Breakout Board Pcb Layout Review

Thread Starter

Avshi

Joined Jan 16, 2015
48
Hello! I have a design here and I would appreciate some feedback before I send it out to be fabricated. The board is breakout board for 3 different atmel microcontrollers (SAMS70J21A,ATUC256L3U, or any 64 pin 0.5mm pitch uC and SAM C20E or similar 32-pin uC). The board also has an RS-485 module broken out and a place for an Ethernet port as well as a 5V to 3.3V regulator. I've included the gerbers and I can export to orcad if that's what someone wants.
Please let me know what you think.
Thanks,
Avi

 

Attachments

Thread Starter

Avshi

Joined Jan 16, 2015
48
I'll add implement the decoupling layout tomorrow or the day after and update the post. Is that the only issue?
 

Thread Starter

Avshi

Joined Jan 16, 2015
48
I've bread boarded U4 and U3 (3.3V voltage regulator and rs-485 module respectively) and confirmed them to be functional in the arrangement in which they're placed. If you want to look into them, the vreg is lm3940-3.3 and the rs485 module is the ISL3259E. P10 is 5V and ground input as well as 3.3V output to power the various pins which have been decoupled in single supply mode according to the documentation.

My goal:
Since there are three micro controllers possible and only one can be present on the board at a time, the related components are labeled with A,B,C,J where A is a cap related to SAMC20E, B to ATUC256L3U, C to SAMS70J21A and J is a jumper in the form of a 0 Ohm resistor. So for example, if I wanted to experiment with ATUC256L3U I would fill in all of the B components and none of the others.
This is what I've got so far.


 
Last edited:

nerdegutta

Joined Dec 15, 2009
2,684
Are one of you ideas that this breakout board will fit into a breadboard? I noticed that you've written 2.8 in grey. Most of my breadboards have 2.8cm between the pinholes far out one each side.

aac_caliper.JPG
If that is you plan/idea, it could be tight...
 

Thread Starter

Avshi

Joined Jan 16, 2015
48
I tried to reduce the board width, but the maximum amount of space I could wrangle was 2.54mm. I guess i'll just have to find a bigger bread board or use two of the narrower ones. Other than that, the layout looks passable?
 

nerdegutta

Joined Dec 15, 2009
2,684
Other than that, the layout looks passable?
I will not vouch for that this board is flawless.

C1 and C3 should change places, (if you see the labelling on the board, and the labelling in the schematic), and I feel that the capacitor on the LM3940 OUT pin is too high. 470uF might be too much.

Skjermbilde fra 2017-01-15 18:29:32.jpg
From datasheet, page 9.
 

Thread Starter

Avshi

Joined Jan 16, 2015
48
The ESR of my capacitor is within the acceptable range. Why would the value of capacitance effect the output if the datasheet says that the capacitance must be over 33uF and can be increased without limit? It even says that larger caps will give better transient response.
 
Last edited:

RichardO

Joined May 4, 2013
2,270
I tried to reduce the board width, but the maximum amount of space I could wrangle was 2.54mm. I guess i'll just have to find a bigger bread board or use two of the narrower ones. Other than that, the layout looks passable?
Sometimes boards like this can be made narrower by rotating the chips 45 degrees. I know this sounds crazy but give it a quick try and see if it works for you.
 

Thread Starter

Avshi

Joined Jan 16, 2015
48
Thanks for the suggestion! I got curious and I managed to reduce the distance between the pins to 25.4mm which leaves one bread board pin on each side (at least on mine). So, would this be good to send to a board fab? I think it looks good to send out, but a second opinion would make me feel better.
 
Last edited:

ErnieM

Joined Apr 24, 2011
8,377
The easiest and most complete way to discover PCD defects is to build one and test it.

The industry average is nearly three turns per design. That's about my number too.
 
Top