Maker Pro
Maker Pro

Power supply return path question

MChapman

Sep 9, 2013
9
Joined
Sep 9, 2013
Messages
9
Attached are the datasheet, schematics and two design options for a switching supply boosting 7.2v to 12v,20mA. My question is intended to be general but I wanted to include my current project as an example. The first design is intended to have a return path shorter than the return path of the second design. I believe I understand the role switching frequency plays on the behaviour of the return path:
1. >MHz frequency and the return path is dictated by inductance.
2. <MHz frequency and the return path is influenced by resistance of the copper plane.

I've seen designs on forums before where a responder would jump on the design for having too long of a return path while it didn't look very long at all. Should I expect behavioural differences in the two layouts here? How would I determine how long is too long of a return path?



1615-1.JPG First Design.JPG Second Design.JPG
 

Attachments

  • 16151fas.pdf
    164.5 KB · Views: 116

AnalogKid

Jun 10, 2015
2,884
Joined
Jun 10, 2015
Messages
2,884
Your generalizations of the characteristics of return paths vs. frequency are too broad. Yes, inductive reactance increases with frequency, but it always is a performance factor. Also, area and thickness are factors that contribute to the effectiveness of a return path design as much as length. For example, reducing the thickness of a trace decreases the inductance but increases the resistance. I'm not familiar with your rule of thumb for return paths, but I have one: less is more.

ak
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
Another factor is the loop area enclosed by the high-frequency conducting traces. Keep the loop area as small as possible to minimize radiated emissions.
 

MChapman

Sep 9, 2013
9
Joined
Sep 9, 2013
Messages
9
Looking for specifics. I would only adjust one variable at a time here. Frequency. Once a board is designed I cannot influence copper pour therefore resistance due to copper is constant through the scenarios.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Attached are the datasheet, schematics and two design options for a switching supply boosting 7.2v to 12v,20mA. My question is intended to be general but I wanted to include my current project as an example. The first design is intended to have a return path shorter than the return path of the second design. I believe I understand the role switching frequency plays on the behaviour of the return path:
1. >MHz frequency and the return path is dictated by inductance.
2. <MHz frequency and the return path is influenced by resistance of the copper plane.

I've seen designs on forums before where a responder would jump on the design for having too long of a return path while it didn't look very long at all. Should I expect behavioural differences in the two layouts here? How would I determine how long is too long of a return path?



View attachment 20901 View attachment 20902 View attachment 20903

What are you interested in? EMC related problems or operability. Signal integrity in many cases is a design specific environment, and it's down to you to decide if it's acceptable. Radiated, conducted emissions and immunity is a requirement by certain standards to be below a certain level.

Signal integrity and EMC often go hand in hand and if you have good signal integrity then you would have good EMC performance and far as emissions is concerned.

So tell us your design requirements and we can give a bit more advice.


Cheers
Adam
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
This is how I would do it for a single sided PCB. You could easily machine this on a PCB milling machine. I would normally use more than one layer if I was doing it for my work but it gives you an idea of the kind of layout that should work.
Adam

LT1615SCH.PNG

LT1615.PNG
 
Top