Tool/software:
I noticed that the ISO1640 has a fantastic I2C output current of 50mA on Side 2, but bus capacitance has a maximum of 400pF. This is rather interesting because typically I2C (standard mode) devices only require an output current of 3mA. I see an opportunity to use this device to isolate and extend a 100kHz I2C bus over a relatively high capacitance (800pF) line with all other issues (termination, EMI, etc.) currently handled other than capacitance. I understand that this limitation may be caused by just following suit with the I2C bus specifications.
If the I2C output is just through a standard open-drain NPN/NMOS transistor, I don't see why a larger bus capacitance isn't possible. If the limitation is peak power of the switch, 50mA provides plenty of headroom for buses with high capacitance at low frequencies (<=100kHz), especially if a load resistor was put between the device and bus.
If the limitation is sustained power, a 1.7MHz/400pF system would dissipate a lot more power in the transistor, than say 100kHz/800pF.
The "same current" case: If peak current in the transistor was restricted to the same amount, up to 17x more capacitance could be allowed with a 17x longer active time (for the same average current), right?
The "same switch resistance" case: If resistance to the bus (load switch resistance) was the same, peak current would increase proportionally to Cbus and the length of the discharge time would also increase with Cbus, for a total power dissipation increasing proportionally to Cbus^2; this would imply that 100kHz/1640pF would dissipate the same power as 1.7MHz/400pF (assuming an the resistor continues to behave ohmically).
If every other device on the bus can support the additional capacitance, would the ISO1640 work for isolation?
Thanks,
Jack
