bertos/cfg/cfg_ramp.h File Reference

Configuration file Ramp algorithm module. More...

Go to the source code of this file.

Defines
#define	RAMP_USE_FLOATING_POINT   0
	Define whether the ramp will use floating point calculation within ramp_evaluate().
#define	RAMP_CLOCK_SHIFT_PRECISION   2
	Number of least-significant bits which are stripped away during ramp evaluation.
#define	RAMP_PULSE_WIDTH   50
	Negative pulse width for ramp.
#define	RAMP_DEF_TIME   6000000UL
	Default ramp time (microsecs).
#define	RAMP_DEF_MAXFREQ   5000
	Default ramp maxfreq (Hz).
#define	RAMP_DEF_MINFREQ   200
	Default ramp minfreq (Hz).
#define	RAMP_DEF_POWERRUN   10
	Default ramp powerrun (deciampere).
#define	RAMP_DEF_POWERIDLE   1
	Default ramp poweridle (Hz).

---

Detailed Description

Configuration file Ramp algorithm module.

Version:

Id

cfg_ramp.h 2582 2009-04-20 12:33:48Z batt

Author:

Daniele Basile <asterix@develer.com>

Definition in file bertos/cfg/cfg_ramp.h.

---

Define Documentation

#define RAMP_CLOCK_SHIFT_PRECISION   2

Number of least-significant bits which are stripped away during ramp evaluation.

This setting allows to specify larger ramps at the price of less precision.

The maximum ramp size allowed is 2^(24 + RAMP_CLOCK_SHIFT_PRECISION), in clocks. For instance, using RAMP_CLOCK_SHIFT_PRECISION 1, and a 8x prescaler, the maximum length of a ramp is about 6.7 secs. Raising RAMP_CLOCK_SHIFT_PRECISION to 2 brings the maximum length to 13.4 secs, at the price of less precision.

ramp_compute() will check that the length is below the maximum allowed through a runtime assertion.

Note:

This macro is used only for the fixed-point version of the ramp.

Definition at line 74 of file bertos/cfg/cfg_ramp.h.

#define RAMP_USE_FLOATING_POINT   0

Define whether the ramp will use floating point calculation within ramp_evaluate().

Otherwise, a less precise fixed point version will be used, which is faster on platforms which do no support floating point operations.

Note:

Floating point operations will be always done within ramp_compute() to precalculate values, so there has to be at least a floating point emulation support.

Definition at line 52 of file bertos/cfg/cfg_ramp.h.