Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Parallel Termination-Reflections at Source Side

Status
Not open for further replies.

hioyo

Advanced Member level 4
Full Member level 1
Joined
Aug 18, 2021
Messages
116
Helped
0
Reputation
0
Reaction score
2
Trophy points
18
Activity points
899
Dear Team,

I know that in parallel termination we are placing a termination resistor at the load side and no termination resistor at the source side.

There is a mismatch between driver output impedance and characteristic impedance of the line and there will be reflections at source side(Please correct me if I am wrong).

Will this reflections cause any problems or this is also taken care by parallel termination resistor.

Or do we need to place termination resistors at both ends
1643641934732.png


Regards
HARI
 

Hi,

for a reflection, you need a signal first. Where does this signal come from?

Klaus
 

You probably want to put the same R in series with the source end or you mismatch Zo to the (almost) zero impedance of the driver output. You will get a signal division by two through.

Brian.
 
You need to specify required signal quality for your application, signal speed, termination and line impedance accuracy. As a general hint, for highest speed (e.g. Gigabit serial protocols), also for long cables double side termination is standard, sub Gigabit signals are often work with single side termination.
 
No,

The driver is the source. It can not be "reflected by itself".

See it like throwing a stone into water. The waves start at the place where the stone meets the water. This is the source.
But the waves will not be reflected there, unless they come from another source.
A reflection is always different in place and the signal is different in time. (With respect to the source)

The source "as source" is not the problem.
Possible problems are:
* if the destination is not properly terminated, then the signal gets reflected and travels back to the source.
--> use proper termination at the destination
* Or another signal is coupled (capacitive, inductive, conductive) into your signal line.
--> use proper cables and proper wiring.

This does not mean that a series termination is useless at all. It has it's benefits mainly on lenghty signal lines with jumps in characteristic impedance, maybe caused by (bad) wiring, connectors, junctions ....

Klaus
 
Thank you.
The output impedance of the driver(Source) is Rd and that of the TX line is Zo.If they are different there can be reflections at the source side.Please correct me if I am wrong
No,

The driver is the source. It can not be "reflected by itself".

See it like throwing a stone into water. The waves start at the place where the stone meets the water. This is the source.
But the waves will not be reflected there, unless they come from another source.
A reflection is always different in place and the signal is different in time. (With respect to the source)

The source "as source" is not the problem.
Possible problems are:
* if the destination is not properly terminated, then the signal gets reflected and travels back to the source.
--> use proper termination at the destination
* Or another signal is coupled (capacitive, inductive, conductive) into your signal line.
--> use proper cables and proper wiring.

This does not mean that a series termination is useless at all. It has it's benefits mainly on lenghty signal lines with jumps in characteristic impedance, maybe caused by (bad) wiring, connectors, junctions ....

Klaus
 

Hi,
If they are different there can be reflections
Again: where does the signal come from?

Take a piece of paper and a pencil and draw the way of the signal with o e color and it's reflection with a different color.

Reflect means "pushed back". Back! Back to where it is coming from (when you are at the end if a cable). When point "S" is the point of reflection, then the signal needs to come from else where.

Klaus
 

I think the discussion is a kind of sophistic. No load termination is perfect, no transmission line is without discontinuities.

The question about purpose of double termination isn't "how can reflected waves arrive at the source?" but "do for- and backreflection affect the signal quality significantly?" The answer can be different depending on the application. If you are e.g. dealing with high speed analog or gigabit signals, necessity of double termination is out of question.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top