Mastering the TDR in 45 Minutes - Eric Bogatin

[@brokenicry]

You know someone has mastered his craft when he can explain it to a beginner and the beginner understands it like a specialist... This is so great

[@leecyrille]

Starts at 6:20

[@alexiscozart3495]

Thank you!

[@優さん-n7m]

This one is Part 1 of 6, where are the remaining parts?

[@diffynou]

What is the difference between a TDR and a couple of function generator + scope ? Rise time is much faster on TDR devices ? No multiple interconnection reflections ? TDR can deliver much more power thus drive very long capacitive cables ?

[@IBasilisvirus]

TDR can see the direction of the current on the return plane. If the current gets inside the TDR's return plane or the current goes out of the TDR's return plane. Think of it like TDR has series output/input resistors on its output and return path so it can see the direction of the current from the voltage drop on these resistors.

[@javier-franciscomartinez4901]

Thank you for this instructive presentation on TDR. It is very refreshing old myths relating transmission line reflections are explained with with clear and simple examples.

[@paulpaulzadeh6172]

@35:14 , you get offset less then 50ohm , less resistive , is it ?! , but in other test you get higher offset , higher resistive , is it ?

[@paulpaulzadeh6172]

this means that on PCB , changing layer by via is not good if you want uniform 50 ohm impedance , is it ?!

[@yeaveny3629]

is it?

[@Bornachiever17]

No you can but maintain return path without screwed so use return vias on layer transition there is proper return path hence no impedance discontinuity

[@TSRHelios]

@@Bornachiever17 with a return via the signal via is like a very short microstrip without impedance control. I remember I have seen a US pattern regarding impedance controlled via.

[@yuraamirov7893]

you are correct :) there is some way to improve it however . you can add some ground vias near by for the return path , by doing so , you will minimize the change of impedance

[@allanknox8216]

Everbody responding is giving you good advice. Avoid layer changes if you can, and when you must, be sure you add some ground vias to compensate in the return path. You can then keep the discontinuity small and (hopefully) insignificant depending on operating frequency.

[@edouardmalot51]

Can we get the small book online ?

[@方方-r7y]

A stupid question: how many reflections happened at each time position on the waveform? Only one reflection or accumulate of multi reflection? Such as there is bounce at the point of discontinue, is the bounce caused by one reflection?

[@tforionnaa7213]

for TDR, since you are measuring at one physical location along the line, you will see multiple reflections manifest as events at multiples of the first reflections distance. I will refer to distance as time to simplify propagation velocity of the medium. So, if the line is 5ns long, the first reflection will show up at 10ns, then another will appear at 20ns, 30ns, etc. The reflection always has to transit twice the distance between the impedance discontinuity so that is why a line that is 5ns will have an event appear at 10ns. The secondary reflections are harder to interpret but will help you confirm if you have issues at the launch into the DUT, DUT impedance, and termination impedance. There are many combinations of mismatches happening so it is best to simulate a transmission line mismatch, load mismatch, and source mismatch in a reliable software first to gain intuition. there will be a reflection for each change in impedance along the line so if your launch into the DUT is farther from the transmitter and receiver and you implemented a poor launch connection, there will also be a reflection there. the reason he mentions the launch as just a spike on the rise edge is because the distances are short and just look like impulses.

[@josephburton92]

Thank you I appreciate this.

[@woutermelon]

Truly shows what a genius this guy is. If one of his books turns out to be full of crap he can still sell it as fiction!