完善资料让更多小伙伴认识你,还能领取20积分哦, 立即完善>
扫一扫,分享给好友
嗨,大家好,
我遇到了一些奇怪的信号特征,我想我可能会找到一些帮助。 我有一个顶级vhdl代码,它有多达5个串联的模块,因为数据从一个模块流向另一个模块。 数据时钟速率为54MHz。 我正在使用chipcope pro,在167MHz的采样频率下进行调试。 现在我注意到,当我将端口映射到第五个模块时,来自第四个模块的一些信号变得非常嘈杂。 第五个模块的任务是读取这些信号并将它们存储到fifo中。 我在下面贴了一些chicpscope pro图片,以便你可以欣赏它们的不同之处。 请注意端口映射后信号8_6和7_7如何变得非常嘈杂。 所有其他信号(8_7,8_11,3_11等)由第五个模块在同一指令读取,但只有8_6和7_7变为有噪声。 我该如何排序这个问题。 我尝试使用中间信号,但无济于事。 谢谢你的时间。 最好的祝福 以上来自于谷歌翻译 以下为原文 Hi guys, I have met with some strange signal characteristics and I thought I might find some help. I have a top-level vhdl code which has up to 5 modules connected in series in the sense that data is being pipelined from one module to the other. Data clock rate is at 54MHz. I'm using chipscope pro, at a sampling freuqnecy of 167MHz for debugging. Now I notice that some signals, coming from the fourth module become very noisy as soon as I port map them to the fifth module. The task of the fifth module is to read these signals and store them into a fifo. I have posted some chicpscope pro pictures below so that you may appreciate the difference. Please notice how signals 8_6 and 7_7 become very noisy after portmapping. All the other signals (8_7,8_11,3_11 etc) are being read by the fifth module at the same instruction yet only 8_6 and 7_7 become noisy. How can I sort this issue. I tried using an intermediate signal but at no avail. Thanks for your time guys. Best regards |
|
相关推荐
5个回答
|
|
嘈杂?
这不是噪音。 它是1还是0.你关心吗? 在问题的时钟边缘应该是什么数据(不是167 MHz)? 可能你只是看到数据在结构中是如何延迟的,但是如果你遇到了计时(在所有路径上都有+松弛),你就不在乎了。 一般来说,我永远不会比我希望调试的数据域的时钟更快地运行Chipscope,因为我将会看到无关紧要的信息。 Austin Lesea主要工程师Xilinx San Jose 以上来自于谷歌翻译 以下为原文 Noisy? It is not noise. It is a 1 or a 0. Do you care? What is the data supposed to be at the clock edge it question (not at 167 MHz)? It may be that you are just seeing how the data is delayed in the fabric, but if you met timing (have + slack on all paths) you don't care about that. Generally speaking, I would never run Chipscope faster than the clock of the data domain I wish to debug, as I will then be presented with information that does not matter. Austin Lesea Principal Engineer Xilinx San Jose |
|
|
|
嗨奥斯汀,谢谢你的答复..是的你是对的,它不是真的吵,但我找不到更好的词来描述它..我真的不明白你的意思“可能是你只是
看看数据在结构中是如何延迟的。“嘈杂的”信号一直在逻辑0和1,当它们在一段时间内保持不变时。 我会理解是否有一些传播延迟,但我发现这种行为很奇怪..再次感谢 以上来自于谷歌翻译 以下为原文 Hi austin, Thanks for your reply.. yeah you're right, its not really noise but I couldn't find a better word to describe it .. I didn't really understand what you meant with "may be that you are just seeing how the data is delayed in the fabric".. The 'noisy' signals go logic 0 and 1 all the time, when they're supposed to be constant for some period of time. I would understand if there is some propagation delay but I find this behaviour very strange.. thanks again |
|
|
|
好,
您的采样频率高于数据的时钟频率。 在允许的时钟周期内,数据从0到1,1到0。 如果您以比所需更快的速度进行采样,您将看到各种中间状态(行为)。 数据必须在54 MHz有效。 这是唯一需要稳定的时间。 如果路径很长,则数据会一直变化到时钟边沿(如果满足时序,则整个周期内存在正斜率,可能只有几百皮秒,少于几百皮秒) ,数据正在发生变化(尚未到达目的地)。 Austin Lesea主要工程师Xilinx San Jose 以上来自于谷歌翻译 以下为原文 OK, You are sampling at a higher frequency than the data is being clocked. The data goes from 0 to 1, 1 to 0 o ver the allowed clock period. If you sample it faster than needed, you will see all kinds of intermediate states (behaviors). The data has to be valid at 54 MHz. That is the only time it needs to be stable. If the path is a long one, the data is changing right up to the clock edge (if timing is met, there is a positive slack, that may be only a few hundred picoseconds -- for the entire period, less those few hundred picoseconds, the data is changing (it hasn't gotten to its destination yet). Austin Lesea Principal Engineer Xilinx San Jose |
|
|
|
>数据时钟频率为54MHz。
我正在使用chipcope pro,在167MHz的采样频率下进行调试。 ChipScope ILA设计为同步状态捕获,时钟必须与用于数据和触发端口的时钟相同。 ------您是否尝试在Google中输入问题? 如果没有,你应该在发布之前。太多结果? 尝试添加网站:www.xilinx.com 以上来自于谷歌翻译 以下为原文 > Data clock rate is at 54MHz. I'm using chipscope pro, at a sampling freuqnecy of 167MHz for debugging. The ChipScope ILA is designed as a synchronous state capture and the clock MUST be the same clock that is used for the data and trigger ports. ------Have you tried typing your question into Google? If not you should before posting. Too many results? Try adding site:www.xilinx.com |
|
|
|
嗨,大家好,
谢谢你的回复。 我已经缩小了chipcope pro采样时钟,正如你所说,信号波形变得更加稳定。 谢谢你的时间。 问候 以上来自于谷歌翻译 以下为原文 Hi guys, Thanks for your replies. I have reduced the chipscope pro sampling clock and the signal waveforms became more constant as you said. Thanks for your time. regards |
|
|
|
只有小组成员才能发言,加入小组>>
2415 浏览 7 评论
2821 浏览 4 评论
Spartan 3-AN时钟和VHDL让ISE合成时出现错误该怎么办?
2292 浏览 9 评论
3372 浏览 0 评论
如何在RTL或xilinx spartan fpga的约束文件中插入1.56ns延迟缓冲区?
2458 浏览 15 评论
有输入,但是LVDS_25的FPGA内部接收不到数据,为什么?
1100浏览 1评论
请问vc707的电源线是如何连接的,我这边可能出现了缺失元件的情况导致无法供电
579浏览 1评论
求一块XILINX开发板KC705,VC707,KC105和KCU1500
441浏览 1评论
2000浏览 0评论
723浏览 0评论
小黑屋| 手机版| Archiver| 电子发烧友 ( 湘ICP备2023018690号 )
GMT+8, 2024-12-21 00:36 , Processed in 1.538699 second(s), Total 87, Slave 70 queries .
Powered by 电子发烧友网
© 2015 bbs.elecfans.com
关注我们的微信
下载发烧友APP
电子发烧友观察
版权所有 © 湖南华秋数字科技有限公司
电子发烧友 (电路图) 湘公网安备 43011202000918 号 电信与信息服务业务经营许可证:合字B2-20210191 工商网监 湘ICP备2023018690号