• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 2 answers
  • Subscribers 31 subscribers
  • Views 4067 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.

LDC1614: LDC1314: Frequency reading changes every reboot

dendama-san
Prodigy 120 points
Part Number: LDC1614

This is Ryo from Japan. 

  • Question

Can we use coils which have different inductance and design at LDC1X14? From the point of temperature drift, are there any concern?

  • Device

LDC 1314 or 1614

  • Condition 

Four coil are connected to LDC 1X14 devices. Although three coils are almost same inductance, the other coils's inductance is different from three coils. The design of three coils are circle. However, the shape of the other coils is  stretched rectangular  like A~G below data sheet about reference coils.

http://www.ti.com/lit/ug/snou136/snou136.pdf

  • Problem

I found out the value of oscillation frequency is much more different than I expected, every time the LDC1x14 device is reboot. In paticular, at the channel which have 13.8 MHz oscillation frequency at the stretched rectangular coil, the oscillation frequency ls 1 MHz  different every reboot. 

  • 0
    TI__Intellectual 2725 points
    Hi,

    The maximum valid sensor frequency for the LDC1x1x devices is 10 MHz, I would suggest changing to a higher inductance or sensor cap value to be below this limit.

    Regards,

    Varn.
  • 0 in reply to Varn Khanna
    Prodigy 120 points

    Dear Varn, 

    Thank you for your reply. 

    Even if I set 8 MHz oscillation frequency at stretched rectangular coil, the oscillation frequency is still different about 80 kHz every reboot.

    Best Regards,

    Ryo 

  • 0 in reply to dendama-san
    TI__Genius 13335 points
    Hello,
    Can you confirm what Deglitch filter setting you are using? Note that the default is 3.3MHz so if you are oscillating higher than this it can likely give you strange results. You should either be setting this value to 10MHz or 33MHz for best results. See section 7.3.4 Input Deglitch Filter section of the LDC1614 datasheet.
    Regards,
    Luke LaPointe
  • 0 in reply to Luke LaPointe
    Prodigy 120 points

    10 MHz is set at Deglitch filter.

    I worry about temperature compensation function, because one coil's shape and inductance is different from the other three coils as I explained at first question. 

    >Condition 
    >Four coil are connected to LDC 1X14 devices. Although three coils are almost same inductance, the other coils's inductance is different from three coils. The design of three coils are circle. However, the shape of the other coils is stretched rectangular like A~G below data sheet about reference coils.
    >www.ti.com/.../snou136.pdf

  • 0 in reply to dendama-san
    TI__Genius 13335 points
    Hello,
    Is the concern that a specific coil starts up with a different value every time? Or is the concern that all four channels do not read the exact same value? If it's the latter, then that is expected due to part-to-part matching of discrete components such as the capacitor as well as different coil geometry as you've mentioned. If it is the first issue then we'd like you to provide more information about the sensors such as inductance value, Rp, capacitance, as well as device settings.
    Thanks!
    Luke
  • 0 in reply to Luke LaPointe
    Prodigy 120 points

    All four channels do not read the exact same value. However, it is not exact every time. The data of four channel osssillation frequency is attached in this post. I reboot device every blank of column in the data. The sampling rate is 1 Hz. 

    • Hardware condition 

    channel 0~ 2 :

    shape: circle

    Rp: 12.6Ω
    C: 18pF
    L: 52.6μH

    channel 3 

    shape: stretched rectangular 

    Rp: 4.0Ω
    C: 200pF
    L: 2.8μH

    External  osssillation frequency: 40 MHz

    • Software condition
      • _Offset = 0;
      • _Rcount = 0x2000; 
      • _SettleCount = 50; 
      • _DriveCurrent = 23; 
      • _dividerREF = 2;
      • _dividerIN = 2;
      •  AUTOSCAN_EN, = 1
      • DEGLITCH =  DEGLITCH_10MHz
      • RP_OVERRIDE_EN = 0 
      • SENSOR_ACTIVATE_SEL =  0 
      • AUTO_AMP_DIS = 1 
      • REF_CLK_SRC = 1 
      • INTB_DIS =  0
      • HIGH_CURRENT_DRV = 0 
      •  DRDY_2INT = 1

    7446.ldc_test.pdf

    Best Regards 

  • 0 in reply to dendama-san
    TI__Genius 13335 points

    Hello,

    Thank you for posting your data. It looks like Channels 0 through 2 are jumping about 5kHz or 8kHz between power cycles and Channel 3 is only jumping about 0.17kHz in the worst case. This is likely the effect of having such a small sensor capacitor. As I mentioned in another thread (https://e2e.ti.com/support/sensor/inductive-sensing/f/938/t/623412#2301344), that with small sensor capacitors (<50pF or so), subtle changes in parasitic capacitance can create a noticeable shift in sensor frequency because they represent a bigger portion of the sensor capacitance. For sensors 0 through 2, you have 52.6µH and 18pF which translates to approximately 5.172MHz. If you have even a 0.05pF shift in parasitic capacitance your new frequency could be around 7kHz off, which would match closely to what you are seeing. If you instead increased your capacitance to about 200pF then the same 0.05pF shift would only result in about a 0.19kHz shift. Therefore I would recommend to increase the capacitance on these channels and see if the problem goes away.

    Additionally, can you clarify your sensor Rp for each of the sensors? Did you mean kΩ instead of Ω? Or did you mean the series resistance (Rs) instead of parallel resistance (Rp)? Note that the minimum value for Rp is 1kΩ for multi-channel measurements. A value of 12.6Ω or 4Ω would be much too low for the device. 

    Regards,

    Luke

  • 0 in reply to Luke LaPointe
    Prodigy 120 points

    Hello, 

    Thank you for your advice. 

    >Additionally, can you clarify your sensor Rp for each of the sensors? 

    The Rp value at previous post is incorrect. The correct value is showed as following "Hardware condition".

    >This is likely the effect of having such a small sensor capacitor. 

    As you recommended us,  100 pF capacitor is used instead of 18 pF. However,  jumping effect is still remaining. Please check the measured data attached in this post. 

    • Hardware condition 

    channel 0~ 2 :

    shape: circle

    Rp: 39kΩ
    C: 100pF
    L: 52.2μH

    channel 3: 

    shape: stretched rectangular 

    Rp: 3.7kΩ
    C: 200pF
    L: 2.92μH

    ldc_test_20170916.pdf

  • 0 in reply to dendama-san
    TI__Genius 13335 points
    Hello,
    Can you try to set the following register?
    RP_OVERRIDE_EN = 1
    Thanks!
    Luke
  • 0 in reply to Luke LaPointe
    Prodigy 120 points

    Hello, 

    Thanks for your advice, there is no more jumping effect at every reboot.

    The data is attached in this post.

    ldc_test_20170920.pdf 

    When the RP_OVERRIDE_EN is set 0, does auto calibration work? How is this parameter different from AUTO_AMP_DIS parameter?

    Best Regards, 

    Ryo

  • 0 in reply to dendama-san
    TI__Genius 13335 points
    Hello Ryo,
    I'm glad to hear that setting this register fixed your issue!
    Setting AUTO_AMP_DIS = 1 and RP_OVERRIDE_EN =1 means that the device will power up with the drive current setting that you set (23 in this case) and maintain that value each subsequent oscillation. This is the preferred mode of operation for consistency on each reboot. If you disable the override setting (RP_OVERRIDE_EN =0), then the device will power on and do an initial calibration to determine the best current drive setting and maintain that value for all subsequent oscillations. Because of this the device may pick a different value each reboot and then you can get the small frequency shifts that you were seeing. Additionally if you set AUTO_AMP_DIS = 0 then the device will continuously monitor the oscillation amplitude and try to adjust the current drive setting to stay within 1.2Vp and 1.8Vp so you could get small jumps between conversions if it chooses a new current drive setting. Typically this is why we recommend to set AUTO_AMP_DIS = 1 unless you have a very large movement in metal proximity where the metal can be either very far from the sensor (max sensor amplitude) and very close to the sensor almost touching (minimum sensor amplitude). In this scenario, you may consider using this feature to maintain the oscillation amplitude at a good level throughout the entire range of metal movement. Otherwise setting AUTO_AMP_DIS = 1 and RP_OVERRIDE_EN =1 is the recommended setting.
    Regards,
    Luke LaPointe
  • 0 in reply to Luke LaPointe
    Prodigy 120 points

    Dear Luke, 

    Thank you for your kind answer!

    My understanding about AUTO_AMP_DIS and RP_OVERRIDE_EN parameter become clear. Our team is going to use  LDC1614 or LDC 1314 as a mass production device. 

    If we have another question, please help us like this post. 

    Thanks!

    Best Regards, 

    Ryo