Some questions about conventions

Álvaro Lopes alvieboy at alvie.com
Fri May 22 21:15:10 CEST 2009


Werner Almesberger wrote:
> Yeah, much clearer ! I would omit the "driver" for nOE, just like
> NXP did.
> 
> There's also the small issue that this is now slightly different
> from the typical way how a driver with inverting enable signal is
> drawn, i.e., without the inverter circle on the enable input (NXP)
> or with separate inverter (TI). So at a quick glance, this may
> look like a non-inverting enable.
> 
> Having one circle on the outside and another the inside would be
> too confusing. Labeling it as nOE helps to disambiguate, so your
> solution is consistent and understandable. Or perhaps we should
> break with the "inverter circle always on the box" convention
> here and only have the circle inside.

Actually I did not draw a circle inside :) But removing the driver of the tristate buffer should make things more clearer.

> I'd lean slightly towards the latter, because it seems to be
> closest to how one would expect a driver with inverted enable
> input to be depicted, based on past experience.

That's actually what TI does on its datasheet. Becomes much more understandable.

Updated version in attach.

Álvaro
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: sn74aup1g125.diff
Url: http://lists.openmoko.org/pipermail/gta03/attachments/20090522/c4401ac1/attachment.txt 


More information about the Gta03 mailing list