I wonder what the current rating on P2 pins in reverse bias is? 500uA like the P1? 1mA?
The clamping is a reverse P-FET substrate diode, so it will present a hard clamp if attempted to push to 5V.
A typical 5V-CMOS output load line would drive 20-30mA into each clamped pin, and one risk is that reverse current then elevates the 3v3 rail (eg if P2 is reset) and so damages P2.
I wonder if using a DDR regulator on a design that might encounter 5V-Driving logic is worthwhile ?
Parts like RT9045 or RT9173 would keep a 3.3v rail at 3.3v
I used bypassible current limiting resistors on the ASC to keep those clamping diodes from frying on 5 volts. For this board though I think it'd be too much. People will just have to use their smarts and accepting of a modicum of responsibility for connecting over voltage inputs. Just like they do for hundreds of other boards, including 3.3v arduinos.
The advantage of starting from another known good design is speed.
Downside is not all the P2 pins going to headers. Still bummed about that, but at least could bring two of them to LEDs and some of the others to probe points.
The advantage of starting from another known good design is speed.
Downside is not all the P2 pins going to headers. Still bummed about that, but at least could bring two of them to LEDs and some of the others to probe points.
You could add 8 more pins to the bottom of the ICSP header like I did on the ASC. (A added 10, but the bottom 2 went to analog inputs on the adc chip) https://1mgh.com/wiki/Asc#Add-on_cards
Comments
The clamping is a reverse P-FET substrate diode, so it will present a hard clamp if attempted to push to 5V.
A typical 5V-CMOS output load line would drive 20-30mA into each clamped pin, and one risk is that reverse current then elevates the 3v3 rail (eg if P2 is reset) and so damages P2.
I wonder if using a DDR regulator on a design that might encounter 5V-Driving logic is worthwhile ?
Parts like RT9045 or RT9173 would keep a 3.3v rail at 3.3v
Great job!
Downside is not all the P2 pins going to headers. Still bummed about that, but at least could bring two of them to LEDs and some of the others to probe points.
Board looks good
https://www.amazon.com/gp/product/B01EUVJYME/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1
You could add 8 more pins to the bottom of the ICSP header like I did on the ASC. (A added 10, but the bottom 2 went to analog inputs on the adc chip)
https://1mgh.com/wiki/Asc#Add-on_cards
Maybe you're the right person to make this kind of board?
I see a lot of Arduino stuff on your website...