With and without MPU block diagrams
andy at openmoko.com
Thu Apr 24 18:59:33 CEST 2008
-----BEGIN PGP SIGNED MESSAGE-----
Somebody in the thread at some point said:
| Andy Green ??:
|> -----BEGIN PGP SIGNED MESSAGE-----
|> Hash: SHA1
|> Hi folks -
|> This PDF looks at the two ways we can go with or without an MPU.
|> I captured everything I recalled us discussing, if a good point one way
|> or the other is missing or wrong feel free to note it and I will amend
|> it. Hopefully tomorrow we will define if we use it or not, and if so,
|> which version.
| Very cool, thanks, and few questions.
| 1. the button should include power button and volume up/down switch,
| should power direct connect to PMU and volume connect to MSP430?
Yes, added, thanks.
| 2. Camera is missing, should direct connect to CPU itself :)
OK :-) I made a v3.
| 3. Any suggestion on FTDI chip's GPIO, I remember there are some
| available could be used.
The main one was reset, but we have that in there already. Thinking
about what I do during development, it's just really reset that I use.
Maybe someone else can have an idea.
If it helps do automated regression testing, we could wire them to
buttons via a series diode, but if we have an MSP430 that is another job
for that really. Dunno if it makes any sense.
Or if there is any step in production test, but I didn't think of
anything that can't be done another way.
| Allen has question about jtag chain, because we never use this before,
| so his first design might not chained together.
Wiring up Jtag chaining isn't very tricky. Each device sees the same
TCK, TMS and if you have it TRST from the TAP controller (tester). The
first device in the chain gets the TDI drive signal from the TAP
controller. Then you pass the TDO from the first device into the TDI of
the second device and so on. The TDO from the last device in the chain
goes back to the TDO input of the TAP controller.
The two issues we have here are 1) we only have one TAP controller
defined in the FTDI :-) so there's trouble having two chains, and a
secondary issue is that I wonder if openOCD is well tested and bug free
if the CPU isn't the first or maybe last device in the chain.
If he doesn't feel comfortable wiring them together, if they come out to
0R pads or a connector we can confirm correct operation chained together
before making the next board.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
-----END PGP SIGNATURE-----
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 41232 bytes
Desc: not available
Url : http://lists.openmoko.org/pipermail/openmoko-kernel/attachments/20080424/4243d764/attachment.pdf
More information about the openmoko-kernel