Part Number: PCA9306
Other Parts Discussed in Thread: TCA9517
Thanks,
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
This device uses a pass FET architecture so the VoL some extra offset of whichever device is driving low. The offset voltage will increase as current increases.
The device does not redrive the signal.
Example: If we assume Ron is 60 ohms and you are pulling 3mA then the extra offset voltage would be about 180mV. So if the VoL of which ever device is driving low is 0.3V then the VoL seen on the opposite side of the PCA9306 will be 0.4V+0.18V = 0.48V
-Bobby
Thanks for your reply,
if I2C master driving OV on I2C lanes, other side of shifter can we see only 0.18V?
The level shifters for I2C (open drain) are NTS0104 and they have maximum V-OL of 0.4V. This seems to be preventing the I2C master from driving the I2C lines to 0V. It still works, but it reduces the design margin for 1.8V logic I2C
"if I2C master driving OV on I2C lanes, other side of shifter can we see only 0.18V?"
The 'offset' voltage of our device depends on the current draw which is why I gave an example calculation. If the current draw (IoL) through our device's channel was 1 mA and the offset voltage assuming 60 ohms (Ron) would be 60mV. The Ron however is also dependent on current draw and biasing voltage (I would expect it to actually be lower than 60 ohms).
"The level shifters for I2C (open drain) are NTS0104 and they have maximum V-OL of 0.4V."
Not exactly.... The NTS device states if the input is 0.15V or less AND the IoL is 1mA then VoL is 0.4V max. If you were pulling low to 0.1V and at an IoL of 500uA for example, then I would expect VoL to be around 0.225V max. I believe their device uses a pass FET architecture and so the FET they are using would be in the linear range of operation in this case. It is similar to our PCA9306 device but has one shots (our PCA9406 would be the 1 channel version of the NTS0104)
"This seems to be preventing the I2C master from driving the I2C lines to 0V. It still works, but it reduces the design margin for 1.8V logic I2C"
If you want to design to have more margin, we may want to design in an I2C redriver device like TCA980x or TCA9517.
Thanks,
-Bobby