• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 32 subscribers
  • Views 649 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TSB41A damage on self-powered node

Peter Kazanzides
Prodigy 90 points
Other Parts Discussed in Thread: TSB41AB2

I have a board that uses the TSB41AB2 to provide two IEEE-1394a ports. This is a research "product" that is used at many different Universities. There have been several failures in the field, where one of the two ports stops working.  The failures have the same characteristics as the "late Vg event" failures reported in SLEA072A.

My nodes are self-powered, however. I am using the 6-pin connectors, but do not use Vp on the board (i.e., Vp is only connected to the two connectors).  The design files (Altium format) are here:  https://github.com/jhu-cisst/mechatronics-boards/tree/master/FPGA1394.  There is also a PDF of the schematics.

I do not have the TPx diodes recommended in SLEA072A, but my connector shells are grounded.  As far as I can tell, the only possible failure (in my case) might be if Vp somehow gets on one of the TPx lines. Is this possible?  Are the protection diodes recommended even for a self-powered node?

  • 0
    Guru 243820 points

    SLEA072A says:

    self-powered applications (a system that never sources or sinks current from the 1394 cable) typically do not see damage due to late Vg events

    However, according to the schematics, your board can source current from Vp, so it is not self-powered according to this definition; it is a power provider

    What can happen appears to be exactly what is described: a cable is inserted at an angle so that only Vp and TPx are connected, and due to the device at the other end, your Vp ends up on TPx.

  • 0 in reply to Clemens Ladisch
    Prodigy 90 points

    Thank you for looking into this, but I am not clear on the definition of "power provider".  There is a jumper (J10) on my board (last page of schematics) that prevents Vp from being connected to my power circuitry (assuming the jumper is not installed, which is how I set it up).  Sure, the Vp lines on the two connectors are connected together, so if power comes in on one of them, that power is passed to the second board on my chain (which doesn't use it either).  For example, if I connect 2 of my nodes, N1 and N2, to a PC via 6-pin FireWire cables, F1 and F2, as follows:

    PC ----- (F1)----- N1 ----- (F2)----- N2

    The PC is clearly a power provider (though nobody is using its power). Since the connector Vp lines are connected together on my board, it seems that you consider N1 to also be a power provider (i.e., it passes through the power from the PC).  Then, if I disconnect F1, N1 could no longer be considered a power provider. Based on this, I would not consider my board to be a power provider.

    Anyway, it seems to me that it doesn't really matter whether we call my board a power provider or not. Since I am always connected to a PC, there is power on the Vp line of every cable in my setup. I still do not see any way that Vp could end up on TPx, unless inserting the cable at an angle actually causes the Vp pin to physically touch one of the TPx lines.  Is that possible, and if so, is there any protection against that?  Because my nodes are self-powered, the cable could be attached when my nodes are powered up or not.  The protection diodes shown in SLEA072A connect to GND and ESD_3.3V.  If my node is powered off, there is no ESD_3.3V.

    I guess the bigger question is whether there is any way to protect against a high voltage (such as Vp, which is usually 12V) being connected to any of the TPx lines.  I have seen this happen when someone forced a cable in upside down.  I know it could happen if there is a defective cable (e.g., short from Vp to TPx in the cable).  I asked above whether it can happen in normal use, by inserting the connector at an angle.

  • 0 in reply to Peter Kazanzides
    Guru 243820 points

    Sorry, I overlooked J10.

    Without anything in your node connected to Vp, this cannot happen in normal use.

  • 0 in reply to Clemens Ladisch
    Prodigy 90 points

    Yes, I agree.  I don't think I am susceptible to the late Vg event as described in SLEA072A.  But, I am seeing failures that look a lot like this, and they seem to happen in groups (i.e., more than one board in a single day).  I am suspecting something like a bad cable that damages a few boards before it is discarded, but I have not yet found the culprit.  So far, the failures have happened at other Universities.

    I am planning another build of boards, so now would be a good time to modify my circuit if there is anything I can add to make it more robust. But, I don't want to incur the expense of modifying the board for a change that is unlikely to make a difference.

  • 0 in reply to Peter Kazanzides
    Guru 243820 points

    In the absence of any other information, bad cables or misuse indeed look like the most likely cause.

    In any case, implementing the SLEA072A protections would help against pretty much anything that can damage a single port.

  • 0 in reply to Clemens Ladisch
    TI__Guru* 76130 points

    Thank you for your comments Clemens.

    Regards.