Chris is correct.
For those of you who wonder why the bow/twist (warpage) measurement is such a big deal, there are several reasons:
1. You may be able to process fabricated boards at the assembly level, but dealing with bow/twist above the limits or guidelines means issues when printing paste accurately. 2. It means parts sometimes not placing accurately also. Or both!
3. Most of all, when you go to put the warped finished CCA into its case or box and you assemble it with hardware, you are typically straightening it out and this puts pressure on the solder joints on the board, especially those that lie perpendicular to the direction of the warp. This stress is somewhat taken up by leaded solder due to the modulus of elasticity and the little bit of one-time relaxation that the lead provides (one more factor overlooked when not performing due diligence during RoHS), but lead-free solder is much more brittle and thus less forgiving and this can translate to cracked solder joints, pad cratering (especially on BGAs, which do not flex much), etc. Experts have studied this and determined warpage less than the limits defined are not likely to fail, but anything more than those limits will most likely result in premature SJ failure.
The more warpage, the more stress.
As far as "placing blame on the design", there are a number of design factors that must be considered. Copper balance is just one, and the most common. Another often overlooked warpage factor is the direction of the design with respect to the strand orientation in the PWB. Sometimes simply changing the orientation of the boards on the flat during fabrication can go a long way in fixing the issue. Quite often I have seen designs with 8 boards on a 24" by 24" flat, and then perhaps 3 more placed at 90 degrees with respect to the rest along one side or along the top or bottom of the flat to maximize the number of fabricated PWBs (or panels) in a flat. This is usually a good practice, but sometimes, for certain designs, this can result in either 8 warped boards or 3 warped boards from a given flat, especially for board thicknesses less than .062" with lots of copper. So orientation on the flat is a factor, thickness of the PWB is a factor, type of PWB material (as defined by IPC 4101B) is another, the fabrication method is another (pressed versus heat-stamped, etc.). The list goes on and on. All of these can work individually or together to cause warpage that may not be seen until after PWB fabrication. But after reflow the CCAs look like they were manufactured at the Lay's Potato Chip factory. (Those fabricated by Old Dutch are not only warped, but burnt with too much salt as well). :-)
-----Original Message-----
From: Chris Mahanna [mailto:[log in to unmask]]
Sent: Tuesday, May 22, 2018 9:02 AM
To: Stadem, Richard D; TechNet E-Mail Forum
Subject: RE: [TN] Bow and Twist measurements per IPC 650 TM 2.4.22, IPC 600-2.11 Flatness
Yep. You've referenced the 6012 bare board as-fabricated requirements that are reprinted in A-610 with the weasel word "should" attached to post assembly.
6012D added good detail regarding how to handle pallets (at completed bare board level). And if you parse the 6012 verbiage carefully, it says that method 2.4.22 is only attached to the *calculation* of percentage of "bow" and "twist".
We call this the "IPC flatness model" as opposed to the common ANSI model.
So, IMO the answer to Victor's question is: measure your assembly using whatever contact/non-contact method you think is suitable, just make sure that you use the IPC model for bow and twist.
If it those percentages exceed 1.5/0.75 PTH/SMT, you need to go look at your bare boards. If they conform, you're stuck with "should" at assembly level. If they don't conform, you may still be stuck with AABUS due to "unbalanced" caveat.
And of course, the unbalanced caveat is the real meat here, because it is what causes the problem during soldering (other variables held equal) and points blame towards design activity.
Chris
-----Original Message-----
From: Stadem, Richard D [mailto:[log in to unmask]]
Sent: Tuesday, May 22, 2018 9:01 AM
To: TechNet E-Mail Forum <[log in to unmask]>; Chris Mahanna <[log in to unmask]>
Subject: RE: [TN] Bow and Twist measurements per IPC 650 TM 2.4.22, IPC 600-2.11 Flatness
Au contraire.
Bow and twist requirements must be met during or after assembly per IPC-A-610 as per figure 10-23, on page 10-13, to be less than 1.5% for PTH, and less than .75% for SMT. IPC-610 also refers you to IPC-TM-650 method 2.4.22, as does the J-STD-001, IPC 6012, and as a condition for fabricated circuit boards, IPC-A-600. They ALL refer you to method 2.4.22, and they ALL have requirements for maximum bow and twist throughout all stages of fabrication and assembly.
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Chris Mahanna
Sent: Monday, May 21, 2018 4:24 PM
To: [log in to unmask]
Subject: Re: [TN] Bow and Twist measurements per IPC 650 TM 2.4.22, IPC 600-2.11 Flatness
To my knowledge, there is no IPC standard requirement (or method) for during/post assembly; much to the package manufacturer's chagrin.
Chris
Chris Mahanna
Robisan Laboratory Inc.
-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of [log in to unmask]
Sent: Wednesday, May 16, 2018 11:04 AM
To: [log in to unmask]
Subject: Re: [TN] Bow and Twist measurements per IPC 650 TM 2.4.22, IPC 600-2.11 Flatness
Folks,
Is there a consensus, especially from committee (7-11D) members, on the above inquiry regarding Bow & Twist for raw board and PCBAs (populated boards)?
Ref: IPC-T-50M (May – 2015) ( Victor’s comments )
Bow - (Fabric) Filing yarn that lies in an arc across the width of a fabric.
Twist - The deformation of a rectangular sheet, panel, or printed board, that occur parallel to a diagonal across its surface, such that one of the corners of the sheet is not in the plane
that contains the other three corners.
Victor,
From: Jose A Rios [mailto:[log in to unmask]]
Sent: Wednesday, May 16, 2018 8:35 AM
To: TechNet E-Mail Forum; Hernandez, Victor G
Subject: Re: [TN] Bow and Twist measurements per IPC 650 TM 2.4.22, IPC 600-2.11 Flatness
2.4.22 is only for bare (unpopulated) PWB’s, and the rigid portion of rigid-flex PWB's. Section 1 (Scope) explicitly excludes PCBA’s from being evaluated using this method.
José (Joey) Ríos, Sr QA Engineer
Mission Assurance Manager
Kavli Institute for Astrophysics & Space Research Massachusetts Institute of Technology [log in to unmask]<mailto:[log in to unmask]>
(617)324-6272
On May 16, 2018, at 7:17 AM, [log in to unmask]<mailto:[log in to unmask]> wrote:
Fellow TechNetters:
Is there a separate Bow & Twist Test/documentation for Printed Wire Boards, PWB, and Printed Wire Board Assemblies, PWBAs? Are they interchangeable......
Victor,
|