Shop OBEX P1 Docs P2 Docs Learn Events
OK tell me how bad I screwed up — Parallax Forums

OK tell me how bad I screwed up

CRST1CRST1 Posts: 103
in Propeller 2
I had my eval board working in my car and only had to get 5 analog inputs working. I have dividers on three inputs to lower the 12 volt signals. Two only get up to 1 volt from the O2 sensors so no dividers. I was testing the three 12 volt inputs to determine the values for 0 to 12 volt thru the dividers. It was going ok until I accidentally touched one of the straight in pins with 8.5 volts! Smile, saw a spark. Now no P2. The red error light is on by the usb connector.
Give me the bad news. Did I fry the whole chip?
«1

Comments

  • VonSzarvasVonSzarvas Posts: 3,279
    If your lucky, just the IO group. Maybe you could isolate it? Which rev eval board do you have? And which usb connector has the red light?
  • CRST1CRST1 Posts: 103
    It's the pc usb conn. It's the new rev b board. PC can't find the P2 anymore.
  • VonSzarvasVonSzarvas Posts: 3,279
    Try removing all the jumper shunts, from 1v8, 5v ldo headers. then supply power from either usb socket (try both alternately), and measure the voltage at the pins of those two voltage headers?

    do you see 5v and 1v8 in any combination?
  • CRST1CRST1 Posts: 103
    Plugged into pc usb there is 1.8 v and 2.5 on the 5 volt header.
    Plugged into the aux usb there is 1.8 v and 5 on the 5 volt header.
    With it plugged into the aux usb and the pc in the pc usb there is no red error light but it still will not find the P2 chip.
  • cgraceycgracey Posts: 14,133
    This is very sad.

    Ken will respond shortly.
  • CRST1CRST1 Posts: 103
    edited 2019-12-19 20:39
    I had three P1's in the vehicle two of them running the 16 pwm controls for the led lighting, 3 counter inputs and 6 digital inputs and one running the 4D systems display and rewired for the P2 and it was handling everything fine with only two cogs and now this. May have to go back to the P1's.
  • VonSzarvasVonSzarvas Posts: 3,279
    sounds like the usb switch needs replacing. maybe the ftdi driver too. And that's before figuring which IO group is possibly blown and isolating that.

    Would need to pull up the schematic and suggest some other tests from work tomorrow. There might be a way to bypass the fault too... Are you ok doing a bit of soldering?

    BTW... Which country is the eval board in?
  • CRST1CRST1 Posts: 103
    edited 2019-12-19 20:43
    I'm in usa. I've done quite a bit smt soldering but not sure on some of the size that's on this board. I have replaced the P1 chip ok.
  • cgraceycgracey Posts: 14,133
    CRST1 wrote: »
    I had three P1's in the vehicle two of them running the 16 pwm controls for the led lighting and one running the 4D systems display and rewired for the P2 and it was handling everything fine with only two cogs and now this. May have to go back to the P1's.

    We'll get you back up and running ASAP. Call Parallax and talk to Christina in sales (press 1). She is checking to see if we have any more boards now.
  • CRST1CRST1 Posts: 103
    Thanks I will call in a bit.
  • VonSzarvasVonSzarvas Posts: 3,279
    CRST1 wrote: »
    Plugged into pc usb there is 1.8 v and 2.5 on the 5 volt header.
    Plugged into the aux usb there is 1.8 v and 5 on the 5 volt header.
    With it plugged into the aux usb and the pc in the pc usb there is no red error light but it still will not find the P2 chip.

    about this... with the voltage shunts removed you won't find the P2 chip. If your familiar with FT_PROG you could try reading the ftdi chip whilst eval is powered from the aux usb socket. That would determine if that part is working.

    Or you could put the 1v8 and 5v headers back on, power the aux usb, then connect pc usb and try detecting the P2. If that works we can get you a little bypass fix worked out that you might be able to apply yourself.
  • VonSzarvasVonSzarvas Posts: 3,279
    in fact, if that test works, I can tell you how to disable pc-usb power with a software config fix that you'd upload once. that would be a very easy fix! no soldering required :)
  • samuellsamuell Posts: 554
    VonSzarvas wrote: »
    in fact, if that test works, I can tell you how to disable pc-usb power with a software config fix that you'd upload once. that would be a very easy fix! no soldering required :)
    I'm curious. Does the process involve a reflash of the FTDI chip with a different "Firmware", so that the DC-DC converter is disabled?

    Kind regards, Samuel Lourenço
  • VonSzarvasVonSzarvas Posts: 3,279
    correct. Its noted on the schematic Ftdi page. Set one of the Cbus ports to tristate. I can't check now as I'm on the mobile.
  • CRST1CRST1 Posts: 103
    I had put the 1.8 jumper on when I tried to connect to the P2 with the aux port powered. I had FT_PROG at one time. I'll see if I still have it.
    The pc does see the usb port still.
  • VonSzarvasVonSzarvas Posts: 3,279
    edited 2019-12-19 21:13
    you'd need the 5v ldo header connected too, before P2 can be found. but to test the ftdi chip only, with ftprog, you can remove both the 1v8 and 5vldo headers.
  • CRST1CRST1 Posts: 103
    Ok, with the headers on and powered from aux it still cannot fide the P2.
    I found ft_prog and it reads the ftdi chip fine with either connector powering it.
  • CRST1CRST1 Posts: 103
    edited 2019-12-19 21:35
    I just realized ther 5 volt header I pulled was the one for the connector rails.
    I went back and removed the 5v and the 1.8v jumpers and there is 5 volts and 1.8 at the ldo pins with the usb connector powered and no red light with the jumpers removed.
    I guess that looks like the chip is gone, right?
  • VonSzarvasVonSzarvas Posts: 3,279
    edited 2019-12-19 21:41
    Seems some additional measurements will be needed. I will be able to suggest a list of checks tomorrow to narrow down the issue.

    All is not lost yet, and you've proven the Ftdi so that's good!

    Actually... with both the 1v8 and 5vldo jumpers on, and powering from usb-aux, you could try measuring all the 3.3v outputs at the edge headers. I'm supposing you'll find one that's not 3.3v!

    The pins will be labelled Vxxxx, like V0007 for the first header.
  • CRST1CRST1 Posts: 103
    Yep, the one on the set of pin for the one that was overvoltaged is dead.
  • Ken GraceyKen Gracey Posts: 7,386
    CRST1: send me your shipping info and I’ll get you a replacement. Early adopter special treatment is in order; don’t fret any further.

    Ken Gracey
    Kgracey@parallax.com
  • CRST1CRST1 Posts: 103
    Ken Gracey wrote: »
    CRST1: send me your shipping info and I’ll get you a replacement. Early adopter special treatment is in order; don’t fret any further.

    Ken Gracey
    Kgracey@parallax.com

    OK on the way, thanks
  • CRST1CRST1 Posts: 103

    VonSzarvas-- I found the 3.3 regulator for that pin header was shorted. I removed it and now the 5 v and the 1.8v are ok with the jumpers on and no red error led.
    But still no communication with the chip.
  • samuellsamuell Posts: 554
    CRST1 wrote: »
    VonSzarvas-- I found the 3.3 regulator for that pin header was shorted. I removed it and now the 5 v and the 1.8v are ok with the jumpers on and no red error led.
    But still no communication with the chip.
    I would guess that the affected I/O bank is the one that communicates with the FTDI. Which pins are fried?

    Kind regards, Samuel Lourenço
  • CRST1CRST1 Posts: 103
    It the bank P24 to P31. I only had P25 and P27 connected.
  • samuellsamuell Posts: 554
    CRST1 wrote: »
    It the bank P24 to P31. I only had P25 and P27 connected.
    That could have affected the internal oscillator, if VIO_24_34 powers it. That means you don't have clock. Just my two cents.

    Kind regards, Samuel Lourenço
  • Ken GraceyKen Gracey Posts: 7,386
    We're sending CRST1 a replacement board.

    @VonSzarvas and @cgracey - let me know if you want his board back for a F/A.

    Ken Gracey
  • evanhevanh Posts: 15,192
    CRST1 wrote: »
    It the bank P24 to P31. I only had P25 and P27 connected.
    Samuell's right. That corner VIO supplies the clock oscillators. With it shorted, the chip will never start up.

  • VonSzarvasVonSzarvas Posts: 3,279
    Ken Gracey wrote: »
    We're sending CRST1 a replacement board.

    @VonSzarvas and @cgracey - let me know if you want his board back for a F/A.

    Ken Gracey

    Hi @CRST1

    Seems you zapped the board on one of the two most sensitive IO groups!

    Glad that Ken could step in with a replacement. Parallax being AWESOME as always!


    It would be interesting to determine if the P2 survived other than IO's P24 to P27. Seems a very high chance. If possible, would you consider posting the old one back to Parallax, for the attention of Ken ?
    If you do that, I'll share the results of surgery back here!

    Apart from testing all the caps and other things in that voltage path, I'd replace the LDO, cut the VIO trace into the P2 chip at VIO2427, and if the short doesn't remain it should be a usable board.
    The P2 is really well segmented, and can keep going if some IO groups are damaged... except for the clock group VIO2831 and if further programming is required then VIO5663.

    For super stout customer application, maybe P28 to P31 should have isolation or protection added right by the P2, so that a short to those pins cannot blow the clock supply. We'll look at that for the next Eval release; it's a good lesson from this destructive event!

  • Ken GraceyKen Gracey Posts: 7,386
    CRST1:

    Ken Gracey
    Parallax Inc.
    599 Menlo Drive
    Rocklin, CA 95765

Sign In or Register to comment.