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.

TSC2046: false PENIRQ signal

Part Number: TSC2046

Dear Specialists,

My customer is evaluating TSC2046E.

Even though the touch panel is not pressed, the Lo signal of \ PENIRQ is output, and then the controller continues to be interrupted.

I'd be grateful if you could advise.

---

When I checked the waveform, the X + terminal dropped to about 1.5V after data acquisition (DOUT) on the controller side.

In the standby state (\ CS is Hi), the \ PENIRQ signal becomes Lo, so interrupts are repeatedly applied to the controller side.

After that, the voltage of the X + terminal gradually rises, and the \ PENIRQ signal becomes Hi.

Could you please see the waveform below.

This is a waveform without interrupt.

The schematic is  listed below

---background

When the touch panel is changed, depending on the touch panel, the Lo signal of \ PENIRQ when the touch panel is not pressed does not appear.

When the capacitance between touch panels increases, the Lo signal of \ PENIRQ tends to continue to be output.

As described in the reference material (SBAA155A), even if the capacitors (circuit diagram: C5, C6) of the X + and Y + terminals are removed, the Lo signal of \ PENIRQ will continue to be output and the X + terminal will drop to about 0V. It takes time to return to the Hi signal of \ PENIRQ.

What is the reason why PENIRQ becomes L even though the touch panel is not pressed?

---quesiton and request

Could you please let us know why the phenomen happen and what is the measures.

Our guess is as follows.

The amount and time of the X + voltage drop, the voltage rise time from there, and PENIRQ are linked.

It seems that the X + voltage changes at the timing of enabling PENIRQ after the Idle state before the LSB output of DOUT shown in Fig. 9 of the data sheet.

Is it correct?

---

I appreciate your great help in advance.

Best regards,

Shinichi

  • Hi Shinchi,

    Please describe what commands you are sending to the TSC2046.  Also explain what you mean by 'when touch panel is changed' - different panels have different parameters which will have a bearing on filter caps and surrounding components.

  • Hi Tom,

    Thank you for your reply.

    I'll confirm your question with the customer.

    When I obtain, I feedback to you.

    I appreciate your great help.

    Best regards,

    Shinichi

  • Hi Tom,

    I could obtain the feedback from the customer.

    Could you please see below.

    (1) command

    → In differential / reference mode

    Commands are sent in the order of Y-POSITION → Z1-POSITION → Z2-POSITION → X-POSITION.

    I tried changing the command transmission order for each POSITION, but the results were similar.

    (2) different panel information

    → We confirmed two types of touch panels from different manufacturers.

    1.  X resistance value: 765Ω Y resistance value: 316Ω X + Y-capacity: 8.9nF

    > PENIRQ frequently occurs even though the touch panel is not pressed

    2.  X resistance value: 774Ω Y resistance value: 247Ω X + Y-capacity: 1.4nF

     > PENIRQ output only when the touch panel is pressed

    The circuit is confirmed with the same product by changing only the touch panel.

    This is the circuit diagram sent you before.

    ---

    I appreciate your great help and cooperation.

    Best regards,

    Shinichi

  • Thanks for the additional details Sinchi, we're look over this and get back to you ASAP.

  • Hi Tom,

    Thank you for your reply.

    I'm looking forward to waiting for your advise.

    Best regards,

    Shinichi

  • Hi Sinichi,

    From your waveforms, and given the increased capacitance of the second touch screen, you may have to adjust your timings.  From the original query about the X+ levels and PENIRQ timing, you are correct.  I can't make out any detail of the command/convert cycle from the pictures you provided.  Have you tried any of the schemes mentioned in the TOUCH SCREEN SETTLING section on page 12 of the TSC2046E datasheet?

  • Hi Tom,

    Thank you for your reply.

    I'll share your suggestion with the customer.

    When I obtain the feedback from the customer, I consult you again.

    I appreciate your great help and cooperation.

    Best regards,

    Shinichi

  • Hi Tom,

    I could obtain the feedback from the customer.

    First of all they could resolve the problem.

    They'd like to know the cause of phenomenon.

    Also I could get the answer regarding your question.

    Could you please see below and advise.

    ---

    We have taken two actions.

    (1) Changed to a program that ignores PENIRQ (timing adjustment)

    (2) DCLK is stopped during the acquisition period

    (Your mentioned TOUCH SCREEN SETTLING section on page 12)

    It was confirmed that the interrupt did not continue to be applied.

    Therefore, I think that there is no problem.

    As a result of our investigation, we think that the cause of false PENIRQ signal is the following.

    Could you please advise about the following.

    (1)According to the datasheet in Fig. 9 on page 15,

    when the device is in idle mode, the three internal FETs are turned on to enable PENIRQ( Fig.10 on page 16.)

    In our usage, PD1, PD0 = (0,0) is set, so PENIRQ is enabled

    In this condition, we are guessing.

    If the amount of X + voltage drop differs due to the touch panel parasitic capacitance and the timing

    difference in which the three internal FETs turn on,

    even though I haven't pressed the touch panel, the Lo signal of \ PENIRQ is output.

     Is this correct?

    (2) According to the datasheet P.16 /PENIRQ OUTPUT,

    "It is recommended that the processor mask the interrupt PENIRQ is associated with whenever the processor sends a control byte to the TSC2046. This prevents false triggering of interrupts when the PENIRQ output is disabled in the cases discussed in this section."

    From this content, we thought, "In some cases, the Lo signal of \ PENIRQ may be output even though the touch panel is not pressed."

    Is this supposition correct?

    ---

    I appreciate your great help and cooperation.

    Best regards,

    Shinichi

  • Progress!  That's great Shinichi!  I'll look into your other questions later today.

  • For Q1 above, yes - you are correct.  Your screen captures already imply that the DCLK is stopped during the acquire phase.  You can see that BUSY goes HIGH on the second DCLK cycle, which implies that the first DCLK is applying PD0 to the TSC2046.  You can also see where the driver turns off as BIT1 is clocked out.

    For Q2, again yes. With the higher capacitance value (and slightly higher resistance) of the alternate touch screen, it will take a little more time to have the voltage settle. 

  • Hi Tom,

    Thank you for your reply.

    I'll share your answers with the customer.

    I think they'll deepen the customer's understanding.

    If they have an additional question, I cousult you again.

    I appreciate your great help and cooperation.

    Best regards,

    Shinichi