• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 31 subscribers
  • Views 3600 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.

DRV2605 Auto Calibration

Scott Valoff
Prodigy 140 points
Other Parts Discussed in Thread: DRV2605

I am using the DRV2605 with the Precision Microdrives C10-100 LAR actuator.  I am having troubles getting the auto calibration to converge.  Each time I try and run auto-cal, I keep getting Diag_Result=1. 

LRA Specs:

Rated Voltage = 2Vrms

Rated Resonant Freq = 175Hz

 

From the LRA specs, I have selected the following settings:

(Address: 0x16) RatedVoltage[7:0] = 0x53h

(Address: 0x17) ODClamp[7:0] = 0x80h

 

Other settings:

(Address: 0x1A) Feedback Control[7:0] = 0xB6h

(Address: 0x1B) Control-1[7:0] = 0x98h

(Address: 0x1D) Control-2[7:0] = 0x80h

 

All other registers are at their defaults.

Before any register values are changed, a 0 is written to 0x01 to take the part out of standby.  After the registers are modified, a value =7 is written to reg 0x01, followed by a value =1 to reg 0x0C to start the calibration.

Afterwards, the  Diag_Result in reg 0x00 is always = 1.  I’m assuming that this means the calibration failed. 

Other questions:

Under the Programming the Rated Voltage section there is a value of 5.28V that is used in equation-2.  Should this number be the supply voltage used, or is this number a constant.  The reason I ask is the DRV2605 is tied to a battery and it will see voltages in the range 4.2V to 3.0V.  The same question goes for the 5.6V used in eq-6. 

  • 0
    TI__Genius 9670 points

    Hi Scott, 

    Thanks for looking at the DRV2605. I double checked your register settings and everything looks okay. 

    When you run auto-calibration how is your actuator mounted?  Typically you will need to mount the actuator to a mass or surface that mimics your final application.  If the actuator is free floating, then the spring-mass inside the actuator and the resulting back-EMF will not reach steady state and the driver will not be able to lock. 

    You can mount the actuator using tape transfer adhesive from 3M, manufacturer part number 3-5-468MP.

    Regarding equation 2, 5.28V is 5.6V (the maximum supply voltage based on the register bits available) divided by a correction factor of 1.06 (6% correction factor).  The correction factor was derived from lab testing.

    5.28V = 5.6V / 1.06% 

    The supply can be any value, but should be high enough to support the output voltage.  3-4.2V will support 2Vrms which it looks like you set in the rated voltage register.

    Thanks,
    Brian

  • 0 in reply to Brian Burk
    Prodigy 140 points

    Hello Brian,

    We have tried calibration with the LRA mounted to a variety of materials.  Currently, I have one epoxied into an SLA housing, and another secured to the TI DRV2605EVK.  In all cases, the auto calibration routine fails. 

    What other trouble shooting steps can I take?

    Is there anything that can be gleaned from looking at the waveforms on the o-scope?

    One difference that I can see is that the period of the waveform on the Eval Board is ~5.7ms (175Hz) with little overshoot.  When I look at the period for our HW, I see 4.3ms and 2.95ms with a fair amount of overshoot.  I'm guessing that the device is not stable.

    Thanks,
    Scott
  • 0 in reply to Scott Valoff
    TI__Genius 9670 points

    Scott,


    Can you please share your scope shots of the DRV2605 output?  Please include OUT+, OUT-, and OUT+ - (OUT-) if possible.

    Also, can you please include the auto-calibration register settings? Registers 0x18, 0x19, and 0x1A after running auto-calibration. 

    In cases where auto-calibration fails, you will see missed pulses in the output waveform. 


    Thanks,
    Brian

  • 0 in reply to Brian Burk
    Prodigy 140 points

    Brian,

    Below are two scope captures.  The first image is the start of auto-calibration for the DRV2605EVK.  The second waveform is from our custom hardware.  Do you know any reason our HW would function differently?  The Out - line on our board does not appear to be driven like the DRV2605EVK.

                                                                 Auto-Calibration DRV2605EVK Board

                                                               Auto-Calibration Custom Hardware

    Register values after calibration:

    Register(0x18): 0x0D

    Register(0x19): 0x6D

    Register(0x1A): 0xA4

     

    Regards,

    Scott

  • 0 in reply to Scott Valoff
    TI__Genius 9670 points

    Hi Scott,

    Okay, thanks. 

    The small triangle waveforms in your scope shots are the back-EMF of the actuator and the tall pulses are the PWM of the output sine wave.  If you were to low-pass filter the waveform than you would see the modulated waveform.

    In your plots it looks like the OUT- is not outputting it's portion of the sine-wave, because there are no "tall" PWM pulses. This indicates the DRV2605 is not locking onto the back-EMF of the actuator.  


    There are two things this is usually caused by:

    1. The actuator is damaged and may need to be replaced.
    2. The actuator and mass are resting on a hard surface (like a table), which is causing the mass to deflect inside the LRA, which in turn makes the back-EMF non-sinusoidal.  If this is the case, try running auto-calibration while holding the actuator and mass in your hand.

    One other option is to lower the rated voltage and then run auto-calibration.  This will reduce the acceleration of the actuator a little and likewise the back-EMF of the actuator.  If it doesn't pass with a lower rated voltage, then it is likely the actuator is damaged.


    I have tested this actuator before and have attached the settings I used.  You may try loading this into the DRV2605 and see if the output looks symmetric driving a waveform.


    Thanks,
    Brian

    DRV2605 Configuration Tool and Design Equations - LRA - C10-100.xlsx
  • 0 in reply to Brian Burk
    Prodigy 140 points

     

    Hello Brian,

    Thanks again for the assistance.  Between the settings you sent out and the removal of an EMI filter on the Out+/Out- lines, I was able to get the auto calibration to pass.  I think the filter network was attenuating the back-EMF signal.

    I appreciate the help,

    Scott

  • 0 in reply to Scott Valoff
    TI__Genius 12485 points

    Scott,

    Just to add, most applications will not require an EMI bead because the DRV2605 has a feature called the Edge Rate Control (ERC) which helps in this situation.

    Regards,

    Gautham

  • 0 in reply to Scott Valoff
    TI__Genius 9670 points

    Scott,

    Great!  I'm glad it calibrates.   Sorry I didn't catch this earlier, but the OUT+ and OUT- test points on the EVM are intended for oscilloscope only.  The filter is used to low-pass filter the 20kHz switching waveform, so that it is "human viewable" on an oscilloscope, meaning that you can see the modulated waveform.

    Glad it works, let me know if we can help with anything else.

    Thanks,
    Brian

  • 0 in reply to Brian Burk
    Prodigy 140 points

    Hello Brian,

    The EMI filter on our board was not taken from the datasheet, or your posting.  It was extra protection we added to the circuit to prevent board noise from being coupled out through the LRA cable.  It is good to hear that the DRV2506 has noise suppression built into the design.  I'm guessing we will not need extra suppression.

    Thanks again,

    Scott