Re: MPC5200 coming out of reset - does it actually work?

Now after another day I can make no progress. This is just
unbelievable.

The one thing I seem to have found out is that when the CPU does
not make it out of reset, the highest 3 address lines may be non-0
as they should be (ad_32, ad_22 and ad_21). Well, they are not
connected to the flash anyway... (I see that via JTAG, don't
know if they float or are driven so, will measure that tomorrow,
I guess).

To ask a more basic question: has anyone seen an MPC5200 of that
kind

MPC5200BV400
L25R
REV 2
QFA0533A

work in that context (come stable and repeatably out of reset)?

Knowing this would be great help, if I knew it could be done
I'd probably manage it somehow. The more I look at the errata ***,
the more "that's how this silicon is" lookes likely (and that "rev 2"
thing... I know rev A is unusable, what if nowadays rev B also is).

Thanks,

Dimiter

On Mar 28, 2:45 am, "Didi" <d...@xxxxxxxxxxx> wrote:
Now I have wasted two days on making the 5200 come out of
reset in a repatable manner - in vain.
I begin to suspect it is a silicon error, judjing by the number
of errata and subsequent revisions etc., it has taken me already
a lot more than typically and it still does not work.
Here is the text on the chip:
MPC5200BV400
L25R
REV 2
QFA0533A

I made the 1.5V come up about half a second after the rest; no change.

I tried to put larger caps on the PLL VCC filters - I have 10 Ohm
resistors
there, tried with up to 1000 (!) uF - no change, 0.1uF or 10 or 1000.

I tried all posible sequence of JTAG TRST and POR; nothing.

I made the POR longer (about 0.5 seconds after 1.5V come up); nothing.

It never comes out of reset when powered on; sometimes, then, after a
few
(rarely just 1) manual retriggering of POR it does come out and then
works
quite stably (part of my monitor still does not work for reasons I
don't know
but the working part is quite stable, I can read/write to memory
etc.).

The PCI clock comes up at 33 MHz regardless of whether the CPU will
come or not. I have set the CPU frequency to 400 MHz, the IPC clock
is still 66 MHz (I will set it twice higher once I move on), I have
checked
countless times the pulups and downs for reset configuration.
I have also put all the pulups on the PCI control signals and on
tsel_0
(pullup) and tsel_1 (pulldown), as well as the suggested pulldowns on
the tmod_0 and 1.
And now I am pretty clueless - it just works whenever it chooses to.
It appears that it takes some time for something to warm up and then
it begins to respond to the manual POR, but this is far from sure
as it does not take long and power cycling alone never works, even
after some warmup.

Hopefully someone knowing the 5200 might have the right guess
what is going on. Hopefully it is not a silicon problem...

Dimiter

------------------------------------------------------
Dimiter Popoff Transgalactic Instruments

http://www.tgi-sci.com
------------------------------------------------------


.