Forum Rule: Always post complete source code & details to reproduce any issue!
Results 1 to 16 of 16

Thread: Custom Teensy Board

  1. #1

    Custom Teensy Board

    Hi all! So this will probably be a bit painful as this is not my field of expertise. I'm a composer for film, tv, and video games with a side hobby in programming. I've been working with a PCB designer to build a MIDI controller based on my specs for a few months now and today I received the boards for testing. (All code and concept was built, tested, and approved on breadboards and using a Teensy 3.2). When I plug in the usb cable to our custom teensy board, nothing happens. No sound and the code compiler can't find anywhere to put the code. I'm not 100% sure what questions to ask so I can provide some pictures and the purpose of the project.

    The project is very simple - just a 16 linear pot design to send 16 different MIDI messages across the board. So I'm only using power, ground, and analog input A0-A15. The concept on breadboard took next to no time to design and code.

    Click image for larger version. 

Name:	FullSizeRender.jpg 
Views:	227 
Size:	119.3 KB 
ID:	10785

    Any idea on why this board won't even power on? The PCB designer made a spot at the bottom for "reset" and "prog" buttons, but they don't exist on the actual board as I didn't want potential customers to be able to open the box and break the system.
    Last edited by Casey Edwards; 06-19-2017 at 05:24 AM.

  2. #2
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    First question would be if this has a PJRC bootloader IC
    https://www.pjrc.com/store/ic_mkl02.html

    or a blank one?

    Down the bottom of that page is the list of possible causes of dead custom Teensys that may be worth a check for the common problems with Vbat and crystal setup.

  3. #3
    I did inform the PCB designer that it had to have a bootloader from the PJRC web-store in order to function as I needed it to. I'll also pass along the info you sent to me. Thanks!

  4. #4
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    Also, you will need to short the program pins at least once to get the bootloader onto the main chip. You don't need to fit a button but will need some method to short those pins while connected to usb.

  5. #5
    Gremlin, is there a way you could explain a way to do this to someone as inexperienced as myself? I'm not familiar with the process.

  6. #6
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    Really basic method is to plug it into USB, then short the program terminals with a wire, paper clip or pretty much anything conductive. All being well your PC will then acknowledge there is a new USB device and you'll be able to load your code.

    If you have the parts a better method would be to solder an actual button onto the two pads for programming since shorting things by hand always has a risk you'll slip and damage something but should be ok to prove that you can at least get USB enumeration going.

    Does the PCB have a ground plain, or is it mostly empty space on the back? Clock crystal may not be behaving well without it.
    Edit: disregard ground plain question since despite the silkscreen markings on the RTC battery it looks like it's designed with ground on both sides of the PCB with plenty of vias tying them together
    Last edited by GremlinWrangler; 06-11-2017 at 02:17 AM.

  7. #7
    I'll try to figure what the "program terminals" are and do what you suggest. I really appreciate the help! And I see you edited your post, but just in case, here is the flip side of the board.

    Click image for larger version. 

Name:	FullSizeRender2.jpg 
Views:	114 
Size:	119.3 KB 
ID:	10786

  8. #8
    It worked!!! Thank you so much for that little tip. I'll pass the info along to the pcb designer for future reference. Hopefully I won't have to return.

  9. #9
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    Glad to hear it worked and sorry for not just saying 'connect where the program button would go'. Depending on how many of these you are building you may want to consider some form of jig to trigger the first program cycle, the normal method would involve pogo pins

    https://www.adafruit.com/product/394

    either as some form of hand held device (a handle with two pins glued to the end linked by a wire) or possibly in some more elaborate setup
    https://learn.adafruit.com/fiddy-ftdi-pogo-pin-clip
    https://learn.adafruit.com/how-to-ma...o-pin-test-jig
    depending on how many you are looking to ship and how long it takes per unit to program and setup. Suspect the hand held version will be the winner of the lowest total cost math unless you are shipping thousands of these and a new spin of the PCB with pads for program, USB and power to mate with a test jig starts to make sense.

  10. #10
    Hey, no worries! I appreciate the help and I'm glad my feeble brain was able to sort it out within seconds of reading your post.

    Unfortunately, I'm back with other problems now that maybe can have some light shed on. The 2 10pin connectors at the top are for individual MIDI fader boards. Each board contains 8 faders, so this control board is for 16 total. The 2 additional pins per connection are for ground and power. Today I received my ribbon cables to finally give this thing the full test. First I plugged in my control board again and the usb sound chimed like a charm thanks to your help! Program was running and confirmed through a MIDI data program. Now it was time to connect the fader boards one at a time to see what would happen. Unfortunately, as soon as the connection is made the usb immediately disconnects, but will reconnect as soon as the ribbon cable is removed. The worst part about this is one of the parts gets VERY VERY hot and it happens very quickly. It's the piece just above the USB plug and to the left of the first 10pin connector. I will post a schematic of the fader board as soon as it's in my possession. Thoughts on what in the world is happening?
    Last edited by Casey Edwards; 06-13-2017 at 01:41 AM.

  11. #11
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    That's the power regulator, so almost certainly you have some sort of miswire in the header to fader arrangement that connects power to gnd and produces a short circuit. Don't keep playing with that till it's fixed since it's a good way to kill things

    Is it possible that the plugs at one end got fliped 180?

    Other more tedious option is that one end was wired
    1 2 3 4 5
    6 7 8 9 10

    and the other is
    1 3 5 7 9
    2 4 6 8 10

    Final option is that a fader is wired with power on the slider, and that slider is at one end.

    Suggest some work with a multimeter may be required here, looking for how power and gnd pins on the Teensy header are getting something less than 10 ohms across them

  12. #12
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    Looking at the photo's power and gnd are on the left hand end pins of both plugs when the PCB is sitting in front of you with text readable. Next trick is to follow where those two wires to go in the cable (which if really badly assembled might short, those wires are side by side) and the fader desk

  13. #13
    Senior Member Constantin's Avatar
    Join Date
    Nov 2012
    Location
    In the yard with a 17' Dia. Ferris Wheel
    Posts
    1,408
    Quote Originally Posted by Casey Edwards View Post
    Thoughts on what in the world is happening?
    I would check the ribbon connectors carefully. My guess is that one connector is put on with the correct polarity, the other one in reverse, sending power where it's not supposed to go. I would consider using female-female jumper cables (i.e. individual pin to pin connections) and see if that solves the problem.

  14. #14
    I scoured my emails and found this picture that he sent to me for the fader boards.

    Attachment 10801

    I'm starting to think there was miscommunication involved, because my documents and 3d mockups to him show 1 10pin header on each board with a total of 2 boards being combined. This picture shows 2 10pin headers per board and the line running doesn't look correct to me. We're using single gang, so ignore the dual-gang lines ran. Per fader connection, it should be Top right = analog, just to the left of that should be 3.3V, and the bottom right should be GND. I don't even see a GND line running on the fader board or this picture. Am I missing something in this drawing that a schematic would show? I also sent the following picture over along with a description of how the control board works with the fader boards.

    Attachment 10802
    Last edited by Casey Edwards; 06-20-2017 at 06:05 PM.

  15. #15
    Senior Member
    Join Date
    Apr 2013
    Posts
    1,885
    Concur that the gnds don't seem to be doing what you would expect there (may be going through the unseen fill, note the connecting vias scattered around), still unclear how that would make a short circuit but it's also very hard to see the two layers on the B&W drawing. Best guess at the moment is that 3.3V is going to a pin on the fader that is part of the metal shell, which is earthed at the other end.

    It may be possible to salvage a prototype at least by skipping the connecter and soldering the 10 wires directly to the faders but think you need a respin of the board once you've better sorted out what it's supposed to do.

  16. #16
    As a follow up, I did a series of tests using the PJRC Teensy 3.2 and my original breadboard design with the custom control board and fader boards to isolate the problem. Turns out it was the fader board, so I passed this info to the PCB designer, and he found his mistake. Much thanks to everyone who chimed in!

Posting Permissions

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