Understanding Crosstalk in PCB Layout - You may wish you knew this before (with Eric Bogatin)
Вставка
- Опубліковано 19 тра 2024
- The best crosstalk explanation I have ever seen. What do you think? Thank you Eric Bogatin. PS: Please share, not because I would like to get views, but I believe it can help many engineers. Thank you.
Links:
- Eric Bogatin: / eric-bogatin-368860
- Eric's courses: www.bethesignal.com/bogatin/
------------------------------------------------------
Would you like to support me? It's simple:
- Sign up for my Hardware design and PCB Layout online courses: academy.fedevel.com/
- You can also support me through Patreon: / robertferanec
- Or sign up for my Udemy course: www.udemy.com/learn-to-design...
It is much appreciated. Thank you,
- Robert
Thank you Eric for a great call and thanks to Martin for creating the subtitles. If you like, leave your feedback for Eric and Martin here, so they know you found it useful. Thank you for watching. Enjoy!
Thank you Robert! This was great.
I'm a soldier of Eric's army
I love how you pause the experts talk for a simpler explanation and then we go back right in to core content :-)
Thank you Amogh :D
Your questions aren't dumb; they help us all to understand what's really going on. Great presentation!
Thank you smeggyhead1
Your videos should be as long as they need to be,
so if your have important material/information on the topic, I don't mind long videos at all.
this is how someone can learn.
Thank you very much bm830810
So true
100% True..
This is one of the most amazing videos ever, on anything.
Both of you are patient with the details. Thanks for the video.
Thanks for this super informative video.
All the contents by Eric Borgatin through roberts channel is priceless...really I am new to this field and am enjoying it as every concept is being visualised in detail and application of undergraduate field theory👍👍....
Thank you very much Siddharth
Very good content and clear explanation. Thanks a lot
Thank you very much Robert!!!
Amazing video!! Just as amazing as the previous one!!
Eric is a great engineer and teacher!! We greatly appreciate his contribution and yours as well of course.
You channel is fantastic thank you
Such a GEM
Thank you very much Robert with such an excellent content in continuation.. Both of your contributions are priceless to the community. Please come with more simulation oriented videos like this with the legend Eric Bogatin.
Thank you very much Sandeep
Hi Robert,
I have read few white papers by Eric but i could not visualize and understand it. But these videos make it super cool and clear.
Thank you very much Robert and Eric for making me feel like awe!
Thanks Robert for bringing this video series about Crosstalk with Eric Bogatin! Really enjoyed watching these videos.
Thank you very much raza
Really great video Robert, thank you very much!
Thank you very much Liad
Thanks Eric for your clean explanantion and thanks robert for your efforts. Your videos are very helpful for building my career....
Thank you soo much for this amazing explanation. Your videos are very useful.
Thank you very much Gireesh
I’m only halfway through watching this video, and understanding crosstalk, reflections, along with their nature and relationship is a university level science doctorate all on its own… kudos to Robert, Erik, and Rick for venturing to provide this information to everyone! Robert, when you were embarrassed at the beginning of the video over a signal characteristic, don’t be. You should keep that type of information in always… we are human. Understanding this information is different than retaining it. I will likely watch this video 10 times just to properly understand it. Retaining the information for the purposes of designing a new circuit is a whole other can of worms! You are doing a great service to those like yourself/myself with a desire to learn Robert!
I am getting many new things to learn through your video. Thanks a lot
Thank you very much Tapan
Awesome. I just realized that i knew nothing. Watching your videos with Mr. Eric and learnt alot. Thank you
Very good content and clear explanation. Thanks from Brazil
Thank you very much Pedro
thank you for this video again. what a nice follow-up of previous one :-)
I love your videos Robert!!! I have told 20+ people to your channel and each person wished they found you earlier!!!
Keep it up!!
Thank you so much tebbo
Amazing! It reminded me all my Physics classes from high school where we were learning about these so basic things about waves, like propagation, reflections. And here is the beautiful presentation of these basics laws of Physics. Robert, Eric - thank you so much!
Thank you Jacek PS: I have similar memories from University. I had no idea why we were learning that and how it can be useful :D
Robert, Eric is your best guest. Invite him more often, please) And of course, you are great too)
Thank you Ялень
Thank you so much for the video, great contribution for engineering all over the world. On the minute 21:48 Eric comments about a rule number 9, I wondering what are the others rules?
Great video... I was learning about the energy flow in a PCB. This helps a lot to someone who r trying to know exactly what is happening in PCB. You should continue this.
Thank you very much Sajin
thank you very much . it was perfect and prfect and perfect
I don't have much of a point in this rant but here it is:
This video will also dispell the weird Veritasium video challenging wether voltage/current/energy/electrcity travels in a wire. Is it electrons propagating in wire or transmission at light speed in all directions. To demo this, he proposed laying a wire 1km long flatly and travelling away 499.5m , the making a 90 degree turn, 1m wire more, then another 90 deg. turn heading back 499.5m keeping a constant 1m distance from outgoing wire.
At starting point of 1km wire was a battery and a 1 pol switch hooked up. I don't remember if the negative side of battery was connected to another 1km wire doing a loop in reverse direction or just a 1m stump to an light source. The challenge was to predict how long time it would take before it would light up with the light source being 1m away from battery/switch. The light source had the property of lighting up at any current present.
It feels like Veritasium makes the case of "spooky electricity", it's not about electrons pushing each other in a wire but "energy" thru air lights up the target at the speed of light travelling directly, jumping the 1m gap.
It is sort of a silly case, because it's basically seen to us as "just crosstalk"... or seen as an antenna transmission.
The only "challenge" in Veritasium video was to "correctly" focus (guess) on the "magic" light source lighting up on the minute crosstalk energy and ignore all other minute details.
To me (the Veritasium vid) it felt like a cheesy video, trying to mystify rather than clearing up different conceptions of reality/theory/fantasy because;
If you are going to make or break the function based on "crosstalk", "magic light source" and asking for a specific time it takes; then anyone sufficiently petty will say question was actually impossible to answer at all because of missing information/test conditions like gravity/time dilation, dielectric properties, thickness of wire and the actual distance between the long runs -is the "1m apart" the c-c wire distance or between the closest wire strands... Oh, if it's an actual wire that means strands, how many? How thick each strand? what twist rate? etc.
I'm not sure why I thought about veritasium video seeing this video but I like this video better.
This is super useful! Thx for sharing! If there’s some chunkable clips, that would be better
Thank you Yuchen
I learned a lot from your videos on cross-talk, Like you guys said , We all are learning from mistakes
Really, Just saying Thanks will not sufficient for your videos , sharing your knowledge with all is such a great thing
I always highly recommend this channel to my friends
Thank you very much Haribabu
Great 👌👍👍
I really like your videos with Eric, specifically because it helps me understand deeper books related to SI/PI, it wrote.
I'm curious regarding hyperlynx, is it free (yes I'm a dreamer).
Keep teaching us guys, I'm feeling clever and clever thanks to you
Thank you very much adjada PS: Hyperlynx is not free. They do not even want me to borrow a license to make videos :(
These videos should be watched day and night!!!
Precious, precious videos for the engineer community.
Now can we move forward into RF areas? XD.
Thank you Zhitai. I am just not sure why the video has only so few views. I believe, this is one of the best videos I created :( PS: I am working on RF, if everything goes well, we should have an RF video in March.
@@RobertFeranec This is really a good question. I suppose that there are many people who still endorse the "it-is-working" principle. Once things get working, they don't really care about how they route the boards. It is the moment when they fail some EMC tests in the lab or simply screw up high-speed communications (even before the EMC test), they will start to think about these SI issues. As things get faster and faster, simply less and less engineers capable of. Not mention those high-end equipment (did someone notice that Eric has a intel i9 CPU with 16 cores running 80% resource!) and expensive software. This is why I appreciate these videos so much. Just like another great channel "The Signal Path" by Shahriar, simply too great to be appreciated. The target audients are by no means beginners. Viewers != Quality.
Hi Robert, when designing a flexible PCB, it is important in some cases to add a stiffener (e.g. FR4) beneath some components to avoid overbending the pads and components. Have you done any videos on how to add FR4 stiffeners to specific locations? Thank you
Awesome. I would like to see the effects of higher output impedance. IE putting a higher value resistor on the output pin of an IC to slow the edge and swamp the trace . Like a trace going from Cmos IC to Cmos IC.
Thank you Richard. PS: I have these videos on my list
At 12:20, Eric adds a 50 Ohm terminating resistor connected to ground. What if you are just trying to prevent crosstalk on the output of a microcontroller? That would kill the microcontroller pin.
It looks like a PCB field solver is becoming a vital part of professional PCB design software. It doesn't seem possible to properly take into account all of these crosstalk, impedance, etc. effects without a simulator. It probably costs a lot, restricting the possibilities to wealthy designers / big companies.
I wonder what is the way to protect the paths of equipment like photodiode?
This video was super informative and very interesting. I had one question regarding the cross talk on the victim trace. I see on the simulation and even on the actual measurements, the cross talk voltage in the victim trace is generated like a few time scales before actually the signal rises in the aggressor. How does this happen? How does the victim trace know even before the signal has arrived.
What is shown is the far end signal of the aggressor. This is not the input signal. The input signal view is disabled in the simulation He shows the trigger in the beginning and disables the view. 4:19 is where the input signal is shown. He tells us to remember the rise time there.
In the live demo he 1st shows the input signal connected directly from the oscilloscope to the oscilloscope and tells us to remember the rise time. 52:33 is the input signal. After which he connects it to the board. Therefore he cannot show the input signal directly.
The cross talk happen at the right time after the input is triggered.
Omg we waste 4 years in class I love this so interesting
Thank you very much Ashwinder PS: I wish I had professor like Eric at University
@@RobertFeranec I wish I had professor like you as well you have changed my life I am following you since 2017
Thanks but I think that simulation tools are still very expensive. A test board or simulation using traces with bends and compensating jog outs would be interesting. Does Eric have a similar test board to test differential pairs - maybe with different lengths to simulate a phase error - how does this affect signal integrity?
Thank you Chromatec. I have this on my list. I may use Cadence instead as Mentor/Siemens doesn't want to provide license for Hyperlynx (yes, simulation software is expensive, so I am very grateful when a company borrows me the license)
Hi Robert Feranec, how near end crosstalk gets high even before the rising edge of the driver signal do it?
the rising edge of the driver which we see in the in simulation is measured at the end of track, in reality the signal starts sooner (at time 0) ... it takes some time to arrive the signal to the point where we measure it in our simulation.
@@RobertFeranec you are rigth. I got it. I let that slip away. Thanks.
To me that looks a lot like a directional coupler design, it even behaves like one
Eric says.... if u understand the fundamental principles... But, duz Eric understand the fundamental principles of what is electricity along a wire? (Nope)(nobody here duz).
I notice that in this youtube & in every other similar youtube that i have watched recently there is no mention of electrons (probly a good thing).
But you didn't compare trace spacing! :'(
It is on my list of future videos.
I don't know how to put it nicely, but here's my problem: it _appears_ as though you can't tell what to expect at the far end of a victim line in a cross-talk scenario. Yet you sell PCB design courses online, some of which claim to teach buyers how to design high-speed boards. In another one of your other videos you try to simulate a split in a plane then ask viewers to help you make sense of the results. I don't get it. Deep down I hope that this is just a pedagogical vehicle, an act you put on to make all your viewers feel comfortable, involved and at ease. If this is the case, you're one hell of an actor and you have your ego under very close control, which is to be applauded.... but if that is not the case, well, then your courses may need a serious overhaul.
Honestly these videos are a bit too long for me. Not that they aren't good. I would like to see some design rules to make a design more efficient on time and space also improve performance. Perhaps 10-20min videos.
I may be doing shorter videos in future from the same topics - playing with simulations and trying different useful stuff which is good to know during schematic design and PCB layout.