I am wondering what the input pins on my GFMPW-1 chips need for logic high minimum voltage and logic low maximum voltage. They have had no issues picking up 3.3V signals so far, but I am still worried about instabilities caused by those levels being technically out of spec.

are you biasing your IOs at 5V with those inputs high at 3.3V?

@Tholin I must admit I’ve not read the GF180 IO cell specs (can you share a reference you’re using for thresholds and recommended levels?) but I assume it’s based on TTL levels where the minimum “high” level is (as I recall) defined as 2V, so I guess therefore that 3.3V will be considered a good “high”. I also assume the IO cell has good buffering before it goes into your designs. BTW, I’ve heard that you can potentially run GFMPW at 3.3V (just more slowly)… I might try that on my own

https://mithro-gf180mcu-pdk.readthedocs.io/en/latest/IPs/IO/gf180mcu_fd_io/gf180mcu_fd_io__5_4.html you're right about the TTL stuff, on this page "input high" Vmin is 2V (Vmax of "input low" is 0.8V)

Oh perfect, so I'm good to interface with 3.3V devices. Thank you!

3.3V inputs to the 5V-powered GFMPW chip should be fine, yes. Its outputs are also fixed at 5V though so of course take care to not damage a non-5V-tolerant device. I think we talked about this a year ago but TXS0108E might be good for fast bidirectional level conversion…? I’m using cheap/basic BSS138-based devices (https://www.sparkfun.com/products/12009) for shifting 5V down to 3.3V for now and they seem to do ok at 25MHz… maybe a little sloppy. I have TXS0108E to solder up and try later. Curious about whether anyone else has tried them or any better alternatives