LCD Datasheet

[danadak]

In the datasheet Cypress states LCD delay functions are independent of clock frequency.

That is not true, they are a direct f() of the clock frequency. I think author felt that
because the API read the OSC register, to get clk setting, that effectively made the
delay f() independent of clock frequency. Needless to say that is not the case, so
datasheet should reflect that.

Regards, Dana.

[bobmarlowe]

Hi Dana,
when I went through the source of the delay-function it looked as if it would work (as long as the clock was internal). Did you test the delay and found any differences or did I overlook something?
Bob

[danadak]

The point expressed in datasheet is delay is independent of clock frequency.
Thats ridiculus, otherwise PSOC has a built in time base, xtal, atomic clock,
cesium beam clock, we are not aware of......

The delay, created by instruction execution countdown, is derived from the clock,
internal or external, does not matter, except that it is derived from a source that
expereinces drift, offset,........ Datasheet author should not have used "independent
of clock", language, simply not true.

Regards, Dana.

[bobmarlowe]

I understand what you mean.
So the datasheet should read ".. independent of the SELECTED Clock frequency" because the API reads out which of the allowed frequencies is selected by the user and calculated (a table entry) the needed count of a loop to perform the required delay.

Bob

[danadak]

No, its dependent on the clock frequency.

The API figures out an approximate effective /n, but that does not matter,
whatever the clock is, it drifts, has an inacuracy, and therefore the delay
is dependent on clock accuracy.

The IMO that sets the instruction rate, is for internal clocks +/- 2.5%, hence
the delay that will be produced from one part to another is +/- 2.5%.

"independent" needs to be changed to "dependent".

Regards, Dana.