[MinnowBoard] System power/reset control architecture with PCIe

Anders, David david.anders at intel.com
Tue Mar 10 20:19:33 UTC 2015


1)      Yes this is an acceptable approach. It's not the cleanest solution, but it should work as long as your power up sequence is followed.

2)       Just remember the i2c interface on the HSE is at 1.8V , so if you are connecting it up to the mPCIe connector you will need to level shift it to 3.3V. make sure you include the appropriate caps in series for the PCIE_TX pair coming from the HSE.

3)      The two implementations we have tested use the PWROK or a basic RC for the PERST#  signal on the mPCIe connector. In fact all of the designs we have tested have zero ohm resistor options so either may be used. I would not recommend using the #PMC_PLTRST_R_V1P8 signal.

Feel free to post schematics for review! Good luck with the design!

Dave Anders

From: elinux-MinnowBoard [mailto:elinux-minnowboard-bounces at lists.elinux.org] On Behalf Of Leigh Gawne
Sent: Monday, March 9, 2015 5:10 AM
To: elinux-minnowboard at lists.elinux.org
Subject: [MinnowBoard] System power/reset control architecture with PCIe

Hi all,

I am in the process of designing a custom expansion board for the MinnowboardMax (really liking my experience with this board so far BTW :-) and have a question regarding PCIe power supplies.

One of the interfaces I am implementing on the expansion board is a Mini PCIe card slot. On my expansion board, I have all the required power suppliers for card slot, and hence don't need to rely on any power coming from the MinnowboardMax expansion connector.

There is a requirement in the Atom e38xx documentation (actually a PCI Express 2.0 spec requirement) that all power rails associated with PCIe should be valid for ~100ms prior to PMC_CORE_PWROK assertion.

>From what I can see in the MinnowboardMax schematic, there is an RC which controls the assertion of SYS_PWRGD providing 100ms delay as part of the power sequencing. My understanding is that as long as I ensure that my PCIe supplies are up before providing 5V to the MinnowboardMax, then I should be meeting this requirement. So a couple of questions just for clarification:

1)      Is this an acceptable approach to take given the MinnowboardMax power supply architecture? I have a seen there is a Lure expansion board that includes mini PCIe and I presume this derives the required supplies from those provided on the expansion header (perhaps the +5VSB supply, which should also mean this constraint is met).

2)      Any other recommended design approach? Backfeeding is always one area of concern I have which I need to manage carefully - just want to make sure there is nothing else I am overlooking/am not aware of.

3)      I am intending on using #PMC_PLTRST_R_V1P8 as the system reset signal for PCIe card slot and some other devices (with appropriate buffers) - any thoughts here?

Many thanks for any information/clarification!



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.elinux.org/pipermail/elinux-minnowboard/attachments/20150310/d607f8ea/attachment-0001.html>

More information about the elinux-MinnowBoard mailing list