Hi,
Below is the explanation of isolation coupler, I have a question to ask. If arbitrary pulse signal(single or continues, frequency is unknow) input from one side of the isolator, can we detect the signal on the output side? Thanks!
Hi,
Below is the explanation of isolation coupler, I have a question to ask. If arbitrary pulse signal(single or continues, frequency is unknow) input from one side of the isolator, can we detect the signal on the output side? Thanks!
Hi Jerry,
This is the edge-encoded architecture used by many of the early devices in the standard data families like ADuM1400/1/2, ADuM2400/1/2, ADuM1200/1, ADuM3200/1... This architecture sends a pulse across the barrier in response to an edge transition detected at the input.
The diagram is a simplification of the internal functionality of the edge-encoded architecture. Because each pulse will take a finite amount of time to be transmitted and decoded, there is a glitch filter on the input side. This will have the effect of limiting the narrowest pulse, and maximum frequency that can be reproduced from input to output. There are Minimum Pulse Width and Maximum Data Rate specifications in the datasheets.
There is also a refresh pulse and watchdog timer. The function of the refresh pulse is to ensure DC correctness of the output. When a channel's input does not detect an edge transition, it samples the level every ~1us. This feature ensures the correctness of a DC signal in the event of an upset. The watchdog timer resets whenever it sees a pulse sent across the barrier. If no pulses are seen for several us, the output is set to default state.
As long as both sides of the isolator are powered and the arbitrary pulse signal is bounded by the datasheet conditions, we would expect it to be faithfully reproduced on the output side.
Regards,
Jason
Hi Jerry,
This is the edge-encoded architecture used by many of the early devices in the standard data families like ADuM1400/1/2, ADuM2400/1/2, ADuM1200/1, ADuM3200/1... This architecture sends a pulse across the barrier in response to an edge transition detected at the input.
The diagram is a simplification of the internal functionality of the edge-encoded architecture. Because each pulse will take a finite amount of time to be transmitted and decoded, there is a glitch filter on the input side. This will have the effect of limiting the narrowest pulse, and maximum frequency that can be reproduced from input to output. There are Minimum Pulse Width and Maximum Data Rate specifications in the datasheets.
There is also a refresh pulse and watchdog timer. The function of the refresh pulse is to ensure DC correctness of the output. When a channel's input does not detect an edge transition, it samples the level every ~1us. This feature ensures the correctness of a DC signal in the event of an upset. The watchdog timer resets whenever it sees a pulse sent across the barrier. If no pulses are seen for several us, the output is set to default state.
As long as both sides of the isolator are powered and the arbitrary pulse signal is bounded by the datasheet conditions, we would expect it to be faithfully reproduced on the output side.
Regards,
Jason