Teensy 4.0 with Audio Shield Rev D short issue

Status
Not open for further replies.
Greetings PJRC,

My partner and I are trying to complete a project using the PJRC Teensy 4.0 with the Audio Shield Rev D. We initially ordered 20 of each unit. Of those 20, 5 had short issues upon assembly with the header pins. The issue is a short between GND and 3.3V. I have tested with a multi-meter. The boards that have worked have not had continuity between GND and 3.3 pads. I was able to successfully program them, and they work as expected. Tonight, I assembled another Teensy 4.0 and Audio Shield. I tested Teensy and Shield for shorts prior to joining them. I also plugged in the Teensy 4.0 to computer prior to assembly, and it is recognized. However, once I join the 2 boards with the header pins, I have short between GND and 3.3V.

Our source code is the ProffieOS by Fredrik Hubinette with MiCom Prop file. The entire OS can be found here: ProffieOSIt is extensive, and I don't want to post it all here, as its irrelevant to the problem, and when the boards don't have this short, they work fine on the ProffieOS.

Here are some photos of the joined pair:

91zOl40s.jpg
Puwiy4hs.jpg
Topside Image Original
Bottomside Image Original

Any help or guidance is appreciated. I don't understand why some of the boards work fine, and some don't. I do have pretty decent experience in joining Teensy 3.0s with Fredrik Hubinette's Teensy Saber sound shield. Not sure why we are getting this short on some of the boards.

Tom
 
Last edited:
I have similar tight packaging (normal 1mm header) and there is still light between and, no surprise, I have no issue.
 
Most probably the SDcard connector on the AudioShield touches one of the SMD components on the backside of the Teensy 4.0 [C20/C9/C17/C29/L3].

Paul
 
Most probably the SDcard connector on the AudioShield touches one of the SMD components on the backside of the Teensy 4.0 [C20/C9/C17/C29/L3].

Paul

I'm using 2.5mm spacers on the header pins. But yes, SMD components are uncomfortably close to SD. Which header pins are suggested then for joining these 2 soundboards? It looks like the suggested header pins are 2.54mm, mine are more like 2.49mm. Could this be the issue? It does possibly look like there may be SMD contacting the SD.

Tom
 
Last edited:
It looks like the suggested header pins are 2.54mm
This 2.54mm is not the height of the plastic insulator but the pitch between the pins.
The height of the SDcard connector is ~1.7mm, the height of C29 and L3 is ~0.9mm, so you need at least 3mm spacing between the boards to allow for some margin. But it's safer to go for MarkT's adviced header pins when you're not height limited.

Paul
 
Any chance you could slide a piece of paper or other thin insulting material between them?

So, I have used the suggested longer pins to join them, and added insulation. I am still getting a short between 3V and GND on several sandwich combos. I have had some successes, and some failures. I really think there is a defect in some of the traces or something. The components are no longer able to touch from Teensy 4.0 to the Rev D audio shield. Still shorts. I don't understand this.

Iron at 350C Needle, using leaded solder. Time on contact on each pad might be a second to a second and a half. Board does not look scorched.

Tom
 
Probably a long-shot, but take a closer look at the area I have circled in red. That small trace running between the GND pin and Pin 0 is connected to 3.3V. Perhaps the solder resist is getting scratched and you are getting a solder bridge between the ground pin and that trace.

Possible solder short.jpg
 
I don't see any shielding issues on this 3.7 trace. I'm sure its a trace issue, I'm going to start probing these boards to see if I can find the short, cut it out, and bypass it. Its really frustrating, and that trace you show above is dangerously close to the ground pad, not sure why they didn't take it by on of the other pads instead of GND. Seems precarious.

Tom
 
Probably a long-shot, but take a closer look at the area I have circled in red. That small trace running between the GND pin and Pin 0 is connected to 3.3V. Perhaps the solder resist is getting scratched and you are getting a solder bridge between the ground pin and that trace.

View attachment 25536

You can't see it well, but from the photo it looks like there is solder on the trace.
This can of course cause the short circuit.
There seems to a small bridge between GND and the trace.

Also, what is that below PIN 3?
 
You can't see it well, but from the photo it looks like there is solder on the trace.
This can of course cause the short circuit.
There seems to a small bridge between GND and the trace.

Also, what is that below PIN 3?

Should the trace not be covered by the green soldermask?
 
That trace just feeds 3.3V to the optional 25K potentiometer. Since it appears you are not using that part, cutting the trace won't affect your setup. If it is a solder bridge, you should also be able to clear it by putting some flux in that area including the trace and reheating the pin so the solder flows towards it.

These boards are constructed to normal standards and are well packaged by PJRC. I have never seen the solder resist scratched, though it is possible. The spacing between the trace and pin is also pretty standard. My guess is you are using a little too large of soldering iron tip and perhaps moving it back and forth while you solder which is scratching the solder resist allowing the solder to flow over to the trace. The tip should just be rested on the pad while also making contact with the pin while solder is being fed in and then lifted straight off. There is no need to move it around while feeding in the solder.

If you are going to be making more of these in the future, a cheap stereoscopic microscope that costs about $200 can make soldering and inspecting for defects like this a whole lot easier.
 
Status
Not open for further replies.
Back
Top