- 积分
- 992
- 注册时间
- 2014-12-4
- 最后登录
- 1970-1-1
- 打卡等级:偶尔看看
- 打卡总天数:105
- 打卡总奖励:319
|
发表于 2018-4-10 10:58:45
|
显示全部楼层
我同意,但关键词最多,并非全部。
主要问题是当一个轴切换到另一个轴时。
首先主轴必须读取0.1ms,然后主机将信息发送给从机,另一个0.1ms。
如果主控制器无需从主轴读取坐标就可以将位置下载到所有轴,那么EtherCat就足够了。
大多数分布式控制网络每次下载一个位置。 EtherCat的默认值是一个只有8字节的CanOpen数据包。这意味着每次下载一个32位的位置。其他4个字节是命令位,并告诉从机包中有什么。
是否曾想过如何使用Sercos,EtherCat和罗克韦尔运动控制器等协议?我知道,因为我的公司为罗克韦尔生产Rockwell 1756-HYD02和1756-M02AS。
这些协议都是相似的。主站每次粗略更新都将位置下载到运动控制器。运动控制器必须内插以为其每个扫描时间生成新的位置
I agree, but the key word is most, not all.
The main problem is when one axis gears to another axis.
First the master axis must be read 0.1ms and then the master sends the information to the slaves, another 0.1ms.
If the master controller is download positions to all axes without having to read from a master axis then EtherCat will be good enough.
Most of the distributed control networks download a position each time. The default for EtherCat is a CanOpen packet which is only 8 bytes. This means a 32 bit position is download each time. The other 4 bytes are command bits and tell the slave what is in the packet.
Have thought about how protocols like Sercos, EtherCat and the Rockwell motion controllers work? I know because my company makes the Rockwell 1756-HYD02 and 1756-M02AS for Rockwell.
These protocol are all similar. The master downloads positions to the motion controllers every coarse update. The motion controller must interpolate to generate new positions for each of its scan times.
|
|