When a Signal Hits The End of a PCB Track - What happens? | Reflections by Eric Bogatin
Вставка
- Опубліковано 20 тра 2024
- Ringing caused by reflections can be fixed. What is happening and what to do? Explained by Eric Bogatin. Thank you Eric.
Links:
- Software used in the video: www.bethesignal.com/bogatin/v...
- Signal Integrity Academy (use FEDSI - 3 months free): www.bethesignal.com/bogatin/3...
- FEDEVEL courses: courses.fedevel.com/
Chapters:
00:00 Understanding reflections
01:06 Electrically Short vs. Long
02:52 The tool we are using explained
03:48 Simulating: Electrically short
08:04 Reflections on classic CMOS and standard board
10:41 Simulating: Electrically long (e.g. DDR3 / DDR4)
15:16 50 OHM impedance is not 50 OHM resistance
16:39 Simulating a buffer example
18:32 Why we get reflections
20:37 What happens when there is a reflection
25:44 Why reflection disappears
30:08 Frequency of ringing
31:49 How to eliminate reflections
35:37 Signal level change caused by termination
39:25 Source termination
------------------------------------------------------
Would you like to support me? It's simple:
- Sign up for online courses hosted on our platform: marketplace.fedevel.education/
- 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 - Наука та технологія
One of the PCB Design Lord King👑! Thank you once again!!
It is useful to conceptualise electrical pulses as having mass and inertia.
Recall playing in a bathtub as a child sending waves and pulses to the end wall of the bathtub and the reflections and splashings that resulted.
Thank you again Eric! Your knowledge is allowing me to feel comfortable with more complex boards - I'm building things I couldn't have dreamt of a year ago.
(although right now I'm having a very easy time with a simple board, and these 600Hz PWM signals are going to be the cleanest signals these motor controllers have ever seen haha!!!)
Really great content. Very important concepts (like the waves traveling), along with important takeaways (like measure signals at the receiver) clearly stated.
Thanks Robert and Eric! Love your videos!
Always great to list to Eric
Thanks for the link Eric - looks like we should avoid stubs at all cost!
very good content as always, thank you!
Wow really nice explanation thank you. The bit about source termination surprised me and I had to stop and really think about it, but it makes sense! The input side sees the sum of the input signal and the reflected signal, but it takes a whole roundtrip for it to settle.
very educational video. Thank you for sharing your knowledge.
Wow! A very useful tools!
Excellent. I’ve been nagging about source termination to 1. Avoid maxing out output drivers into capacitive loads and 2. Avoid reflections/ringing.
This video is fantastic!
really informative
I was thinking that the tool should perhaps show (an interpretation\comprehension) of the interaction between the forward and reflected signals as a vortex rather than a diagonal edge - similar to a 'pipeline' beach wave. With that comprehension, it allowed me to understand how the reflected signal combines with the source.
This kind teacher i want in my life. Oh god he nailed it ❤❤❤❤❤❤.
Thanks 👍👍👍
I was just theoristing reasons for why reflections occur with my new found knowlege of how signals propergate and then I find this video. Let's hope the universe doesn't decide to stop reflecting or it will blow up.
It's better to say that it's about conservation of energy. The sender can't know what the receiver is doing due to the speed of light (in the cable)
Hi, can you please explain, where do I have to add a resistor to increase an impedance of a source? Do I have to add it parallel to a pin or how?
👍👍 Nice lesson! I did something similar by calculating the waveguide impedance and the reflection factors. In case somebody is interested to complement his know-how ..
Danke!
Thank you very much @xenofontzaras2741
And thats why you had 50 ohm bnc terminators on both ends on old coax network cable.
Very interesting thx!
Thank you very much @remy-
That’s so great, thank you again Eric, by the way, I don’t know how to use FEDSI to get a 3 months free, could you help me?
Great walkthrough. My question is what if the output resistance of drive is 50 ohm, it seems termination resistor is not required. But what is the downside?
high driving current
Thanks!
Thank you very much
Isn't this just all standard transmission line theory.... ie matched send build out resistance, line impedance and load impedence required to retain signal integrity !.
Tak!
Thank you very much
50 ohm impedance is desired from end to end but this gives half the voltage of the output. Isnt that a problem?
Watch the source termination 39:25 Notice the voltage on the input.
Seems that load should be higher impedance to cause reflections to build up to the full voltage of the reciever. So load being same impedance with source and line is not a good thing.
Why dont you terminate with 50ohm with capacitor to 0V to avoid halfing the voltage level?
Signal edge is built from many different frequencies and we need 50 ohm for every frequency. An ideal capacitor's impedance for low frequencies is very high (DC doesn't flow through a capacitor), an ideal capacitor's impedance for high frequencies is very low (capacitor is a short circuit for high frequencies). For real capacitors it is a little bit different, but still, a standard capacitor will not have 50OHM impedance for wide range frequencies - that is why resistor is a simpler solution. PS: watch 39:25 about halfling the level
@@RobertFeranec Yes, I think I was unclear about what I mean. Ringing is often rather high frequent. I want to avoid 50ohm loading on logic DC-levels (and low frequent digital pulses) and still maintain the logic levels without halfing it. Placing a capacitor in series with a resistor to 0V will give zero current (i.e. low power consumption) at DC and low frequencies. Problematic ringing is only a troublesome issue in the fast signal transitions. My intention is that the capacitor shall be a short circuit at high frequencies and thus the 50 resistor terminates high frequencies well and att low frequencies the capacitor is high impedant and thus disconnects the terminating resistor. This is how we sometimes handled emc-issues on high frequency databus-es when I was working at Flextronics.
Thank you for great videos Robert, (a little bit toooo long videos, but very interesting videos! :-) )
Yes you can do a AC termination system if the change of the signal is low enough compared to the capacitance of the wire. you need a capacitor at least 10 times the capacitance of the entire trace. and the signal can't change a a rate faster than the RC time constant. I use AC termination in many places where I want sharp edges without reflections, but my application meets the requirements. Edit: this is a Capacitor and a resistor in series. The frequencies only see the resistor and the capacitor only needs to hold on till the energy is absorbed.
@@LaserFur Yes, ofcourse... AC-temination is the word for it! Thanks
Wait, people still use Adobe Flash in 2022? 😂
Jokes aside, another great video, Robert. Thanks!
🙏 𝓟Ř𝔬𝓂𝔬𝐒ϻ