Author Topic: LGA solder pad problem  (Read 1023 times)

0 Members and 1 Guest are viewing this topic.

Offline bittumblerTopic starter

  • Contributor
  • Posts: 37
  • Country: de
LGA solder pad problem
« on: August 01, 2021, 10:47:48 am »
Hi,

I am trying to design a PCB for a BME680 sensor.
That is a LGA 3x3mm package with a metal top.
The 8 pads on the package are 0.4x0.4mm.

My problem is, i do not get proper solder contacts on all pads.

My process uses a PCB from one of the chinese PCB services, stainless steel stencil for solder paste, hand placement, reflow oven.
This works fine for similar sized packages, eg. 3x3 WDFN.
However WDFN have their pads on the edge, LGA somewhere in the middle.

My question is, are there any tips/optimizations i can do to improve results:
- make pads on PCB bigger? (Currently i use size per data sheet)
- extend PCB pads over the edge of the package?
- have a via in the PCB pads to be able to rework individual pads from bottom side?
- do i need a special solder paste for this?
- other ideas?

Best Regards

Matthias
 

Offline AmnevaR

  • Contributor
  • Posts: 40
  • Country: ua
Re: LGA solder pad problem
« Reply #1 on: August 01, 2021, 08:24:39 pm »
It should work as is. I'll suspect temperature profiles for reflow oven first.
Yeah I know how it sounds. But I've seen it a few times. It's either the temperature profile is not right, or the PCB has a big copper planes that steal heat, or even tall components that ''create turbulence'' or "shadow" for airflow.
Also there is a solder paste that is better suited for BGA work.

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: LGA solder pad problem
« Reply #2 on: August 02, 2021, 01:26:18 am »
Check that you don't have too much solder paste on there. On pads under a package there's nowhere for excess solder to go and it can physically hold the package up in the air and prevent other pads soldering properly. Check the manufacturers recommendations as to solder paste area and stencil thickness. Don't be surprised if you have to go to as little as 50% paste coverage on your pad, you almost certainly don't want more than 80%.

Edit: Also, make sure that your pads have enough allowance over the size of the LGA pad to allow for fillets - say 100um on every side a solder fillet will form.
« Last Edit: August 02, 2021, 01:29:06 am by Cerebus »
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 

Offline bittumblerTopic starter

  • Contributor
  • Posts: 37
  • Country: de
Re: LGA solder pad problem
« Reply #3 on: August 02, 2021, 04:49:47 am »
Thanks for the tips.

There also seem to be two types of pad designs:
"Solder Mask Defined": pad is bigger than solder mask.
or
"Non Solder Mask Defined": pad is smaller than solder mask.

Any experience, which variant gives better results for newbies?
 

Offline bittumblerTopic starter

  • Contributor
  • Posts: 37
  • Country: de
Re: LGA solder pad problem
« Reply #4 on: August 02, 2021, 05:31:53 am »
Hi AmnevaR,

the only thing i can think of stealing heat, is the metal lid of the package.
Do you know how the heating of the pads is supposed to work?
From the top through the package, like in BGA?
Through the edges of the package?
From the bottom?
The bottom of the package seems to be a tiny PCB itself.
 

Offline AmnevaR

  • Contributor
  • Posts: 40
  • Country: ua
Re: LGA solder pad problem
« Reply #5 on: August 02, 2021, 06:54:54 am »
the only thing i can think of stealing heat, is the metal lid of the package.
The metal lid of the package could be a problem if your reflow oven has infrared lamps (metal lid could act like a mirror). At least I've seen this behavior with 4G modules that also have metal lid.

I would also recommend to check manufacturer's soldering instructions:
https://www.bosch-sensortec.com/media/boschsensortec/downloads/handling_soldering_mounting_instructions/bst-bme680-hs000.pdf

Ideally, if your reflow oven doesn't have a conveyor, or at least doesn't have very long conveyor ;D you could attach a thermocouple to the side of the package (or to the PCB next to the package) and check if the temperatures inside of the reflow oven correspond to the soldering profile.

There a lot of factors, I hope my suggestions are useful.

Offline bittumblerTopic starter

  • Contributor
  • Posts: 37
  • Country: de
Re: LGA solder pad problem
« Reply #6 on: August 02, 2021, 08:14:38 am »

The metal lid of the package could be a problem if your reflow oven has infrared lamps (metal lid could act like a mirror). At least I've seen this behavior with 4G modules that also have metal lid.

This might be the problem. My oven is a Puhui, which uses infrared heating.
Thanks for the tip.
 
The following users thanked this post: AmnevaR

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: LGA solder pad problem
« Reply #7 on: August 02, 2021, 12:51:38 pm »
Thanks for the tips.

There also seem to be two types of pad designs:
"Solder Mask Defined": pad is bigger than solder mask.
or
"Non Solder Mask Defined": pad is smaller than solder mask.

Any experience, which variant gives better results for newbies?

When dealing with parts where coplanarity is a issue, like DFN, QFN and LGA solder mask defined pads can be problematic because the height of the solder mask may not be particularly even (depending on the method used to produce the solder mask). You're also at the mercy of the adhesion of the solder mask to the pad, and the definition (i.e. resolution) of solder mask versus copper. Generally most board shops will be more precise in where the copper is placed versus where the solder mask is placed.

I'd say that, in general, unless you've got an explicit good reason to use a solder mask defined pad don't, make non-solder mask defined pads your default.
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 
The following users thanked this post: AmnevaR


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf