Simple PCIe FMC carrier SPEC issueshttps://ohwr.org/project/spec/issues2019-02-12T09:17:58Zhttps://ohwr.org/project/spec/issues/34V1.1 - 3V3 on FMC mezzanine too low2019-02-12T09:17:58ZErik van der BijV1.1 - 3V3 on FMC mezzanine too lowMeasurements show that there is too much resistivity in the 3V3 path to
the FMC mezzanine. See attached file where one can see that the voltage
is lost over the fuse, inductor and the FMC connector.
E.g.. at 2.29 Amp, the mezzanine gets only 2.78V. At 1.35A it gets 3.02V
and at 0.7A it gets 3.194V (starting from 3.35V on the input).
V1.1 of the board (not in EDMS) had an inductor and a resettable fuse in
it's path to the FMC connector.
Solution adopted for V2 of the board
- removed inductor
- replaced resettable fuse by a fuse that can really blow
- widened the track size of the 3V3 to the FMC connector
### Files
* [3752_001.pdf](/uploads/ca179c98c9e20490d85a78471265d906/3752_001.pdf)https://ohwr.org/project/spec/issues/21V4 - D32227_A1x incompatible motherboard2019-02-12T09:17:54ZFederico VagaV4 - D32227_A1x incompatible motherboardOn the FUJI D32227\_A1x motherboard the SPEC doesn't work most of the
time.
Slot x1 and x4 doesn't work at all: it freeze the system. Depending on
the driver:
- it freeze on FPGA configuration (fine-delay)
- it freeze after few minutes of usage (fine-delay with timer, BE-RF
obs-box \[no mezzanine\])
- receive only one interrupts, after a while (~30s) it freeze (BE-RF
obs-box \[no mezzanine\])
On slot x16 it works most of the time, but it's not 100% reliable
It's impossible to get error messages from the kernel because the system
is completely frozen. It happened only a single time that it prints
messages about: hardware error, Process context corrupter, CPU
synchronization failure
### Files
* [P1020296.JPG](/uploads/2e4bbff7d520ae656e918a1ed1fe458f/P1020296.JPG)Erik van der BijErik van der Bij