Limp home mode

[Bob Wooten]

Thanks. I am going to have to print this up & follow along with the print.
from what I recall of the print this makes sense to me.  what I am
ultimately trying to do is to pull the ~RESET line so that I can shut down
the CPU.  with the emulator that I just bought.  I figured that I could tie
into Pin 2 but it appears that this is not working as I guessed it would  by
holding it down it appears that the entire CPU is still "dead" & the
emulator does not like that.  I guessed that this is the source of the
~RESET but all I want to pull down is that line.  my apprehension is that if
I tie right to it that I am going to try to sink a ton of current through
the emulator from the 34992 & let the smoke out.  I am under the dash &
being 6'4", this is not working well for me.  I think that I am going to end
up yanking some of the harness out so that I can @ least put it onto the
Floor of the car to figure out what line I need to pull to assert the
~RESET.  unless someone else has a better idea, or has been down this road.

thanks again Ludis, you are a great help.  sure wish I was not so anal about
making things tidy & stuck the ECM up there in such a hard to reach place.

Bob


-----Original Message-----
From: owner-gmecm at diy-efi.org [mailto:owner-gmecm at diy-efi.org]On Behalf
Of Ludis Langens
Sent: Sunday, March 11, 2001 4:26 PM
To: gmecm at diy-efi.org
Subject: Re: Limp home mode


Bob Wooten wrote:
>
> I am wondering if there is a document that describes the limp home mode?
> specifically what gets shut off in the ECM & what continues to run.  I
have
> been looking @ the print for the 7730 & it appears that some of the
devices
> that the CPU use when running normally get the CS taken away when in limp
> home mode.

The LIMP signal originates in U13.  It gets asserted when the watchdog
isn't fed - that's one of the bits in the SPI interfaced control
register in U13.  In LIMP mode:

The CTS pullup is probably set to 4K ohms. (In other words, U13 stops
sourcing 5V from it's pin 49.)

The output port bits of U10 are set to zero.  I forget if the output
register gets cleared, or if the outputs are simply ANDed with ~LIMP.

The 2000 and 4000 ppm VSS ouputs from U24 stop toggling (because their
clock is shut off).

U10 forces the ECM to magnetic VSS mode.

Two of the quad drivers are shut down.  The third keeps working - to
drive the SES light, a fan relay, and two more general purpose outputs.

U10 forces the fan output on - but it waits for several seconds of
continuous LIMP mode.  This keeps the fan from getting turned on briefly
when the ECM first powers up.

The IAC driver outputs get tri-stated.

The ignition module is switched to bypass mode.  (This also happens when
the ECM doesn't command a spark for some short interval.)

Control of the injector(s) is switched over to U9, the RFD chip.

> on a related note in the 7730, U17 looks to be kind of a watchdog type of
> chip.  mine only has the GM number (16034992) on it (maybe a cross
reference
> guide?).

The 34992 (U17) is the power supply chip.  It appears to have no
watchdog functions.  It is very likely a GM custom.  I haven't found any
industry standard power supply chips with this exact functionality.  I
just checked several GM ECMs - the 34992's were marked with the
manufacturer code of 466, and used inverse video marking.  Both of these
imply that this is a GM produced part.

> what are inputs, what are outputs & what is
> the logic.

The reason that the LIMP signal connects to the 34992 is for CPU
controlled ECM powerdown.  When the ignition key is turned on, the 34992
turns on it's 5V supply output.  The 5V will stay on until the key is
turned off _and_ the CPU causes LIMP to be asserted (by stopping the
feeding of the watchdog.)  GM uses this delayed poweroff to move the IAC.

--
Ludis Langens                               ludis (at) cruzers (dot) com
Mac, Fiero, & engine controller goodies:  http://www.cruzers.com/~ludis/


----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org

----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org