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>
我在找到两个输入时钟之间的多路复用可路由解决方案时遇到了问题。
该设计在40MHz范围内的Bank 0上有一个差分时钟,另一个在75MHz范围内的Bank 3上是单端。慢速时钟当前直接进入PLL,但其中一个输出通过原始频率和 需要找到它的BUFGMUX方式,以便我可以在它(40MHz)和另一个(75MHz)输入之间进行选择。 我在Place& amp; Route抱怨IOB无法沿时钟路径到达BUFGMUX。 在这一点上,我无法找到这个问题的全貌,或者应该采取什么措施来解决它。 有没有办法在器件的相对两侧的时钟之间进行多路复用,或者是PLL的输出问题? 现在,路径看起来像: 慢时钟,Bank 0 - > IBUFGDS - > PLL_BASE - > BUFG - > BUFGMUX 快速时钟,银行3 - > IBUFG -------------------------------------- ^ 目标:xc6slx16 如果我忘记了任何重要信息,请告诉我! 以上来自于谷歌翻译 以下为原文 I'm having problems finding a routable solution to mux between two input clocks. The design has one differential clock coming in on Bank 0 in the 40MHz range, the other is single ended on Bank 3 in the 75MHz range. The slow clock currently goes straight to a PLL, but one of the outputs passes through the original frequency and needs to find it's way to a BUFGMUX so that I can select between it (40MHz) and the other (75MHz) input. The error I get during Place & Route complains about the IOB not being able to reach the BUFGMUX along a clock path. At this point, I can't find a complete picture of what this issue is, or what should be done to fix it. Is there any way to mux between clocks on opposite sides of the part like this, or is the problem with the output of the PLL? Right now, the path looks like: Slow clock, Bank 0 -> IBUFGDS -> PLL_BASE -> BUFG -> BUFGMUX Fast clock, Bank 3 -> IBUFG --------------------------------------^ Target: xc6slx16 Let me know if I've forgotten any important information! |
|
相关推荐
1个回答
|
|
与Spartan-6上的BUFGMUX输入的连接非常有限。
对于任何给定的BUFGMUX,只有少量(大约4个)时钟输入可以到达MUX - 这些在Sparan-6用户指南UG387图1-3和1-4(及其附近的表格)中有记录。 如果你的两个时钟不在同一个BUFGMUX的I0和I1输入端,那么你就不能将它们(直接)复用在一起。 DCM / PLL输出和BUFGMUX输入之间的连接更加稳健,但仍然不完整。 器件上半部分的PLL / DCM可以达到32个BUFG的前16个,底部的DCM / PLL可以到达底部的DCM / PLL。 时钟输入可以达到哪个PLL以及哪个DCM更加复杂和混乱的限制! 所有这些限制都在UG387中详细说明。 它们很复杂,因此您需要阅读本节(介于3到17次之间)以确保您理解它们。 只有这样你才能弄清楚这个MUX是否可以完成。 显然你现在正在做什么(一个是PLL而另一个是直接的)不起作用。 也许在另一个上使用DCM,或在两者上使用PLL或...... 祝你好运。 Avrum 以上来自于谷歌翻译 以下为原文 The connections to the BUFGMUX inputs on the Spartan-6 are VERY limited. For any given BUFGMUX there are only a small number (around 4) clock capable inputs that can reach the MUX - these are documented in the Sparan-6 User Guide UG387 figures 1-3 and 1-4 (and the tables near them). If your two clocks are not on a pair that can reach the I0 and I1 inputs of the same BUFGMUX then you simply can't mux them together (directly). The connections between the DCM/PLL outputs and the BUFGMUX inputs are more robust, but still not complete. The PLLs/DCMs in the top half of the device can reach the top 16 of 32 BUFGs, and the DCM/PLLs on the bottom can reach the bottom ones. The restrictions on which clock input can reach which PLL and which DCM is even more complex and confusing! All of these restrictions are spelled out in UG387. They are complex so you will need to read this section (somewhere between 3 and 17 times) to make sure you understant them. Only then will you be able to figure out if this MUX can be done. Clearly what you are doing now (a PLL on one and direct for the other) doesn't work. Maybe with a DCM on the other one, or a PLL on both or... Good luck. Avrum |
|
|
|
只有小组成员才能发言,加入小组>>
2370 浏览 7 评论
2786 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?
2255 浏览 9 评论
3330 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?
2420 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?
743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
531浏览 1评论
求一块XILINX开发板KC705,VC707,KC105和KCU1500
344浏览 1评论
747浏览 0评论
1948浏览 0评论
小黑屋| 手机版| Archiver| 电子发烧友 ( 湘ICP备2023018690号 )
GMT+8, 2024-11-14 13:18 , Processed in 1.238010 second(s), Total 79, Slave 62 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