第七篇?Rico board实现4路“DA”输出

时间：10-02 整理：3721RD 点击：

因为我要通过4路DA输出，控制两路电机，而Rico Board上又没有直接输出DA的部分，所以我考虑为了节省资金，准备通过4路PWM输出来实现4路DA的输出功能。

第一步：

先把4路DA正常输出吧，用的hrt高精度时钟，主要为了实现比较平滑的输出，pwm的频率越高，低通滤波后的输出越平滑，这也是我所希望的。

硬件上，用了4个GPIO实现4路输出，如下所示：

实现的效果如下：

我同时测试了两路输出，因为示波器只有两个通道啊。

测了三组：

测试的硬件连接图：

示波器的显示如下所示：

今天焊上了滤波电路，也测了三组，终端显示如下：

分别表示1650mv，330mv，3267mv。

滤波后的输出波形，示波器测试如下：

但是把信号加入到电机驱动时，发现电压被拉低或是被太高，准备明天再加一级跟随试一试。

下图是我的滤波电路和测试现场图：

驱动代码如下：

/*************************
*
*[url=home.php?mod=space&uid=1455510]@file[/url] mada.c
*@date 2016.11.12
*author iysheng
*
*************************/
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
MODULE_LICENSE("GPL");
MODULE_AUTHOR("iysheng ");
MODULE_VERSION("0.1");
#define MADA_NUM 4
#define MADA_NAME "MADA"
#define NAME_SIZE 10
#define MADA_TIME_SCALE 1000
#define VALUE_SIZE 8
#define MADA_MAJOR 0
static dev_t mada_devnum;
static int mada_major = MADA_MAJOR;
#define P_VOLTAGE 33
#define MADA_VOLTAGE(devp) (P_VOLTAGE*(devp->mada_high)/10)
int mada_gpio[MADA_NUM] = {164, 165, 166, 167};
struct mada_dev{
struct cdev cdev;
struct device *device;
int mada_gpio;
unsigned int mada_high;
unsigned int mada_voltage;
dev_t devnum;
char mada_name[NAME_SIZE];
struct hrtimer mada_hrt;
bool mada_gpio_value;
bool mada_hrt_state;
} *mada_devp;
struct class *mada_class;
/*
int MADA_VOLTAGE(struct mada_dev * devp)
{
int voltage;
voltage = P_VOLTAGE * devp->mada_high;
return voltage;
}
*/
enum hrtimer_restart mada_hrt_callback(struct hrtimer *hrt)
{
struct mada_dev *devp;
devp = container_of(hrt, struct mada_dev, mada_hrt);
devp->mada_gpio_value = ！devp->mada_gpio_value;
gpio_set_value(devp->mada_gpio, devp->mada_gpio_value);
if(devp->mada_gpio_value == true)
hrtimer_forward_now(hrt, ns_to_ktime(devp->mada_high * MADA_TIME_SCALE));
else
hrtimer_forward_now(hrt, ns_to_ktime((1000 - devp->mada_high) * MADA_TIME_SCALE));
return HRTIMER_RESTART;
}
int mada_open (struct inode *inode, struct file *filp)
{
struct mada_dev *devp = container_of(inode->i_cdev, struct mada_dev, cdev);
filp->private_data = devp;
if(devp->mada_hrt_state == false)
{
hrtimer_init(&devp->mada_hrt, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
devp->mada_hrt.function = mada_hrt_callback;
hrtimer_start(&devp->mada_hrt,ns_to_ktime(devp->mada_high),HRTIMER_MODE_REL);
devp->mada_hrt_state = true;
}
printk(KERN_INFO "mada_open func.\n");
return 0;
}
int mada_release (struct inode *inode, struct file *filp)
{
printk(KERN_INFO "mada_release func.\n");
return 0;
}
ssize_t mada_write (struct file *filp, const char __user *buf, size_t count, loff_t *ppos)
{
int ret;
char value[VALUE_SIZE];
struct mada_dev *devp = filp->private_data;
ret = copy_from_user(value, buf, sizeof(buf));
if(0 ！= ret)
{
printk(KERN_INFO "error in writing and errnum is %d.\n", ret);
return 0;
}
else
devp->mada_high = (unsigned int)simple_strtoul(value, NULL, 10);
return sizeof(buf);
}
ssize_t mada_read (struct file *filp, char __user *buf, size_t count, loff_t *ppos)
{
int ret;
struct mada_dev *devp = filp->private_data;
char value[VALUE_SIZE];
devp->mada_voltage = MADA_VOLTAGE(devp);
sprintf(value, "%d", devp->mada_voltage);
ret = copy_to_user(buf,value,sizeof(value));
if(0 ！= ret)
{
printk(KERN_INFO "error in reading and errnum is %d.\n", ret);
}
else
printk(KERN_INFO "%s mada_voltage is %s.\n",devp->mada_name,value);
return ret;
}
struct file_operations mada_fops = {
.open = mada_open,
.release = mada_release,
.read = mada_read,
.write = mada_write,
};
int mada_setup(struct mada_dev *devp, int min, int gpio_num)
{
int ret = 0;
printk(KERN_INFO "mada_setup function.\n");
if(！gpio_is_valid(gpio_num))
{
printk(KERN_INFO "invalid gpio_num:%d.\n", gpio_num);
return -1;
}
else
{
gpio_request(gpio_num,"sysfs");
gpio_direction_output(gpio_num,false);
gpio_export(gpio_num,false);
}
devp->mada_gpio = gpio_num;
devp->mada_high = 500;
//devp->mada_voltage = MADA_VOLTAGE(devp);
devp->mada_hrt_state = false;
devp->mada_gpio_value = false;
devp->devnum = MKDEV(mada_major, min);
sprintf(devp->mada_name, "mada%d", min);
cdev_init(&devp->cdev,&mada_fops);
ret = cdev_add(&devp->cdev,MKDEV(mada_major, min),1);
if(ret)
{
printk(KERN_INFO "add mada%d fail. error is %d", min, ret);
memset(devp, 0, sizeof(struct mada_dev));
gpio_free(gpio_num);
gpio_unexport(gpio_num);
return ret;
}
else
devp->device = device_create(mada_class,NULL,MKDEV(mada_major, min),NULL,devp->mada_name);
return 0;
}
void mada_del(struct mada_dev *devp)
{
hrtimer_cancel(&devp->mada_hrt);
gpio_unexport(devp->mada_gpio);
gpio_free(devp->mada_gpio);
device_destroy(mada_class,devp->devnum);
unregister_chrdev_region(devp->devnum, 1);
memset(devp, 0, sizeof(struct mada_dev));
kfree(devp);
}
int __init mada_init(void)
{
int ret = 0, i;
printk(KERN_INFO "mada_init begin.\n");
mada_devp = kzalloc(MADA_NUM * sizeof(struct mada_dev), GFP_KERNEL);
if(IS_ERR(mada_devp))
{
ret = PTR_ERR(mada_devp);
printk(KERN_INFO "vmalloc_fail and err is %d.\n", ret);
goto fail;
}
if(mada_major == 0)
{
ret = alloc_chrdev_region(&mada_devnum,0,MADA_NUM,MADA_NAME);
mada_major = MAJOR(mada_devnum);
}
else
{
mada_devnum = MKDEV(mada_major, 0);
ret = register_chrdev_region(mada_devnum,MADA_NUM,MADA_NAME);
}
if(ret < 0)
{
printk(KERN_INFO "register_fail and err is %d.\n", ret);
goto fail;
}
mada_class = class_create(THIS_MODULE, MADA_NAME);
for(i=0; i<MADA_NUM; i++)
{
mada_setup(mada_devp+i, i, mada_gpio[i]);
}
printk(KERN_INFO "mada_init！\n");
return 0;
fail:
return ret;
}
void __exit mada_exit(void)
{
int i;
for(i=0; i<MADA_NUM; i++)
mada_del(mada_devp+i);
class_destroy(mada_class);
printk("mada_exit！\n");
}
module_init(mada_init);
module_exit(mada_exit);

复制代码

主要部分是4路高精度时钟。

上一篇：DDR3的MIG（xilinx）无法自动初始化和校准
下一篇：FPGA 如何与ADC max11300 通信

栏目分类