Forum Rule: Always post complete source code & details to reproduce any issue!
Results 1 to 17 of 17

Thread: Cutting Vin from Vusb and bricking Teensy 3.2 in the process

  1. #1
    Junior Member
    Join Date
    Aug 2020
    Posts
    19

    Cutting Vin from Vusb and bricking Teensy 3.2 in the process


    ---------------------------------------------------------------------------

    Admin Edit:


    If your Teensy 3.2 is bricked after cutting too deep, info & a solution can be found here:


    https://forum.pjrc.com/threads/65611...l=1#post264557

    ---------------------------------------------------------------------------



    Hi,

    I was about to switch from using my breadboard prototype to a full blown manufactured PCB, and I needed to route the USB power out from Teensy and then back into Vin. So as instructed, I cut the trace on the back of the board, and I soldered an extra pin to the Vusb hole.
    And my Teensy is bricked It looks like I somehow lost the connection from Vin to the voltage regulator.
    Is anyone able tell me how this trace actually goes and when during the process I might have damaged it during the process? I dont seem to have connection between Vin and the input of the regulator nor the decoupling capacitor there, but i'm not quite sure what I might have done wrong (and I want to do it on another Teensy ;-)

  2. #2
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    13,940
    If the right spot only between the two pads by Vin as shown on the rear of the Teensy card was cut that should not be a problem. The trace is Only between those two equal sized pads wiring VUSB to VIN.

    Once that is cut the Teensy can run on external power Only applied to Vin pin (ignoring 3v3 pin). If a USB cable is attached when the Teensy is thus powered - it should still program and communicate normally over USB.

    The only time I had an issue was external USB power for a 32x32 LED matrix that wasn't grounded well back to the supply, when connected to computer USB to program it was feeding over 1 AMP through the USB Shield GND and would be intermittent to program until sufficient GND path to the supply was made.

  3. #3
    Senior Member
    Join Date
    Jul 2014
    Posts
    3,156
    any picture of your cut?

  4. #4
    Junior Member
    Join Date
    Aug 2020
    Posts
    19
    Here's the picture of the cut, I did it along the whole pads.

    Click image for larger version. 

Name:	cut.jpg 
Views:	41 
Size:	148.8 KB 
ID:	21858

    Is the trace connecting Vin to the voltage regulator anywhere there? I clearly must have damaged that somehow, and I'm not really sure how. Ideally i wouldnt want to do the same thing again ;-)

  5. #5
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    13,940
    Is that picture in a mirror or with selfie cam? ... orientation isn't like IRL

    What is the power source back to the Vin pin when that USB trace cut? It is noted that : "route the USB power out from Teensy and then back into Vin"

    Sound like it is still USB powered but taken through other 'unseen' things before it can get back to the Teensy.

  6. #6
    Junior Member
    Join Date
    Aug 2020
    Posts
    19
    The image is mirrored for some reason, I just hooked up the microscope to my PC, I didn't bother with any setup.

    Right now, none if this external circuitry is connected (but basically the USB power goes through a battery charger/power path controller, and then through some regulators, back to Vin pin). All the tests that I'm doing right now are just with Teensy hooked up straight to usb and a cable running from Vusb to Vin.
    It seems to me, that I might have just cut too deep and damaged the trace from Vin to the regulator, but it would be great if someone confirmed that it actually is somewhere there.

  7. #7
    Senior Member+ defragster's Avatar
    Join Date
    Feb 2015
    Posts
    13,940
    The next thing/stuff under that trace is the PCB board layer. That would have to be penetrated well to hit anything that may be under that.

    So at this point perhaps the VIN could be powered from VUSB power removing the external battery and circuitry? Does it work then?

    A DVM on those two pads should show they are not connected and one connects to 5V and the other to VUSB internal pin.

  8. #8
    Junior Member
    Join Date
    Aug 2020
    Posts
    19
    Like I mentioned, my Vin just doesn't get to the on-board voltage regulator.

    Teensy 3.2 is 4 layer board, right? I don't really know how it's routed, I don't know how deep my cut was compared to the layer thickness, but it looks like the input voltage goes to the regulator through that path, originating on the top layer from the marked via, the Vin of the regulator is pin 6

    Click image for larger version. 

Name:	top.png 
Views:	36 
Size:	162.8 KB 
ID:	21859

    it looks a bit like this is the corresponding via on the bottom

    Click image for larger version. 

Name:	bottom.png 
Views:	42 
Size:	137.2 KB 
ID:	21860

    but I dont see any paths from Vin pin leading to that spot, which suggests that this trace is on one of the internal layers. Looking at the board, it wouldn't be crazy to image that it's below the pads from Vusb to Vin, and I might have damaged it.

  9. #9
    Junior Member
    Join Date
    Oct 2020
    Posts
    2
    I have had the same problem. We made a fairly deep cut (about 0.3mm) using a utility knife and bricked 2 Teensies. I probed the input of the regulator and found that it was not connected to the Vin pin.
    Looking under bright light it appears there is a trace in the layer below that comes from the Vin pin, goes under the bridge between Vusb and Vin then heads off towards the regulator. Would like to have this verified and maybe add a warning to only cut lightly to separate Vin and Vusb.

  10. #10
    Junior Member
    Join Date
    Oct 2020
    Posts
    2
    I bridged the via just under the letter A in AGND to Vusb and was able to get it working again.

  11. #11
    After cutting hundreds of Teensy 3.2 for my stuff successfully, with the latest batch of Teensy I got there suddenly was several non responding ones. I started checking them before and during build process, and found it was when cutting the trace they stopped working. Looking further I found the connection between Vin and the voltage regulator input was lost on all of them. Have I suddenly started to cut too deep, or has the routing of the trace to the voltage regulator been changed lately? Hopefully I can un-brick them with the procedure above, but I find this a bit worrying.

  12. #12
    Senior Member PaulStoffregen's Avatar
    Join Date
    Nov 2012
    Posts
    24,108
    First, the easy question:

    Quote Originally Posted by trasselfrisyr View Post
    or has the routing of the trace to the voltage regulator been changed lately?
    Teensy 3.2 PCB layout has been exactly the same for many years.


    Quote Originally Posted by trasselfrisyr View Post
    Have I suddenly started to cut too deep
    However, we have used several different PCB vendors. The distance between the layers may have changed. They are all supposed to be using FR4 Tg170 material, but we don't have any way to check which material they actually used.

    A couple years ago we started having all the boards made with a little 6-bit code on the silkscreen, between the USB connector and pin 1. Those 6 bits represent which PCB vendor made the bare boards. If you see any correlation between those 6 bit codes and the failures, I'd really like to know!

  13. #13
    Quote Originally Posted by PaulStoffregen View Post
    Teensy 3.2 PCB layout has been exactly the same for many years.

    However, we have used several different PCB vendors. The distance between the layers may have changed. They are all supposed to be using FR4 Tg170 material, but we don't have any way to check which material they actually used.

    A couple years ago we started having all the boards made with a little 6-bit code on the silkscreen, between the USB connector and pin 1. Those 6 bits represent which PCB vendor made the bare boards. If you see any correlation between those 6 bit codes and the failures, I'd really like to know!
    Thank you, Paul! This makes sense.

    I checked the failed boards for the code, and they all had this one:

    Click image for larger version. 

Name:	IMG_0452.jpg 
Views:	29 
Size:	93.1 KB 
ID:	22847

  14. #14
    Deleted User
    Guest
    I had one 3.2 with the same feature- unfortunately already gutted. Obviously bad PCB. Anyway, it is not good practice to route a signal 1 or 2 layers under a cutting jumper, as nowadays insulation layers can become extremely thin, if a manufacturer / PCB pool uses the 16L process even for a 4L booard to simplify flow.

    Then there was this T4.0 a year ago that had those unexplainable issues until it got re-fluxed and re-reflowed. Bitcode unknown.
    Others on the forum with T4.0 and intermittent mysteries, looked like the same bad soldering of the BGA.

    Looks like there is a strait of quality problems. QC?

  15. #15
    Just replying to this thread to link in a thread I started, not realizing this one existed.

    Paul request; Separate Vin/VUSB Teensy 3.2

  16. #16
    Quote Originally Posted by gwideman View Post
    Just replying to this thread to link in a thread I started, not realizing this one existed.

    Paul request; Separate Vin/VUSB Teensy 3.2
    ... and that other thread now has images of the PCB layers involved, and schematic detailing the exact damage caused.

  17. #17
    Administrator Paul's Avatar
    Join Date
    Oct 2012
    Posts
    391

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •