From: Ash C. <as...@gu...> - 2013-09-17 19:09:36
|
Hi Philipp, Short answer: there is a 10K pull-up to 1.8V internal to the DuoVero COM on the HDQ_SIO line. Longer answer: I agree---this is an error in the design (my error in fact for which I sincerely apologize). I've logged this as a bug internally as something we should wrestle with for the next revision. If you are comfortable with a soldering iron and need to pull these off, I can provide the location of the resistors on the current revision of the COM. I reviewed the schematic for the DuoVero COM this morning and I don't see any supplemental pull-ups on lines that are broken out. --Ash On Tue, Sep 17, 2013 at 1:56 AM, Philipp Lutz <phi...@dl...> wrote: > Hi, > > I'm working with the Parlor/Crystal combo and wanted to change the > default pinmuxing to have some GPIOs available on the 40pin header. For > my application it turned out that the signal HDQ_SIO / GPIO127 would be > convenient to use. > However I always wondered why the signal is driven high as soon as I > changed the pinmuxing and changed the direction to input. When I > activate the build-in pull-down via PADCONF registers, I see a voltage > of around 1V, so it seems there must be one resistor driving the pin to > VCC. In the parlor schematics I didn't find any pull-up, the signal > coming from the gumstix high-density connectors seems to be directly > routed to the 40pin header. > So again this pull-up is supposed to be on the COM board which I really > don't like (to say the least). We already figured that out for I2C2 and > I2C3, see > http://gumstix.8.x6.nabble.com/I2C-pull-ups-on-the-DuoVero-td4967758.html > I don't understand why one would do peripheral related things like that > on a COM board, this renders I2C2, I2C3 and the HDQ interface useless > for other pinmuxing modes, even for custom peripheral boards.... > > @Ash: Are there even more of these build-in pull-ups on the crystal COM > board? > > Cheers > Phil > |