We are using this chip for voltage translate between 3.3V and 1.8V. The 1.8V side is not going to ground. What would be the likely cause of this?
I thought I might have had pullups in line but it turns out they are not populated. Here is the schematic as the pull up is happening on pin 15.
Sorry here is the schematic of the part being used. Pins 15 and 13 eventually go into a processor.
Hey Alan,
Thanks for the schematic. There are a couple items that I have questions/comments on:
First one comment - don't leave CMOS inputs floating. For your system, I'd recommend to pull-down 1A2 and 2B2 -- assuming you never change directions for 1DIR and 2DIR.
My main question is: what's connected at UART_FTDI_TX? From the scope shot, you have bus contention on that line (I am also assuming this is the line you showed previously -- it's the only one that makes sense).
My guess, from the schematic and your previous comments is that the UART driver is forcing the line high, and the AVC device is only able to pull it to about 1V.
Given that the output of the AVC device's output at 1.8V has a resistance around 30 ohms, we can easily calculate the required resistance to pull the line down to 1V using the voltage divider equation: Vout = Vin*R1/(R1+R2):
1V = 1.8V * 30/(30+R2) --> R2 = 24 ohms
While this is an estimate, it's pretty clear that there's something actively driving that line high -- it's not a dead short, nor is it a pull-up resistor.
Could that cause the problem with 1A2 and 2B2 left open?
The I2C side of it works fine and the UART side works fine when the S3 is off and S4 is set to UART which doesn't use a level shifter. The pullup happens when S4 is off and S3 is set to UART which is using the level shifter.
