Author Topic: Any ideas for solving boards bending problems  (Read 3167 times)

0 Members and 1 Guest are viewing this topic.

Offline ErickbenTopic starter

  • Contributor
  • !
  • Posts: 15
  • Country: au
Any ideas for solving boards bending problems
« on: June 29, 2016, 10:09:48 am »
Recently, one of my friends said the PCB he received is bending, so that he can't place and pick the components on exact way. Then we can't help but wonder Why cause PCB bending? Hope it can help you avoid this situation if meet such an "accident".
(Welcome to add what's missing factors and correct my obvious mistakes, thank you.)

We summarize as following:

1.Uneven copper on PCB will worsen boards bending and  warp.
2. The circuit board layers connecting point (vias) will limit the board to expand and contract.
3. The circuit board itself weight will cause deformation of the board recess.
4. V-Cut depth and connecting bar will affect the amount of deformation panel.

And we also think about how to avoid this?

1. To reduce the effect of temperature on the board's stress.
2. Using high Tg plate.
3. Increase the thickness of the circuit board.
4. Reduce the size of the board and reduce the number of panel.
5. Use the furnace tray fixture.
6. To use Router replace V-cut.

Welcome any suggestions from all of you!!!
try everything!!!
 
The following users thanked this post: nail10

Offline Rerouter

  • Super Contributor
  • ***
  • Posts: 4694
  • Country: au
  • Question Everything... Except This Statement
Re: Any ideas for solving boards bending problems
« Reply #1 on: June 29, 2016, 10:17:07 am »
Biggest factor i have found is vastly different copper density on each side, e.g. 99 vs 15%, flood filling with ground on both sides on the second version solved that issue, as it brought it closer to 99 / 85%.
 

Offline Alice Rayming

  • Newbie
  • !
  • Posts: 2
  • Country: sg
Re: Any ideas for solving boards bending problems
« Reply #2 on: June 29, 2016, 10:24:22 am »
Thanks for your post.
There is another suggestion to avoid this issue for your reference:
can bake the working plate before production.

Alice
2016.6.29

 

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: Any ideas for solving boards bending problems
« Reply #3 on: June 29, 2016, 11:36:28 am »
In my experience this is usually down to very different copper density on the two sides, double sided RF stuff with a solid copper pour for a ground plane on the back and low copper density on the front is notorious for doing this.

You also seem to be using a thin substrate (0.8, 1mm) that will make this worse.

Personally I would be sending that back to the board house, but they might tell you they need you to amend the gerbers to resolve the issue.

Regards, Dan.
 

Offline PCB.Wiz

  • Super Contributor
  • ***
  • Posts: 1535
  • Country: au
Re: Any ideas for solving boards bending problems
« Reply #4 on: June 29, 2016, 10:13:58 pm »
Recently, one of my friends said the PCB he received is bending, so that he can't place and pick the components on exact way. Then we can't help but wonder Why cause PCB bending?

That's some serious bending you have there.
Was this shipped or stored bent ?
Is the treatment the same on both sides ? - hard to tell from the image, but is that side all solder, no mask ?
If it is all solder, then applying and thinning that would be tricky, to avoid bending.

Did maybe someone decide solder was a cheap way to increase metal thickness ? - best to avoid that, and specify thicker copper on your boards.
 

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2441
  • Country: us
Re: Any ideas for solving boards bending problems
« Reply #5 on: June 30, 2016, 01:46:45 am »
Hi

One solution is to tie it down when you do the pick and place. Even with a much flatter board, that's a pretty good idea. As others have said, the board needs to be "same / same" on both sides. My normal solution is to pour ground or power on every layer. Both sides get solder mask over > 90% of the surface. The same plating goes on both sides of the board.

Another thing to look at is layer stack up. You don't mention how many layers are in the board design. The same symmetry issues apply on stack up. If you have 4 layers with three 3 mil cores on the top, you put 4 layers with three 3 mil cores on the bottom. If your top layer is 0.5 oz copper then the bottom layer is 0.5 oz copper. That might give you a 5 mil inner core to get you to 0.8 mm.

Lots of variables ....

Bob
 
The following users thanked this post: montemcguire


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf