-->
Page 4 of 5

Re: Rx input problems

PostPosted: Sat Aug 15, 2015 5:21 am
by MikeBolton
My simple scheme with a resistor and 'zener' worked fine. The zener was actually a white LED and a 1N4148 in series. Gave a clamp at 3.35V. I wasn't worried about the current as this was supplied by the HIN232. As this is a temporary adapter, I used what I had. No point in buying a MOSFET for this purpose, even if it is 'better'. The use of a white LED enabled me to see if there was traffic on the ESP input. It flashes with data. No speed issue at 115200B.

Still no specification on the ESP input current though. Will make the design of any level shifter difficult.

Mike B

Re: Rx input problems

PostPosted: Sat Aug 15, 2015 8:27 am
by eriksl
Why don't you simply measure it? From a basic no-nonsense setup, just pust a current measurement tool in series and keep sending 1's first and then 0's (or v.v.). I can't imagine the impedance is any lower than 100k, the input is probably nothing more than the two gates of a cmos pair + some protection diodes.

Re: Rx input problems

PostPosted: Sat Aug 15, 2015 9:19 am
by martinayotte
If you were using 2 MOSFET circuits mentioned above tied together (only 1 pullup needed on the 3V side), you will end up with an AND configuration, so no needs for inverter ...

Re: Rx input problems

PostPosted: Sat Aug 15, 2015 2:38 pm
by MikeBolton
eriksl wrote:Why don't you simply measure it? From a basic no-nonsense setup, just pust a current measurement tool in series and keep sending 1's first and then 0's (or v.v.). I can't imagine the impedance is any lower than 100k, the input is probably nothing more than the two gates of a cmos pair + some protection diodes.


As I said before, this is OK if you know the logic low value for the ESP and can be bothered to do it. This sort of information should be readily available from the manufacturer.

Mike B