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.

Failsafe function of SN65HVD102.(CQ voltage)

Other Parts Discussed in Thread: SN65HVD102, SN65HVD101, SN65HVD101EVM

Hello.

Please tell it about failsafe function of SN65HVD102.


My customer checked on an action in CQ voltage = OPEN of SN65HVD102.
Then the High output became Low after that in around five seconds.

Because be written as RX=H in Table 2.(P.12) of the SN65HVD102 data sheet, High should continue being output.
As for this, what kind of state will be correct?

Checked in EVM, but be similar states.
Checked as follows in EVM.
1.Remove R1 from EVM board to make CQ line OPEN.
2.Become RX=High to Low when I turn on power.
3.Make CQ line pull-down, and it is with RX=Low to High
4.Make CQ line OPEN, and keep RX=High for around five seconds, and it is after that with Low.

Best regards.

k-hashi.

  • Hi k-hashi-san,

    I looked at this in the lab today on the SN65HVD101/102 EVM like you and I cannot repeat the same problem that you are seeing on my board. When I have L+ and VCC_IN connected with CQ open like you my RXD output continuously stays high like the datasheet states.

    Can you tell me:
    1. Are you supplying a Vcc voltage to the SN65HVD102 on the VCC_IN pin? The SN65HVD102 does not contain an internal regulator like the SN65HVD101 so the VCC_OUT pin is a NC. This means you have to externally supply VCC_IN from another source.
    2. For item 3 above, how are you pulling the CQ line down? It may be possible that you are loading the CQ pin to much and forcing the part into thermal shutdown.
  • Hello Michael.

    Thank you for your check on EVM.

    I understood a correct function by your answer of RXD.

    I will ask customer to your question.

    Best regards.

    k-hashi.

  • Hello Michael
    I asked a customer your question.

    The answer of the customer
    1. externally supply 3.3V to VCC_IN pin of SN65HVD102.
    2. 100k ohm pulldown the CQ line to GND.

    Furthermore, there is the following comment from a customer.

    ===========================================
    When this test, attach a probe to only the Rx pin. (not attach to CQ line)
    CQ line does not become the open state in the impedance of the probe when I attach a probe to the CQ line.
    When I attached a probe to both RX and CQ line and examined it, the RXD output continuously stays high like the datasheet states.
    ===========================================

    Will you please check once again?

    Best regards.
    k-hashi.
  • Hi k-hashi,

    I doubled checked my setup in the lab today on the EVM and the CQ pin is open. I was not able to reproduce the situation on my board.

    I am going to check with the design team tomorrow on the definition OPEN in the datasheet because to me open looks like open is based on voltage thresholds and not a no load case.

    Is there a possibility that this part was damaged in some way by ESD? If design cannot give us any insight into the issue I suggest that you collect the board from the customer and send it to me and I will send you a new board to deliver to the customer. I would like to evaluate the chip for damage or possible poor soldering. 

  • Hi Michael .

    Thank you for your double check the function.
    And were able to hear the comment from a designer?

    To hear your check details, the customer checked it once again as follows.
    Please tell it if different from your check.

    =====
    board:SN65HVD101EVM 6569277 REV A
    CONDITIONS.
    JMP1: Connect to GND
    JMP2: Connect VCC_IN-VCC_SET
    JMP8: Connect VCC_IN-VCC_OUT
    Removed R1 resistance
    Connect the probe of the oscilloscope to only Rx terminal
    Lp-Ln: Apply DC24V
    VCC_IN: Check on 5V

    Action 1.
    Turn on.
    Became RX=High to Low
    (LED_Rx lighting to lights out)
    Action 2.
    Make CQ line pull-down.
    Became RX=Low to High
    Action 3.
    Make CQ line OPEN.
    Keep RX=High around five seconds; is Low after that.
    Because it has already become High to Low at Action1, expect that the effect of the pull-down has nothing to do.

    In addition, connect a probe or R1 to the CQ line, or becomes Rx=H.
    =====

    Best regards.
    k-hashi.
  • Hi k-hashi,

    I was able to reproduce this in the lab today on the EVM.

    I apologize for not being able to see this before but we were never specific about what to do with the EN pin. During my initial testing I was pulling the EN pin high which makes the driver active and forces the receiver into a deterministic state.

    I was able to reproduce your problem by pull EN low (JMP5) which places the device in a receive only mode. In receive only mode the device uses an internal reference voltage to make a decision about which way to pull the RX pin but any bit of noise or loading on the CQ pin will cause the RX pin to flip one way or the other. For example if you just touch the CQ pin with your finger you will see the RX line go high for a short period of time as well. What we are seeing is the effects of a high impedance impedance and its ability to consume load current. It takes ~5 seconds in the case of the 100k ohm load but it take much shorted in the case of a smaller load (i.e. touching it with your finger).

    You can avoid this behavior by adding a weak pull up or pull down resistor to the CQ line to make the RX output deterministic in nature.

    I hope this helps answer your question. Please let me know if you have anymore questions?
  • Hi Michel,

    Thank you for doing reproduction in EVM.
    We went to a customer this morning, and talked about this movement reproducing.
    The customer was relieved at their boards not being abnormal.

    The customer is curious about follows.
    1. This movement adds a postscript to a data sheet (e.g., at the time of version up)
    2. Is this movement included in the next new product?(or do change it to become function same as a data sheet exactly?)

    Best regards.

    k-hashi.
  • Hi k-hashi,

    I will talk to the systems team about adding a small section in our new device datasheets that discusses this behavior but I am not sure we will go back and update the old datasheet since this is not a simple process. Thanks and let me know if the customer has anymore questions.