• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 34 subscribers
  • Views 1082 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TCA8418E:Key scan problem

toshi-e2e
Genius 4830 points
Other Parts Discussed in Thread: TCA8418, TCA8418E

Hi all,
My customer is faced with the key scan problem as follow;

When a key is pushed down about 10,000 times continuously,  the interrupt doesn't occur any more suddenly,
and even if a key is pushed down after that, interrupt doesn't occur.

When a problem has occurred, a Low pulsed output of COLx of scan stops,
and the information which pushed down a key as a result doesn't seem readable.

After such phenomenon has occurred, TCA8418  normally returns by H/W RESET .

Can I have the advice to settle this phenomenon?

Dumped register data when problem occured.

address read value
00 -
01 19
02 0
03 0
04 0
05 0
06 0
07 0
08 0
09 0
0A 0
0B 0
0C 0
0D 0
0E 0
0F 0
10 0
11 0
12 0
13 0
14 FF
15 FF
16 3
17 0
18 0
19 0
1A 0
1B 0
1C 0
1D FF
1E 7F
1F 0
20 0
21 0
22 0
23 0
24 0
25 0
26 0
27 0
28 0
29 FF
2A 7F
2B 0
2C 0
2D 0
2E 0
2F -

COL3 output wave form:

Regards,
Toshi

  • 0
    TI__Mastermind 23226 points

    Hello Toshi,

    I need to get some additional information about this application and I need to get more details about what your are doing.  What I see on the scope plots looks very concerning.  What is causing the triangle saw tooth part of the waveform? 

    How are you implementing the 10,000 key stroke?  What are the reads and writes you are sending during the testing? If you have code then please send that along too.    I have not seen this waveform before. I would like to understand where it is coming from.  Even the first one has the problem but is less severe and then stabilizes. 

    I would like a schematic so that I can verify hardware is done properly.  If we can't see anything wrong there then we may have to look at the layout.

    -Francis Houde 

  • 0 in reply to fhoude
    Genius 4830 points
    Hi Francis-san,
    thanks for your support.
    I am sorry that I have a limited information from my customer at this time.

    When they confirmed the symptom,
    they found out COL3 was in the state of HiZ, and a triangle waveform appears on the oscilloscope at the state that a finger touches other keys with which COL3 is shared. (The key of a shape is a dome type and is not pushed down.)
    When a finger was separated from the state which touching a key, COL3 become to be low.

    During the idle key scan mode, all COLs are low, aren't these?
    When the error occurs, it seems the COLs are the HiZ state.

    Is there anything for the condition that COLx will be in the state of Hiz?

    Best regards,
    Toshi
  • 0 in reply to toshi-e2e
    Genius 4830 points

    Hi Francis-san,

    I have an additional question.

    Should unused COL or RAW pins be tied Vcc through an external pullup resistor (10 kΩ)?

    Anyway, I have received the schematics from my customer as below.

    And they mention that currently when they press the key more than 90,000 times by hand, the same phenomenon is reproduced.

    I hope this information  helps for your analysis.

    Best regards,
    Toshi
     

  • 0 in reply to toshi-e2e
    TI__Genius 16510 points

    Toshi-san,

    To answer some questions:

    1) During idle key scan mode, all COLs are low? YES, when all keys are idle/not pressed and voltage all rows is high, the TCA8418E will put all COLs into output low mode. During the key scanning process (1 or more keys is being pressed), the TCA8418 will sweep through all COLs by having all but 1 COL be high-Z.

    2) Should unused COL or ROW pins be tied to VCC through an external pull up resistor?  YES, it would be best to use a 10k or 100k pull-up to VCC on unused pins.

    Note about your waveform: COL3, the triangle waveform you note is from when a user touches another key that COL3 shares. Since COL3 is floating in this instance, this might not be a huge concern, because COL is not supposed to be held to any specific voltage, and is not an input. However, it does seem to suggest that you might want to consider some way to prevent a user from touching the actual signal path, creating this noise.

    One thing to try, since this issue seems to be a result of noise getting coupled onto the COL pins during a sweep, I ask that the customer try a pull-up resistor on the COL pins to assist in eliminating noise. I suggest trying a 100k pull-up resistor on each COL pin. By doing this, ICC will increase, but it should help get rid of the noise coupling issue we are seeing.

    The best solution I think is to better isolate the noise from the signal lines, but I think using a 100k pull-up resistor (or 50k) on each COL pin to VCC should help.