Symmetry is essential for unshielded Ethernet. The cable must meet this symmetry requirement in its construction. The impedance must also correspond to the nominal value to prevent reflections of the signal.
Thank you Anil for the question! Sorry for the late reply. This is indeed something interesting. Let us look into the topic and how we can answer this.
Thank you Dr. Lars Völker. I have a question regarding Automotive Physical layer. In conventional Ethernet PCS uses block encoding 4b/5b and scrambling to ensure while transmitting bits at physical layer by MDI for asynchronous and NRZ protocol there are bit toggles for synchronisation. While for Automotive Ethernet PCS just does clock conversion 4b/3b and coding 3B/2T which if input bits are 0 continuously then at voltage will be represented by -1 since both the ternary symbols are -1 or for input bits 1 continuously then +1 since both ternary symbols are +1. How does a NRZ asynchronous protocol like Automotive Ethernet handles it?
Hello Satheesh Kini! Thank you for your question! I am not sure I understand it correctly but it seems to suggest that 100BASE-T1 does not use a Data Scrambler? In my humble opinion this is not the case. If you take a look at IEEE 802.3-2018 figure 96-8 you can see that a Data Scrambler is part of the 100BASE-T1 PCS right after the 4B/3B conversion and before the 3B/2T. For a good description you might want to also consider "Matheus; Königseder: Automotive Ethernet 3rd Edition", page 159 and following. Best regards, Dr. Lars Völker
@@technica-engineering Thank you for the reply. I was asking more from a point of how synchronisation is achieved in 100 Base T1 if data is all 0s or 1s are transmitted since it's NRZ protocol. I understand 4B/5B encoding ensures there are no consecutive bits continuously while 3B/2T as it seems not the case for consecutive 0s and 1s.
Hey @@KinzRulz, Yes, using 3B2T does not guarantee that the clock is constantly available. This is the same as for 1000BASE-T on which 100BASE-T1 was based and in contrast to 100BASE-TX. For 1000BASE-T and 100BASE-T1, the master generates the clock. In the first phase of the link training the slave synchronize to this clock. In addition, the master determines the phase of the slaves signal in the second phase of link training. Later during transmission, the much long scrambler sequence compared to 100BASE-TX helps that signals basically change often enough. And of course the start and end of the data is marked with specially 00 symbols. Best regards, Dr. Lars Völker
@@KinzRulz You can have a large run of identical symbols in a row. However on average the DC content is zero, so over long term there are transitions. This is exactly why the scrambler is in place so that you have transitions available. The actual RX PMA implemenation is vendor specific. Essentially there is a CDR that tracks the datastream and recoveres a clock based on the data input. For that you dont rely on an edge per symbol, just on a reasonable numer of edges on average. Another effect caused by a run of symbol is BLW which essentially shifts the slicer point based on the history. So the receiver needs to keep track of what has been received in order to know where the slicer level is. If you google for baseline wander you will find a bit of background.
MOST was able to handle most of these use cases in the past. While for premium Audio transport MOST was really well suited, for other use cases like ADAS and AD, MOST was never a preferred solution. Automotive Ethernet in contrast can handle all different use cases in vehicles. More than 10 years ago, OEMs started to reduce the usage of MOST, while they were adding Automotive Ethernet. For many well-known brands and OEMs, Automotive Ethernet has already successfully replaced MOST as well as other technologies.
![Megan Thee Stallion - Neva Play (feat. RM) [Official Video]](http://i.ytimg.com/vi/TpYTyAaTRts/mqdefault.jpg)
Can you explain SOME/IP
Great - Would like to know any Difference between CAN & Ethernet Cables used - Like shielding , Material Difference / Number of Cables etc.. Thank you
Symmetry is essential for unshielded Ethernet. The cable must meet this symmetry requirement in its construction. The impedance must also correspond to the nominal value to prevent reflections of the signal.
Great explanation. Some humble feedback, the background music is disturbing and its volume creates a distraction.
Thank you for your feedback. We are working to improve our Videos.
Thanks its very useful, is it possible to have a detailed session on AUTOSAR ethernet stack , atleast overview about each module that are involved
Thank you Anil for the question! Sorry for the late reply. This is indeed something interesting. Let us look into the topic and how we can answer this.
Thank you Dr. Lars Völker.
I have a question regarding Automotive Physical layer.
In conventional Ethernet PCS uses block encoding 4b/5b and scrambling to ensure while transmitting bits at physical layer by MDI for asynchronous and NRZ protocol there are bit toggles for synchronisation.
While for Automotive Ethernet PCS just does clock conversion 4b/3b and coding 3B/2T which if input bits are 0 continuously then at voltage will be represented by -1 since both the ternary symbols are -1 or for input bits 1 continuously then +1 since both ternary symbols are +1.
How does a NRZ asynchronous protocol like Automotive Ethernet handles it?
Hello Satheesh Kini!
Thank you for your question! I am not sure I understand it correctly but it seems to suggest that 100BASE-T1 does not use a Data Scrambler? In my humble opinion this is not the case. If you take a look at IEEE 802.3-2018 figure 96-8 you can see that a Data Scrambler is part of the 100BASE-T1 PCS right after the 4B/3B conversion and before the 3B/2T. For a good description you might want to also consider "Matheus; Königseder: Automotive Ethernet 3rd Edition", page 159 and following.
Best regards,
Dr. Lars Völker
@@technica-engineering Thank you for the reply. I was asking more from a point of how synchronisation is achieved in 100 Base T1 if data is all 0s or 1s are transmitted since it's NRZ protocol.
I understand 4B/5B encoding ensures there are no consecutive bits continuously while 3B/2T as it seems not the case for consecutive 0s and 1s.
Hey @@KinzRulz,
Yes, using 3B2T does not guarantee that the clock is constantly available. This is the same as for 1000BASE-T on which 100BASE-T1 was based and in contrast to 100BASE-TX.
For 1000BASE-T and 100BASE-T1, the master generates the clock. In the first phase of the link training the slave synchronize to this clock. In addition, the master determines the phase of the slaves signal in the second phase of link training.
Later during transmission, the much long scrambler sequence compared to 100BASE-TX helps that signals basically change often enough. And of course the start and end of the data is marked with specially 00 symbols.
Best regards,
Dr. Lars Völker
@@KinzRulz You can have a large run of identical symbols in a row. However on average the DC content is zero, so over long term there are transitions. This is exactly why the scrambler is in place so that you have transitions available. The actual RX PMA implemenation is vendor specific. Essentially there is a CDR that tracks the datastream and recoveres a clock based on the data input. For that you dont rely on an edge per symbol, just on a reasonable numer of edges on average.
Another effect caused by a run of symbol is BLW which essentially shifts the slicer point based on the history. So the receiver needs to keep track of what has been received in order to know where the slicer level is. If you google for baseline wander you will find a bit of background.
@@paxer Thank you. I will search on mentioned topics.
Hello, thanks for the video. I thought for all that usecases is used MOST Bus. Can you explain why not?
MOST was able to handle most of these use cases in the past. While for premium Audio transport MOST was really well suited, for other use cases like ADAS and AD, MOST was never a preferred solution. Automotive Ethernet in contrast can handle all different use cases in vehicles.
More than 10 years ago, OEMs started to reduce the usage of MOST, while they were adding Automotive Ethernet. For many well-known brands and OEMs, Automotive Ethernet has already successfully replaced MOST as well as other technologies.
What about OPC UA?
To the best of our knowledge, OPC UA is not used in cars but has been considered at for other types of vehicles
I wanted to sell ur product