USB cable melting and catching on fire.

Home › Forums › TinyG › TinyG Support › USB cable melting and catching on fire.

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • June 29, 2015 at 10:08 pm #7901
    jmjcoke
    Member

    So power is wired correctly, triple checked before power up, powered up board and leds did what wiki said they should, I then plugged in usb and heard my computer make the connection sound, so I started walking over to start following the instructions and the connection ding started going off over and over, I then smelt that smell that all electricians know. I looked over and smoke is blasting out of the enclosure, by the time I get it open the usb cable is dripping melted plastic to the point that I can see the wire connections glowing and then it started a small, localized fire. Uhhhh I have been a field service engineer and controls system engineer for 20 years working on Siemens 840D and Fanuc CNC controls and this is my first USB fire. The board is still working as far as the leds are still on and I did check for voltage and a short at the USB plug.
    I was running a smoothieboard 1 hour prior with the only new thing being the Tinyg board. Thoughts???

    Thoughts…

    June 30, 2015 at 7:49 am #7903
    cmcgrath5035
    Moderator

    My 1 second guess is a foreign voltage offset between tinyG GND and computer GND.
    Since you have plenty of experience, I’ll skip the longer BE CAREFUL lecture.

    I have seen some weird effects of the interconnection of multiple “floating” power supply bricks.

    I think I would start with checking GND to GND potential with both AC and DC voltmeters with no connection between the tinyG and the computer

    June 30, 2015 at 9:13 am #7904
    jmjcoke
    Member

    I am using 2 12V power supplies wired in series, I have made sure they are isolated from ground and they have worked on three different boards
    1. Rambo
    3. Azteeg X3 Pro
    4. Uberclock Smoothieboard

    I did find that there was voltage between the usb cable outer metal plug and the outer metal surrounding the usb on tinyg (12v).
    The board is toast so I will get another and use a regular 24v PS.
    Someone please add this tinyg dislike of series PS to the wiki because if this had taken out my main computer with $5K of software, the tears would flow, and we don’t need that.

    June 30, 2015 at 10:32 am #7905
    cmcgrath5035
    Moderator

    As you describe it, I don’t see how this can be a tinyG issue but I don’t have one in my hand either.

    I believe that the tinyG USB shield is connected to the Gnd lug of the Vmotor input. I’ll say I assume your other three boards were same.

    If I am reading you correctly, you are measuring 12V between the neg side of your series connected 24V (=12V+12V) supply and the chassis of your computer, ?
    If true, why?
    Seems one of your two 12V supplies is therefore referenced to whatever the computer chassis is connected to.

    I am just curious, do you a a PWM spindle supply connected as well?
    I find it very difficult to figure out what ground is there.

    • This reply was modified 9 years, 4 months ago by cmcgrath5035.
    June 30, 2015 at 12:08 pm #7907
    jmjcoke
    Member

    Between the metal shield around the USB port on the tinyg and the shield of the cable plugged into my computer tower had 12V with the cable unplugged. What lead me to check that was the sparks that happened when I tried to plug the usb extension cable into tinyg.
    I ordered a new Tinyg and will use a single 24V 8A PS.The other cards are working fine, I disconnected tinyg and installed smoothieboard to the exact wiring as the tinyg, it is running right now on the series PS. I am not saying it is a tinyg “problem”, just that you had said there has been weirdness with this setup before, and as much as I like spending $160 2 times in one week, others might not.

    June 30, 2015 at 12:15 pm #7908
    jmjcoke
    Member

    Yes I have a spindle drive connected to the PWM.

    June 30, 2015 at 6:24 pm #7909
    cmcgrath5035
    Moderator

    I frequently encounter high impedance chassis to chassis voltage offsets due to (??) leakage to “gnd” in switching regulators and supplies, a good example would be a DVD player and an LCD TV, producing a mini-arc when connecting one chassis to the other via a coax, hdmi or USB. Sounds to me like you encountered a low impedance version.

    I don’t have any experience with the other three boards you are using.

    Do you have the PWM spindle driver output somehow connected to (referenced to) “gnd”?
    Most of the hobbyist cnc systems lack the concept of ground bus management

    If of use, you will find the tinyG schematics here

    July 1, 2015 at 11:05 am #7913
    cmcgrath5035
    Moderator

    I am still curious as to the origin of a 12V offset between tinyG GND and PC chassis.
    The only logical cause I can come up with is that unintentionally the center tap (middle) of your series connected power supplies is referenced to earth (e.g. green wire to the connection in a three wire power cord), but that is not what you decribe.

    Does the offset go away if you disconnect from your PWM controller?

    If so, can you describe what PWM controller and Spindle you are using?
    Do you have a separate PwrSup for the Spindle? What V?

    July 3, 2015 at 3:23 am #7925
    Zootalaws
    Member

    It very much depends on the PSU as to whether it will accept operating in series or parallel.

    Some will work in parallel, some won’t – the same that won’t will work in series, the ones that did work in parallel won’t work in series – some will work in both and some will work in neither.

    It depends 🙂

    One thing that is sure, for it to work it MUST be wired like this: Series PSU

    July 3, 2015 at 7:50 am #7932
    cmcgrath5035
    Moderator

    Thanks for posting the reference.
    Thinking about it for a bit more, I was distracted by the comment that it worked with the other 3 controllers.

    As a general statement, putting two “buck regulated” [transformerless] supplies in series or in parallel is probably not a great idea, there will always be a reference, through the input diode bridge, between the outputs and input AC Line and Neutral.

  • Author
    Posts
Viewing 10 posts - 1 through 10 (of 10 total)
  • You must be logged in to reply to this topic.