FSRN (Correction Factor)

Hello everybody,

While I was checking the FSRN logic, I just saw when Isochronous active by L83SCI_CMD, FSRNI will be like this:
FSRNI = FSKRN3*(TNRI-TNH)- FSRN + FSR }+FSRNI

Why FSRN and FSR should be added to this Formula?
By FSR in this formula, means Current active FSR?

thanks for reply ...
 
no_bug,

I don't believe you've properly copied the formula; at least one left French curly brace is missing, and I'm not sure about all of the signal names, either.

Does the unit you are working on have "regular" Droop or Constant Settable Droop?

And, what problem are you perceiving, or what change are you contemplating making?
 
You are right. the formula is like this:

FSRN = {(FSKRN2*(TNR-TNH))+FSKRN1} + {FSKRN3*(TNRI-TNH)- FSRN + FSR}

Right now, I don't understand why FSRN and FSR should add to end of this formula which it is FSRNI?
 
no_bug,

"Straight" ("Normal") Droop Speed Control is proportional control. Constant-Settable Droop Speed Control uses and integral control function in addition to the proportional function.

If I remember correctly (and I don't have any Mark* software to look at as I write this) Isochronous Speed Control is proportional-plus-integral control. I still think you haven't properly listed the Isoch speed control formula. And, I think the fact that Isoch speed control includes integral action in addition to proportional action answers your question. Have a closer look at the "blocks" and mathematical "primitives" used in the calculation of Isochronous Speed Control in the Mark* at your site. I think you will find there is an integral function being performed.

Hope this helps!
 
P

PC Load Letter

Hi No_Bug,

Based on the formula you have given, the machine at your site has standard leniar droop implementation.

As CSA has mentioned you don't have the formula correctly written. You forgot about the Z^-1 term. This is the Z transform version of unit time delay. The correct equation needs the (t) or time element where its the value at time t.
<pre>
FSR(t) = { (TNR - TNH(t)*FSRKN2 + FSRKN1}
+ { FSRNI(t-1) + (TNRI -TNH(t))*FSRKN3}
+ {FSRN(t)- FSR(t)}

</pre>
The first term in the curly braces " {(TNR - TNH(t)*FSRKN2 + FSRKN1} " is the proportional control equation. This is your standard linear droop control.

The second term in the curly braces { FSRNI(t-1) + (TNRI -TNH(t))*FSRKN3} is the intrgral potion which is used to control the speed of the machine to TNRI value (isochronous speed). You may not be used to this equation; it's what is called the "Velocity From" equation for a integral control. It's an implementation used in discrete time or digital control systems such as the Mark VI control.

The third term under normal operation condition will be zero as FSRN and FSR will be the same. (All the various FSR values FSRN, FSRT, FSRACC, etc. go to a min select block). That term is kept for transition purposes when FSRN is not equal to FSR of the system.

So, to summarize:
Droop control - Propotional Control
Isochronous Control - Propotional + Integral Control

Note - This PI control is valid around the 100% mark +/- FSKRN6 value; if for some reason the FSRNI value is more than 0.5%, then the droop reference TNR is raised or lowered. During that time LISOK signal goes low cutting off the Integral part of the equation. The droop reference is raised or lowered till the error value FSRNI is below 0.5%. after that LISOK goes high and the normal PI control resumes.

CSA would have more to add to this I am sure.


 
Top