Discussion:
[EE]: 2 layer PCB layout. 12v and 5v tracks stacked. Good or bad
Justin Richards
2018-06-28 03:36:34 UTC
Permalink
Working on a 10cm x 10cm board.

Vin is 12V, gets converted to 5v and 3.3v.

Bottom plane is ground and top layer will be 3.3v flood.

As always, space is constrained.

I have routed the 12v (Vin) 2mm trace on the top layer around half of the
perimeter.

I want to route the regulated 5v 2mm trace directly underneath the 12v
trace.

Is this a bad idea. Should the 5v trace be routed such that it also runs
on top of the ground plane.

The 12v trace will be supplying an inductive load that gets switched on and
off and draws about 500mA.

Attached top half of the brd. Tried to make attachment as small as
possible.

Cheers justin
e***@dea.spamcon.org
2018-06-28 05:12:06 UTC
Permalink
Justin Richards wrote:
<snip>
Post by Justin Richards
I have routed the 12v (Vin) 2mm trace on the top layer around half of the
perimeter.
I want to route the regulated 5v 2mm trace directly underneath the 12v
trace.
Is this a bad idea. Should the 5v trace be routed such that it also runs
on top of the ground plane.
The 12v trace will be supplying an inductive load that gets switched on and
off and draws about 500mA.
The last is the bit that worries me i.e spike coupling to the 5v rail.
As the pcb designer I would throw it back as a query to the customer.
:-)

George Smith
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Randy Dawson
2018-06-28 05:33:16 UTC
Permalink
Well heck, make that 500mA trace as large as possible, and decap it to the gnd plane with a bulk (10uf) and several ceramic NPOs near the switcher, on a plane to the pin and very nearby, with vias to gnd.


________________________________
From: piclist-***@mit.edu <piclist-***@mit.edu> on behalf of Justin Richards <***@gmail.com>
Sent: Wednesday, June 27, 2018 8:36 PM
To: Microcontroller discussion list - Public.
Subject: [EE]: 2 layer PCB layout. 12v and 5v tracks stacked. Good or bad

Working on a 10cm x 10cm board.

Vin is 12V, gets converted to 5v and 3.3v.

Bottom plane is ground and top layer will be 3.3v flood.

As always, space is constrained.

I have routed the 12v (Vin) 2mm trace on the top layer around half of the
perimeter.

I want to route the regulated 5v 2mm trace directly underneath the 12v
trace.

Is this a bad idea. Should the 5v trace be routed such that it also runs
on top of the ground plane.

The 12v trace will be supplying an inductive load that gets switched on and
off and draws about 500mA.

Attached top half of the brd. Tried to make attachment as small as
possible.

Cheers justin
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Justin Richards
2018-06-28 05:54:59 UTC
Permalink
So no real issues it seems with having 12v tracks sitting directly above 5v
tracks.

However a sprinkling of 10uf capacitors along the 12v trace with some
ceramics near the switcher.

I will assume ceramics near and on the output of the switcher and others
near the input to the devices it is supplying.

And 10uf caps near and on the input to the switcher.

Cheers justin
Post by Randy Dawson
Well heck, make that 500mA trace as large as possible, and decap it to the
gnd plane with a bulk (10uf) and several ceramic NPOs near the switcher, on
a plane to the pin and very nearby, with vias to gnd.
________________________________
Sent: Wednesday, June 27, 2018 8:36 PM
To: Microcontroller discussion list - Public.
Subject: [EE]: 2 layer PCB layout. 12v and 5v tracks stacked. Good or bad
Working on a 10cm x 10cm board.
Vin is 12V, gets converted to 5v and 3.3v.
Bottom plane is ground and top layer will be 3.3v flood.
As always, space is constrained.
I have routed the 12v (Vin) 2mm trace on the top layer around half of the
perimeter.
I want to route the regulated 5v 2mm trace directly underneath the 12v
trace.
Is this a bad idea. Should the 5v trace be routed such that it also runs
on top of the ground plane.
The 12v trace will be supplying an inductive load that gets switched on and
off and draws about 500mA.
Attached top half of the brd. Tried to make attachment as small as
possible.
Cheers justin
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Brent Brown
2018-06-28 12:00:59 UTC
Permalink
Post by Justin Richards
Working on a 10cm x 10cm board.
Vin is 12V, gets converted to 5v and 3.3v.
Do you have any/much circuitry directly running off the 5V? I.e. if the 5V rail is
merely a pre-regulated supply for 3.3V linear reg(s) then there is a degree of
isolation - and a convenient place to add filtering/suppression.
Post by Justin Richards
Bottom plane is ground and top layer will be 3.3v flood.
As always, space is constrained.
I have routed the 12v (Vin) 2mm trace on the top layer around half of the
perimeter.
I want to route the regulated 5v 2mm trace directly underneath the 12v
trace.
Is this a bad idea. Should the 5v trace be routed such that it also runs
on top of the ground plane.
Not to say it's a bad idea, but on top of each other does intuitively seem to be the
formula for best possible coupling. The toss up you say is space, and if space wins
that's fine, as long as effects of coupling are negligible on performance. If
performance proves to be a factor, then cost is probbaly the next thing to look at...
e.g. 4 layer board.
Post by Justin Richards
The 12v trace will be supplying an inductive load that gets switched on and
off and draws about 500mA.
Nice thing about inductive loads is current ramps up and down slowly. Propblem of
course when there is no flyback diode/suppression. If the inductive load is on the
circuit board it's somewhat easier to account for. Off board needs a little more
thought... long wires pick up noise from "outside" and couple back to your 5V, etc.

Just my 2c worth.
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Justin Richards
2018-06-29 09:35:13 UTC
Permalink
Post by Brent Brown
Nice thing about inductive loads is current ramps up and down slowly. Propblem of
course when there is no flyback diode/suppression. If the inductive load is on the
circuit board it's somewhat easier to account for. Off board needs a little more
thought... long wires pick up noise from "outside" and couple back to your 5V, etc.
Inductive load is off board, up to 4 meters. I make a habit of adding
suppression diodes as close as possible to the load (magnetic locks).

I have seen designs where they are placed on the board with the customer
simply connecting the inductive load without giving it a second thought.

Perhaps both is best.

Thanks to all. I will get these boards made then bench test and scope the
rails etc to see how it all behaves.
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Loading...