Hello PVWB ,
Thank you for contacting the Intel Embedded Community.
In order to better understand your questions, we would like to address the following questions:
Could you please clarify us if the PacketRunner2i is related to the device mentioned at the following website?
Could you please give us the pictures of the top side markings of the 82C226 units related to your questions?
Please give us all the information that may answer these questions,
Thanks in advance for your collaboration.
The link you have is for the board we are working on. The picture is from a
previous version with a Cavium processor, as the data sheet predates the
board. The block diagram is correct however.
I am attaching a sketch that shows the location of the various numbers.
I don't have a suitable camera to take an image of the entire part.
Hello PVWB ,
Thanks for your reply.
In order to better understand this situation, we would like to address the following questions:
Could you please confirm us if the affected design worked properly? Has it started to fail when you implemented changes on it? Please let us know these changes.
Could you please inform us if the schematics and layout related to this project have been reviewed by Intel, please give us all the information related to these processes?
Could you please tell us how many Intel(R) C226 Chipsets are related to this situation and where they have been acquired? By the way, it will be appreciated from our side if you give us the markings of the affected units.
Thanks in advance to give us all the requested information.
We have had three revisions of this board. We only get this
problem on the latest revision, which is Rev 2. We made two
changes to the PCH device. Our BIOS suppliers could never get
24 MHz out of pin AT9 for the SuperIO chip, so we just used a
buffer from an existing 24MHz oscillator. Pin AT9 is now not
connected to anything.
We added a device to the LPC bus and we used pin AN9 as a clock
source for the additional device.
We have compared the Gerber files for the boards and there do
not seem to be any additional, unwanted changes to the PCH area
of the board.
The Rev 0 design was reviewed by Intel and we incorporated most
of their advice into the Rev 1 design. Most of the differences
were due to our use of an FPGA instead of the discrete logic
liked by Intel. In the Rev 2 design, we fixed DDR3 routing
lengths and 10G Ethernet signals on the board, as well as the
PCH changes listed above.
The connection between VCCASW and ground is in the device.
After removing the part from the board, the short on the board
disappears. We can measure the short on the balls on the part.
The short is not present on parts we have in stock, not assembled
onto boards, based on a small sampling.
The board from which we removed the part has never had power
applied, so the short is not the result of a power supply voltage
being too high or in the wrong sequence.
Rev # Serial PCH part markings
Rev 2 115 SR179 11E406C086 MD420160A04158 G31426 01 ERJ has short
Rev 1 107 SR179 11E406C086 MD420160A04149 G31426 01 ELN good
Rev 0 101 SR179 11E406C086 MD420160A04962 G31426 01 EEP good
Loose SR179 11V5518066 N1551066A05345 G31426 01 YOK12 No short on VCCASW
I'll look into the source of these parts. Also, based on the above marking
information, are these all 82C226 devices?
I have read through the thermal document.
I did notice that this document considerably
predates the 82C226 device.
Has the thermal profile changed for the 82C226
at any time? Have other customers experienced
this rather unusual problem with the power supply
shorts inside the PCH part after soldering? We are trying to find
the cause of this so we can avoid the problem
reoccurring in the future. We had four out of four
boards have shorts between one or more power
rails and ground. The shorts appear to be in the
82C226 devices after soldering. We tested other parts
in stock and they do not have the shorts to ground.
Thanks for your update.
The information that may help you is stated in section 2.2, on page 9 of the Intel 8 Series/C220 Series Chipset Family Platform Controller Hub Thermal Mechanical Specifications and Design Guidelines (TMSDG) document # 328906.
Please let us know if this information is useful to you.