Hello
My customer intends to use I2C multiplexer with 1:5 .
I will promote one LTC4312 and one MAX7367 for this configulation, but would you please let me know your opinion?
I hope your advice soon .
Best Regards
LTC4312
Production
The LTC4312 is a hot-swappable 2-channel 2-wire bus multiplexer that allows one upstream bus to connect to any combination of downstream busses or channels...
Datasheet
LTC4312 on Analog.com
MAX7367
Production
The MAX7367/MAX7368/MAX7369 bidirectional, four-channel I²C switches/multiplexer expand the main I²C bus up to four extended buses. The MAX7369 1:4 multiplexer...
Datasheet
MAX7367 on Analog.com
Hello
My customer intends to use I2C multiplexer with 1:5 .
I will promote one LTC4312 and one MAX7367 for this configulation, but would you please let me know your opinion?
I hope your advice soon .
Best Regards
Hello Shiba183,
Using the LTC4312 + MAX7367 to realize a 1:5 I2C multiplexer will certainly work and I'd suggest using the LTC4312 as the upstream mux with the MAX7367 as a downstream mux. This will give you the benefit of the LTC4312's bus buffers to electrically partition the mux cascade from the rest of the upstream bus. This will also allow the LTC4312's rise time accelerators to help support the rising edges through the downstream mux's series R. Depending on the application's bus topology, this can be beneficial.
One potential drawback with this combination is that these parts have different control methods - the LTC4312 is controlled by two enable pins while the MAX7367 is controlled through its I2C registers and so the control method for the composite mux will need to take this into account. This means that the control device needs to drive 1 or 2 GPIO in addition to having the I2C control functions. Assuming the MAX7367 is downstream from the LTC4312, then in order to access its control registers the LTC4312 first needs to be set to select the MAX7367 and then the controller can set the MAX7367 appropriately. This requires additional pins, time and incrementally more complicated code.
An alternative might be to use the LTC4305 which is a 2-channel mux similar to the LTC4312 but is controlled via I2C registers similar to the MAX7367. This would remove the requirement for the 1 or 2 GPIO pins and make the entire control scheme be similar; however, there will still be some delay and sequencing details involved in changing downstream channels.
Another alternative would be to consider simply using a 1:8 multiplexer and just not using the unneeded 3 extra channels. The MAX7356, MAX7357 or MAX7358 could be candidates for this option. They don't included the bus buffer or RTA's of the LTC4312/LTC4305, but would provide a single part solution for just a bit more than the MAX7367, This would simplify the control to a single interface and remove the control sequencing complexity of the cascaded mux. They could still add a standalone bus buffer or RTA on the mux input to regain those benefits if desired.
Eric
Dear Mr Eric
Thank you for your answer soon! I will introduce MAX7357 all of whose register can be accessed at power up.
Best Regards