0
\$\begingroup\$

I have a multi layer board and have 2.5V plane as one of the layer. Some how during testing something go shorted on 2.5V. I am using the Toneohm 950 to detect the plane short after removing whatever components possible. I am left with 12 BGA ICs which cannot be reworked at lab.

But the surprising part is the instrument is detecting the short near power entry. Is it possible that my 2.5V somehow sparked to reach the ground via inner plating?

Please refer the picture below. This is an inner layer of 2.5V(yellow) and blue color vias are ground thru-hole vias.

The minimum distance between the 2.5V and ground via is 15mils.

enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ "during testing something go shorted"...as in the short was not there and then occurred during testing or would it be more correct to say a short that was there from the time the board was made was detected during testing? \$\endgroup\$ – Norm Jul 20 '18 at 13:46
  • 1
    \$\begingroup\$ Are there any grounded metal parts on those connectors? Any traces running under them, or stray solder to the 2.5V line? A hot air pencil may deal with the latter. \$\endgroup\$ – Brian Drummond Jul 20 '18 at 14:25
  • \$\begingroup\$ @Norm The board was absolutely fine initially, in course of testing I soldered some wires for I2C bus line for external programming. some how those wires got touched on some other part of board. This led to the short cicuit. Thanks!! \$\endgroup\$ – pavan Jul 23 '18 at 6:25
1
\$\begingroup\$

It seems unlikely, 15mils clearance is pretty generous.

Make sure you have removed all ceramic bypass capacitors. They can go short.

Also check that there is pullback of the plane from the edges of the board and look at what's stuck into in unplated holes especially if they don't have pullback.

\$\endgroup\$
1
\$\begingroup\$

I know your question was specifically about how your board might have shorted around the power entry area, but on the topic of detecting shorts...

From my admittedly limited use of ToneOhms, I found them not very useful for detecting shorts. Do you have access to a thermal camera and very low voltage lab DC power supply? If so...

With your board disconnected from its regular power supply, wire the lab supply to your shorted rail and set it to zero volts. Turn it on, then VERY SLOWLY increase the voltage until you're pulling a hundred miliamps or so. Look for hot spots on the board using the thermal cam. If you don't see any, continue VERY SLOWLY increasing the current until a visible hot spot develops.

If you don't have a thermal camera, a laser spot thermometer works OK too, it's just a bit more tedious.

Obviously don't increase the voltage beyond the regular operating voltage of the rail (2.5V in your case), but this should be a moot point. If you really have a short, you'll get some fireworks long before you get that high.

\$\endgroup\$
  • \$\begingroup\$ yes, I felt the same in using toneOhm. I will try if i have an infrared thermometer and to avoid any fireworks, I will limit the current from the DC power supply. \$\endgroup\$ – pavan Aug 29 '18 at 5:22

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.