Re: SHC CONNECTION WARNING WITH WeMos R32 with CNC V3 Shield #wemos #esp32


Howard Dutton
 

On Sun, Sep 12, 2021 at 02:32 AM, Ken Hunter wrote:

First let me say this looks like a nice design, I like that you "thinned it out".

If it is not removed, the resistor will "Pullup" the signal lines to 5V. Some think this is dangerous
to the ESP32 MCU. It might improve the connection and... I have used it both ways during the development of my SHC.
I think you're saying the ST4 interface signal lines running to OnStep are pulled up to 5V?  If so, I really don't see the point of this or why it should improve the connection in any meaningful way.  All four signal lines are driven as outputs by the two MCU's so the voltage levels stay essentially the same 0V LOW and 3.3V HIGH.  Additionally you will be pulling some ST4 lines above 3.3V on the OnStep side during the initial connection phase, which for some MCU's is definitely NOT ok.

Decrease the discharge time constant by changing the 100Mfd capacitor value to a lesser value.
     I find the capacitor can be decreased to as low as 2.2Mfd and still stabilize the 5V line sufficently
     while providing a much quicker recovery time. (Measured in seconds instead of minutes).
I don't understand why the ESP32 goes low power, nor did I research this (nor did I design an ESP32 based SHC for that matter.)  But 100uF isn't that large of a value and the I figured the MCU would drain that in an instant.  And then there's the keypad pullups, display, etc. too.  But yea, if this isn't draining down the cap fast enough when unplugged add the resistor to do it.  I just felt given the flimsy RJ connections and possible voltage drop on the thin gauge cord it'd be best to have some useful capacity near the MCU to help cover momentary power requirement fluctuations and help clean up induced noise.

Join main@onstep.groups.io to automatically receive all group messages.