ShortSniffing – ShortSniffer: World's Best PCB Short-Circuit Locator

The key operating concept is simple:

Follow the Sound to the Short.

ShortSniffer Troubleshooting Hints

After you have reviewed the Practice Board exercises and watched the videos, a little real-life practice should give you confidence to track down any short circuits you encounter. Here are some hints to remember when your short circuit is making you doubt your abilities:

Most shorted nets are only shorted at a single location. The ShortSniffer should guide you to that location.
For most short circuit location tasks, the ShortSniffer’s built-in speaker may be easier (fewer cables to tangle) than using the headphones.
Complicated short circuit geometries (strange conductive paths, wires, ground planes, etc.) will benefit from using the headphones, with right-left stereo imaging for current flow direction and louder sounds for weaker signals.
If you are receiving confusing sonic feedback from your pickup probe positioning, change your current injection locations to contact points at the other ends of the traces (or other corners of planes) on your shorted nets.
A parallel 1uF capacitor will pass current like a 10-Ohm resistor. If you are having trouble finding the strongest current path on shorted nets with parallel capacitance, remove any caps that are more than 1uF to concentrate the received signals as you follow the sound to the short.
Most PCB or solder shorts are under an Ohm. I had one PCB fabrication short that was 100 Ohms, but it could not be seen without extreme magnification and enhanced lighting conditions. Any shorts greater than an Ohm may not be the real shorted nets, but only electrically close to the shorted nets.
On shorted nets with resistance measurements more than an Ohm, check for resistive – vs. – semiconductor junctions. Change ohmmeter polarity (swap the leads) or change ohmmeter range looking for nonlinear response. If you have nonlinear response to your resistance measurements, check nearby net combinations for the real (lower resistance) shorts.

Troubleshooting Examples –

Troubleshooting process on a batch of production fallouts: This video consists of an introduction explaining the problem of a group of non-functional boards that were received from our assembly house, followed by the process of finding symptoms of the cause, progressing to finding the shorted nets. The video finishes with the ShortSniffer identifying the exact location of the short circuit. These were real PCB faults on “Electrically Tested” boards. (can you spell fraud?) . . . Watch Video on YouTube (7 minutes)

The video shows the ShortSniffer action on a defective PC board. Since there were 8 boards with similar problems, I spent about an hour finding the 8 short circuits, fixed those problems with a microscope and an Exact-o knife, then tested the boards again with 6 boards passing. Repeating the process on the remaining two bad boards, I found two more shorts, which were quickly removed. Here are some pictures of the shorts I found:

Pictures of Shorted traces from this batch of boards (click for larger views)

hard to see, scraped to see copper, cleared with deep scrape

easy to see when you get close and know where to look, cleared with deep scrape

easy to see when you get close and know where to look

Here are some more videos of ShortSniffer action. Watch and Hear as we follow the sound of the current to the short circuit location. The links (below) are MPEG presentations of ShortSniffers in action. They show how rapidly you can pinpoint the location of your shorts when guided by sound.
picture sequence #1 sniffing short #1:
Video Segment 1A
Video Segment 1C
close-up still of Short01
picture sequence #2 sniffing short #2:
Video Segment 2A
Video Segment 2B
Video Segment 2C
Video Segment 2D
Video Segment 2E
Video Segment 2F
Video Segment 2G
Video Segment 2H
close-up still of Short02
Segment #3 is a sequence showing a short sniffed from the opposite side of the board, simulating finding a short under an IC or on an inner board layer.
Video Segment 3A
Video Segment 3B
Video Segment 3C

Note: The previous sets of videos featured shorts that were intentionally placed on the boards to demonstrate the functionality of the ShortSniffer products. Most real world ShortSniffer successes are immediately repaired, without a chance to document them. As impressive “real world” successes are captured on video, they will be added to this page.

Here’s the SS2 in Action with Finding Real World Shorts:

Here are some shots of a 18 layer prototype board that had been electrically tested (good) before assembly. There is a 3.3V short to Ground somewhere on the board. The BGAs were X-Rayed and looked good. Where is the short? This is a job for the SS2! Connecting the current signal into the power pins, I chose the Sensitive probe and waved the probe across the board on a 1/4 inch grid . . . nothing . . ? I soldered some wire leads across a bypass tantalum cap at the other end of the board and clipped the current leads there. Another “grid wave” of the probe showed a “hot” area where the power plane current was concentrating (at the short). I checked the other side of the board and found a stronger signal running under an IC (near pins 31 and 32 which were power and ground)! To narrow the possible area of the short, I tack-soldered a pair of #30 wires to the two pins and connected the SS2 current injection leads to the wires. Changing to the Tiny probe, I found that the current didn’t go past the inside dimensions of the PCB pads for those pins. By rotating the probe 90 degrees, the maximum signal (cross over point) seems to be near the inside third of the pads. The board owner was able to X-Ray the board to verify the position of the short. Check out the X-Ray to see how well I did!
Power-Plane-to-Ground Segment 4A
Power-Plane-to-Ground Segment 4B
Power-Plane-to-Ground Segment 4C
X-Ray of short found under IC

Locating Inner Trace Shorts to Ground Planes
Here are sequences of pictures showing shorts between nets and ground planes. The videos show the process of following the sound to the location of the short, coming from both ends of the shorted net (with respect to the ground plane). The place where the “sound trail” ends is where the current density drops as it goes into the ground plane from the shorted net. By noticing this effect while coming from either end of the shorted net, the location of the short circuit can be found within a few mm. I’ve asked the board owner to X-Ray the boards and send the pictures for this page show what a good job I’ve done (I hope).
Board #1:
First end point to short
Second end point to short
Picture of shorted area with PCB artwork showing inner layer for reference

Board #2:
First end point to short
Second end point to short
Picture of shorted area with PCB artwork showing inner layer for reference
Picture of Gerber file showing 0.005 clearance at area of short

Here are some pictures of a short that was extremely difficult and confusing:

Changing the current injection points provided no additional clue as to the location of the short. Every place I injected the test current showed the current going into the wires, but there was no path over the board to the short. It was acting as if every contact point (through-hole pins, and vias) in the Vcc net was shorted to the GND net. Finally I shined a bright light through the board and noticed an amazing difference between a good board and a bad board.

Good Board and Bad Board (shifted power plane)

Do you have a challenging short circuit finding problem? Let me know the details and maybe I can feature your board here. (you pay postage, both ways)

Finding the most difficult shorts is like solving a puzzle, and I like puzzles. I’d like to convince you that you deserve a ShortSniffer so you can demonstrate your puzzle-solving skills, too.