- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Due to powersave requirements, my board has the ability to switch the FPGA / DDR memory / sensor part off, if not used for a certain time. I consider this similar to a hot-socketing behaviour, which is supported by Stratix III devices.
I'm now facing the problem, that the FPGA pwrreset_n input (on-board POR going to FPGA, STM32 and FX2 devices) is tied low by the FPGA IO during a short time when ramping the power rails up, causing the other devices (FX2 / STM32) to reset, which is unwanted. According to manuals "...the Stratix III device does not drive out until the device is configured..." so I'm a little confused about that. Further fact: Only boards assembled with EP3SL70 show the behaviour, EP3SL50 don't tie low the mentionned IO in an otherwise unchanged environment. I attached two CRT shots showing my power sequence, currently using ratiometric tracking of 2V5 for 1V8 and 1V1VCCL. Coincident tracking of the same rails showed the same behaviour. To my understanding, the powerup sequence is within the recommended behavior concerning rise time and monoty. Any idea why this IO drives low at this time where it shouldn't?Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
After having investigated all my fife boards, I observe:
- None of the two dips occur on the board with an EP3SL50 device (I have only one)
- The first dip (at the end of 3V3 ramp up) occurs on all 4 EP3SL70 boards
- The second dip (during 2V5 and lower voltages ramp up) occurs on only one of the EP3SL70 boards.
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page