Counting encoder pulses

Islam Elnady said:

the needed clock for the counter is external

My friend - firm/I have done pretty much the same as you describe - yet we've "not" found the need to employ a "counter" clock.   Does your magnetic encoder datasheet (clearly) state that such an (external) clocking IS required?

If freed from that "external clocking" - might an MCU Timer - configured for "Edge Count" mode - very well, "meet your objectives?"

Should you, "buy into" the above - a second Timer - this configured as "One Shot" can serve as a, "Gating Control" for the (earlier mentioned) Edge Counter.   The one-shot Timer (only) enables the Edge Counter for a (well determined) interval.   Upon the one-shot's "time out" - the Edge Counter is frozen (disabled) - its contents read & stored - and then processed to determine motor speed.   (i.e. Counts per unit time is directly proportional to motor speed)   And - if you (well) choose the one-shot's duration - your "Counts to RPM calculation" can be substantially simplified.

Firm/I prefer the "One Shot" Timer mode to "Periodic" as it proves less "demanding" upon the Edge Counter's, "Read, store, speed calculate & display."   (note that we also display the motor's Actual Speed, Target Speed, Current draw, Advance/Retard settings (if any) & Torque etc.)   Only when completed do we (again) clear the Edge Counter & launch a new, "Count acquisition sequence."  (and that's why the "flexibility" & "accommodation" of the one-shot is our choice over the "too demanding" (inflexible) periodic.)

It would be useful if you'd link to your magnetic encoder.   Firm/I employ some from "CUI" (and are unrewarded) for such mention...   (can that be fair?)

It should be noted that "if" your encoder produces "Quadrature Pulse Output" AND if your MCU includes a QEI peripheral - that implementation may work well.   (and the quadrature method supplies "motor direction" - absent in the method I've (earlier) described...)

That said - method presented (up top) serves as a, "Universal Solution" - able to succeed even w/"lesser" MCUs - and as it well complies w/KISS (via 2 part, individually managed design) it's faster, easier & more instructive than a QEI module - where (most) facts & methods are "hidden."