On Tue, Jun 30, 2020 at 05:59 AM, Dave Schwartz wrote:
P.S. Although I am pretty explicit with the orientation in the assembly instructions for the STM32 kit, I've still works with several who have installed it backwardsPerhaps it kind of works but not too darn well.
Installed backwards I think it goes like this:
RA- is then tied (through 2k resistors) to both 3v3 and (as a pull-up/down) to DE-,DE+,RA+.
If RA- gets pulled low DE-,DE+,RA+ all go low.
If any of DE-,DE+,RA+ get pulled low everything goes to ~1.6v (as a voltage divider through RA- to 3v3.) If more than one of these are pulled low the voltage goes even lower.
For an SHC if it manages to make it into data mode the pull-up/down are "forced" since all four pins are driven as outputs. The resistors will however add noise to the signal lines since the data in/data out/clocks are all weakly tied together. It's difficult to wrap my mind around what that would look like or if it could possibly work at all; not worth the effort to really analyze it.
Makes me think about Khalid saying that the internal pull-ups on the STM32 ST4 must be on. There is no reason that should be needed... on the other hand it might add some stability to the above situation since it would tend to boost the voltage levels a bit on the signal lines and turn what might be a logic low to the high it should be in marginal cases.