Author Topic: (Potentially stupid) crazy BGA pad design  (Read 4640 times)

0 Members and 1 Guest are viewing this topic.

Offline blueskullTopic starter

  • Supporter
  • ****
  • !
  • Posts: 367
  • Country: cn
  • BA7LKP
(Potentially stupid) crazy BGA pad design
« on: February 25, 2017, 08:41:59 am »
Hi all. I'm working on a ultra small form factor design which has a goal of using as TINY parts as possible.
As part of the challenge, I need to find a PMIC that can power my CPU while being as small as possible.
All WLCSP parts I can find are offered in 0.4mm pitch, not 0.5mm, therefore standard 0.1mm/0.1mm PCB process without microvia cannot layout this.
I only need to bring a few balls out, not all.
Essentially, I cut 1/4 of the 0.2mm ball pad to satisfy 0.3mm spacing between 2 pads, so that I can route 0.1mm trace in between.
Do you think this is a viable solution, or a stupid one?
 

Offline radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 687
  • Country: us
Re: (Potentially stupid) crazy BGA pad design
« Reply #1 on: February 25, 2017, 05:14:44 pm »
Wouldn't the ball just re-form into a sphere when reflowing?
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26751
  • Country: nl
    • NCT Developments
Re: (Potentially stupid) crazy BGA pad design
« Reply #2 on: February 25, 2017, 09:24:16 pm »
I'd keep the total pad area equal so there is a chance the solder blobs are at least the same in height.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Scrts

  • Frequent Contributor
  • **
  • Posts: 797
  • Country: lt
Re: (Potentially stupid) crazy BGA pad design
« Reply #3 on: February 26, 2017, 03:36:24 am »
I'd keep the pads, break the routing rule and let PCB manufacturer to eat it...
 

Offline Someone

  • Super Contributor
  • ***
  • Posts: 4509
  • Country: au
    • send complaints here
Re: (Potentially stupid) crazy BGA pad design
« Reply #4 on: February 26, 2017, 05:21:08 am »
I'd keep the pads, break the routing rule and let PCB manufacturer to eat it...

I've talked with the Wurth guys, they told me they will not accept non-manufacturable designs.
Also, making non-manufacturable designs means if my boards failed e-test, I will be held responsible. I don't want to gamble with my own money.
You know, students usually don't make much more money than they need to survive ;).
Boo hoo hoo, you poor struggling student who can't afford a mid priced board house....
Also, I need one that is light enough that I can carry it with me without a backpack, and I need the kind of warranty that gives no-bullshit repair or return. I will use it in a real HV lab/field environment, and accidents do happen.
I will run Altium Designer on it, which on its own is an $8k software. Its license agreement allows up to 2 installations at a time, so essentially I'm locking a $4k license on the new machine. Therefore, saving pennies is not my main focus on this purchase.
The reason I say this is because I paid one of the previous Scope Month grand prize winners $8.5k to buy a fully loaded 6GHz MSOX6004A
It pays to be careful when you've got a public profile that keeps all this visible. Showing off about wealth is particularly repellent in most cultures, and then to cry poor just makes you seem entirely insincere.
 

Offline hendorog

  • Super Contributor
  • ***
  • Posts: 1617
  • Country: nz
Re: (Potentially stupid) crazy BGA pad design
« Reply #5 on: February 26, 2017, 07:35:08 am »
No idea if there are any other issues but I just checked PCBCart and they do 3mil trace width and spacing. That is 0.0762mm.

http://www.pcbcart.com/pcb-fab/standard-pcb.html
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: (Potentially stupid) crazy BGA pad design
« Reply #6 on: February 26, 2017, 08:09:09 am »
I asked around the DFM guys and was recommended to use LDI (Laser Direct Imaged) soldermask for fine CSPs.

I wouldn't recommend putting a trace inbetween pads (fortunately I didn't have to; the central pins were power connections in parallel with the sides, so no trace-weaving was necessary).

If you have to, the tighter alignment of LDI should help at least.  Try making the solder mask expansion small (0 to 2 mils).

Shaving pads I think would be okay, as long as you adjust the overall diameter to keep the same pad area.  It's kind of getting into the realm of the superstitious though... who knows.

For sure, regular fab mask isn't going to cut it.  See if LDI adds much cost.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf