1
Projects, Designs, and Technical Stuff / Re: Drive 36" long pcb traces
« Last post by davethomaspilot on Today at 09:52:08 pm »Found a couple of useful "nuggets".
Typical propagation delay of a signal with FR4 substrate is 6 inches per nanosecond.
The empirical data found in the test is that when the signal delay on the pcb trace is higher than 20% of the rising edge of the signal, the signal will produce significant ringing.
So, one approach would be to add a series resistor at the source to slow the rise and fall times so they are slower than 5x the signal propagation delay. If the resulting tr/tf is too slow, break the long trace into segments driven by separate buffers.
36" is 6 nanoseconds prop delay by the above guideline. So, Tr/Tf would need to be longer than 30 nanoseconds.
That seems reasonable for 2 Mhz--60 nanoseconds for Tr/Tf and (500-60)/2 = 220 nsec for the up and down times of the the clock signal).
So, I can pick resistor value based on what I actually see once I get pcbs back.
Comments?
Typical propagation delay of a signal with FR4 substrate is 6 inches per nanosecond.
The empirical data found in the test is that when the signal delay on the pcb trace is higher than 20% of the rising edge of the signal, the signal will produce significant ringing.
So, one approach would be to add a series resistor at the source to slow the rise and fall times so they are slower than 5x the signal propagation delay. If the resulting tr/tf is too slow, break the long trace into segments driven by separate buffers.
36" is 6 nanoseconds prop delay by the above guideline. So, Tr/Tf would need to be longer than 30 nanoseconds.
That seems reasonable for 2 Mhz--60 nanoseconds for Tr/Tf and (500-60)/2 = 220 nsec for the up and down times of the the clock signal).
So, I can pick resistor value based on what I actually see once I get pcbs back.
Comments?