transfer Function directly? (or is there another technique for these sort of things?)

 

[jim hardy]

I think so. It was mid 1960's when i took controls course, so i am really rusty and shudder at the thought of all that algebra now.

But looking at that sketch it seems to me --

they could have drawn it as one horizontal chain
then the X2 line wouldn't have to cross X1 where X2 feeds back to sum with u, so it would look less formidable.

isn't it just two cascaded lags with output of 2nd fed back to input of first one?

i hope you'll post the transfer function.

Think it'll oscillate?

old jim