??? 08/09/10 21:54 Read: times |
#177917 - Series resistors will do the trick! Responding to: ???'s previous message |
Raghunathan said:
Any tips on the interfacing aspect and traps to watch out ? I agree with Per: Put a series resistor of 47...470R into each output line and input line, right where the cable leaves the board, best next to the micro. If a flat ribbon cable is used, connect each second line to signal ground at both ends of cable. This in order to minimize the loop area formed by the signal currents and ground return currents and to minimize capacitive cross coupling between the individual lines. The resistors do many tricks: 1. They act as a simple but efficient measure against ESD by softening the edge of ESD currents (in combination with input stray capacitance of micro) and by this allowing the micros's internal protection diodes to shunt ESD. Carbon surface resistors like the CMB0204 or CMB0207 from Vishay, e.g., are especially useful for this purpose. 2. They provide sort of series termination of output lines and by this reducing the risk of reflections and ringing. When chosen higher than characteristic impedance of cable they can provide some very useful low pass filtering. 3. They provide some input low pass filtering (especially if higher resistance values are used) again in combination with the input stray capacitance of micro. 4. They provide current limiting and can substantially reduce the risk of destruction of micro during short circuiting conditions. If common mode noise induced radiation of this flat ribbon cable is an issue, for instance, when a solid ground plane is missing on the PCB or when the cable leaves the board at a place that is not "quiet" (in the sense of ground noise) thread the whole cable through a soft ferrite core and locate this core right where the cable leaves the board. Today there are even snap-on flat cores available especially for this job to ease the handling. Kai Klaas |
Topic | Author | Date |
Pulse degradation | 01/01/70 00:00 | |
For many purposes, slow flanks are not a problem | 01/01/70 00:00 | |
Slew-rate limiting | 01/01/70 00:00 | |
Extending SiLabs TestBoard I/O | 01/01/70 00:00 | |
Solved.. | 01/01/70 00:00 | |
piggy back PCB? | 01/01/70 00:00 | |
maybe not totally related | 01/01/70 00:00 | |
Quite related and valuble | 01/01/70 00:00 | |
Brief Heavy or Not | 01/01/70 00:00 | |
Series resistors will do the trick! | 01/01/70 00:00 | |
Pro vs Amatuers | 01/01/70 00:00 | |
I'd like to rephrase | 01/01/70 00:00 | |
Theory first, then lots of testing | 01/01/70 00:00 | |
Environment... | 01/01/70 00:00 | |
! Amused![]() | 01/01/70 00:00 |