1
\$\begingroup\$

I have a product that will be subjected to reliability testing (random vibration and thermal shock) and I need to design a footprint for an SMT power inductor that will give the best mechanical strength to resist damage during vibration testing. I've looked through the IPC 7351 standard but I cant find anything that talks about how different pad size (resulting in different fillet shape) will affect the mechanical strength of the solder joint. The part I am working with is Murata LQH3NPN1R0MMEL. It has a mass of approximately 0.05 grams.

Does anyone have a method for calculating a land pattern for optimal mechanical strength?

\$\endgroup\$
  • 4
    \$\begingroup\$ Just use glue on the sides and fix it to the PCB after soldering. \$\endgroup\$ – Andy aka Jun 23 '17 at 18:00
  • \$\begingroup\$ @Andyaka wait, really? \$\endgroup\$ – tuskiomi Jun 23 '17 at 19:01
  • \$\begingroup\$ @tuskiomi maybe you have a suggestion? \$\endgroup\$ – Andy aka Jun 23 '17 at 19:07
  • \$\begingroup\$ @Andyaka I've got no clue. \$\endgroup\$ – tuskiomi Jun 23 '17 at 19:08
  • 1
    \$\begingroup\$ Not a bad idea Andy. But I don't have the budget for the extra cost of glue. \$\endgroup\$ – Sidearm Jun 26 '17 at 16:13
3
\$\begingroup\$

If you're serious about this, you must take into account the mounting geometry of the pcb, the material stiffness, and the location of the inductor. You'll need to do a FEA vibration analysis to identify nodes on the board and coupling of vibrational energy into the inductor footprint.

If you don't have the in-house expertise for that, and don't have access to an outside expert you can hire, you're best off making a blank pcb with just the inductor and shaking/shocking the hell out of it until it fails.

EDIT Sometimes the physical straightforward approach can save a lot of time over careful analysis. There's a story about Edison. Supposedly he assigned a task to a new engineer - determine the capacity of a an oddly-shaped glass bulb. The engineer spent a day taking careful measurements of the bulb and calculating precisely how much volume this entailed, after making due allowances for the (measured, of course, at several places) thickness of the bulb. Edison, it is said, took the bulb, walked over to a water tap, and filled the bulb, then poured the water into a graduated cylinder.

\$\endgroup\$
  • \$\begingroup\$ Percussive tuning? \$\endgroup\$ – tuskiomi Jun 23 '17 at 19:03
  • 1
    \$\begingroup\$ @tuskiomi - My favorite kind. \$\endgroup\$ – WhatRoughBeast Jun 23 '17 at 19:14
  • \$\begingroup\$ Thanks Beast. Unfortunately I don't have enough information to do a proper analysis. My board is a subsystem that will be mounted on a huge number of host PCBs with various thicknesses, sizes, and mounting arrangements. I guess I'll use a guess and check method, build up a few panels with different size footprints (maybe I'll adjust the paste coverage too) and see what one performs best in the vibration test. I have a feeling a larger pad with maximum paste coverage will be the best. I guess we will just have to wait and see. I'll share my results when they come back from testing. \$\endgroup\$ – Sidearm Jun 26 '17 at 16:11
  • \$\begingroup\$ @user36072 - "My board is a subsystem that will be mounted on a huge number of host PCBs with various thicknesses, sizes, and mounting arrangements." In that case, why are you doing vibration tests, and what standards are you using? Without knowing the environment the board will be operating in, any test will be useless. \$\endgroup\$ – WhatRoughBeast Jun 26 '17 at 16:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.