We are facing a problem during development stage,
every time when we have some engineering samples ready.
A few of them just can't boot up.
We are using the boot from fuse method for the bootup,
and we burn the fuse such that iMX53 is booting up from eMMC.
Most of our boards work OK, but there are around 3% of the boards that won't be able to bootup.
Or it will bootup one time after many reset.
We tried MFGTOOL, and we ran DDR stress test and able to get up to 470MHz with 4 x DDR3 configuration
or as high as 550MHz with 2 x DDR3 configuration
so we think eMMC and DDR3 doesn't have any problem.
When we tried and check the signals,
we are able to see the eMMC clock change from 3xxkHz to 40MHz.
Then it will stop.
And there is no activity on the DDR3 signals
After much testing and debugging, we found out if we increase the damping resistor of the eMMC to 36 or 47 or even up to 100ohm, and with a 10pF capacitor connected to ground.
Most of these boards is able to bootup again.
But adding such large damping resistor will cause the clock waveform to become like a sine wave, and we couldn't meet the eMMC rise time and fall time spec.
I found an errata ENGcm12290, which says there is a timing issue in SD SDR mode of eSDHC3 port.
We are using this port to connect eMMC too.
Although it says eMMC should not be effected,
but is there a problem with the setup time of eMMC for this port too?
Because if we increase the value of the damping resistor,
it basically make the clock rise time a little bit later,
which may overcome the setup time problem.
We are ready to do mass production very very soon.
Within weeks of time.
We need help immediately on this.
Anyway has similar problem and has any good fix?
Hi Donny,
I am facing nearly the same problem as you desrcibe, but with i.MX6Q and SD4.
When the issue happens, i.MX6 enters serial downloader (USB).
It would be great, if you can investigate a bit in how you fixed the problem.
Many thanks in advance.
Klaus
I cannot remember exactly,
but I remembered that we had added a R-C circuit to the eMMC clock.
We are using 39R and 10pF and the problem doesn't seem to occur anymore.
The reason why it falls back to serial downloader mode is because the CPU has problem finding the eMMC,
so it just fall back to serial download mode automatically.
We still don't understand why this problem happens, but seems to be able to overcome it if we add some RC to the eMMC clock.
Hi Donny,
thanks for your answer. The problem is, that the issue appears only in a small
temperatur window between 37°C and 50°C. But I will test your proposal with the RC
of course.
Br Klaus
Donny
We have not got your response yet and will close the diacussion in 3 days. If you still need help, please feel free to reply with an update to this discussion.
Thanks,
Yixing
Hi Yixing Kong,
We had this problem fixed already.
But it has been a while, and I can't remember how we fixed it.
Anyway, thanks for looking into this problem.
Donny
Donny
We are sorry for getting back to you so late. Are you still stuck with the issue? If you have somehow to resolved the issue, can we close the discussion? If you still need Freescale help, please feel free to contact us.
Thanks,
Yixing