Facebook
Google
GitHub
Linkedin
SPI Connection
- Thread starter lucas03
- Start date
In normal circumstances I would say no. There may be some instances where extremely fast clock pulses and long traces might benefit. Tell us more about the conditions you are worried about.
I found a very logical explanation for our question we can argue about it :
The need for series resistor can be explained with respect to the maximum power transfer theorem. The law states that to transfer maximum amount of power from the source to a load, the load impedance should match the source impedance.
While transmitting data from one device to another using SPI or I2C, the two devices are connected via a transmission line. For maximum power to be transferred, the transmission line impedance should match the source and load impedance. If the impedances aren’t matched, maximum power will not be delivered. This means that the load does not absorb all the power that is transmitted and will reflect back towards the source and the therefore, the signal is lost. To prevent the loss of signal, series terminating resistors are added for impedance matching.
Series resistors are recommended only in the case of very-high-frequency communication (more than 400 kHz) or when the communication line traces are over a large trace distance. A trace is considered as a long trace if the time it takes for a signal to travel down a trace and reflect back is significantly longer than the rise time of the signal.
Series resistors can also be used for protection against high-voltage spikes on communication lines owing to high frequency. Also, the voltage swing on the bus reduces. However, a high value of the series resistor would slow down the rising edges of the signal, which makes them susceptible to noise. Therefore, the right value for the resistors should be chosen to reduce high-voltage spikes in the communication lines without causing it to slow down the signal drastically.
I found a very logical explanation for our question we can argue about it :
The need for series resistor can be explained with respect to the maximum power transfer theorem. The law states that to transfer maximum amount of power from the source to a load, the load impedance should match the source impedance.
While transmitting data from one device to another using SPI or I2C, the two devices are connected via a transmission line. For maximum power to be transferred, the transmission line impedance should match the source and load impedance. If the impedances aren’t matched, maximum power will not be delivered. This means that the load does not absorb all the power that is transmitted and will reflect back towards the source and the therefore, the signal is lost. To prevent the loss of signal, series terminating resistors are added for impedance matching.
Series resistors are recommended only in the case of very-high-frequency communication (more than 400 kHz) or when the communication line traces are over a large trace distance. A trace is considered as a long trace if the time it takes for a signal to travel down a trace and reflect back is significantly longer than the rise time of the signal.
Series resistors can also be used for protection against high-voltage spikes on communication lines owing to high frequency. Also, the voltage swing on the bus reduces. However, a high value of the series resistor would slow down the rising edges of the signal, which makes them susceptible to noise. Therefore, the right value for the resistors should be chosen to reduce high-voltage spikes in the communication lines without causing it to slow down the signal drastically.
Why would you think power transfer has anything to do with communication? You are trying to apply a principle where it has no meaning!
I do agree that for very fast clock edges on long PCB traces, series damping resistors may have some benefit, but as noted they will interact with the load capacitance to produce rounded edges.
As a matter of principle you can do anything you want to do. If you're so damn smart why bother asking us for our opinion?
I do agree that for very fast clock edges on long PCB traces, series damping resistors may have some benefit, but as noted they will interact with the load capacitance to produce rounded edges.
As a matter of principle you can do anything you want to do. If you're so damn smart why bother asking us for our opinion?
Nonsense. Trying to understand proper handling of digital logic signals in terms of power transfer is only going to hinder your understanding. Learn when to use the maximum power transfer theorem, and when not to. Its use in this case is just plain idiotic.
You have the right ideas but in the wrong context and applications.
You have to beware that when you learn something new it does not mean that it is applied universally.
You describe about three or more different phenomena that are to be applied under different circumstances.
SPI communications is not one of them.
You have to beware that when you learn something new it does not mean that it is applied universally.
You describe about three or more different phenomena that are to be applied under different circumstances.
SPI communications is not one of them.
The primary reason for using a series resistor on a digital signal line is to counteract the effect of distributed inductance along the trace. This applies to traces in excess of about 4 or 5 inches. It leads to ringing on very sharp edges in the nanosecond range. This comes directly from a consideration of the RLC circuit familiar to most college sophomores. The resistor controls the damping factor of the 2nd order ODE. It moves the system from being underdamped toward being critically damped. It has nothing to do with transmission lines or impedance discontinuities or reflections.
