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.

DRV421: About the demag cycle

Part Number: DRV421
Other Parts Discussed in Thread: DRV401

I found DRV421 demagnetization is different from DRV401,  the waveform of  VIcomp in DRV401 is  invariable, it doesn't change with the demag current .

when the demag current is large (Rshunt is small enough  such as 50ohm), drv421 will skip  demagnetization without any waveforms in VICOMP . Is it due to the over saturation of the fluxgate sensor beacuse of

the larger demag current?  as below

Therefore, the magnetization current is difficult to control to ensure the demagnetization is correct

my core parameters: 

magnetic Gain =800-1000uT/A

N=1000 turns

air gap =0.3mm

L=200mH

  • Hello User4308845,

    If the sensor/core is saturated for any reason, be it core magnetization or large primary current, then the “smart” demagnetization will not function properly because it requires feedback from the saturated sensor to evaluate which way to apply current to the compensation coil. The search function can help to try to identify the direction and intensity of the saturation prior to demag if it is not beyond the limits of the compensation coil field generation. The DRV401 has a different demag cycle that doesn’t evaluate the offset of the core/sensor; it just forces current back and forth in a fixed pattern to demag the core.
  • Hi Patrick

    How should I Demag the core correctly when Rshunt is large such as 1k ohm ? the parameters of the core is as above.
  • Hello User4308845,

    I would consider an external H-bridge circuit with a microcontroller to create a PWM signal that decays in pulse width. We don't have an example circuit to share at this time but we are investigating it. The idea would be to create a waveform like plot C in figure 7 of the DRV401 data sheet. When we have a working circuit, we will update the DRV421 data sheet. I encourage you to go to the product folder (www.ti.com/.../drv421) and click the "Alert me" button to get updates when documentation on this product changes.
  • Hi Patrick
    Can you describe the whole process of the unique closed-loop demagnetization in DRV421?
  • The demagnetization is supposed to operate as follows:
    1. Device measures the field from pre-magnetized ferromagnetic core.
    2. If the core is saturated, the device skips demagnetization. Otherwise the device continues the demagnetization.
    3. Device applies full scale current in secondary. This current is applied for a few tens of ms.
    4. Device applies 0V to both IComp pins to discharge secondary.
    5. Device measures field.
    6. If the field now in a different direction, next pulse of current will be in opposite direction. Otherwise the next pulse is in the same direction. The duration of the pulse should be shorter than the previous.
    7. Steps 4 through 6 are repeated until 500ms are reached.

    If you look at figure 61 in the datasheet, you can see the waveforms corresponding to the current pulses and the vout from the device. For the Time axis label, the units should be (s).
  • Now I have two questions

    1.   Device applies full scale current in secondary. This current is applied for a few tens of ms. 

    ==> What is the value of this current?   When Rshunt is large ,the DEMAG waveform  is wrong, the pulse of current  is all   in the same direction.

    2.  If the field now in a different direction, next pulse of current will be in opposite direction. Otherwise the next pulse is in the same direction. The duration of the pulse should be shorter than the previous.

    ==> How to judge the direction of the field in different directions

  • Hello user4308845,

    1.  With a large Rshunt you are limiting the current across the compensation coil.  During Demag the Icomp pin drives to the full scale voltage and the current (limited to 210mA @5V VDD) determine by the Rshunt and the Impedance of the compensation coil.  If the coil does not saturate with this current then it will not get magnetized in the direction we are expecting.  If that is the case it will continue to drive the current in the same direction.

    2.  The internal flux-gate sensor makes that determination of the field.  If the current is not enough to saturate the core it will never get magnetized in the direction it is driving.  

    This process will continue to decay bring the core magnetization close to zero.  If the current is not enough it will never change any properties of the compensation core.

    The DRV401 does not use the sensor during the demag.

  • ==>  " If the core is saturated, the device skips demagnetization. Otherwise the device continues the demagnetization."

    this is from Patrick Simmons. 

    but you think:  The internal flux-gate sensor makes that determination of the field.  If the current is not enough to saturate the core it will never get magnetized in the direction it is driving.

    and  the compensation loop is active after the demagnetization cycle which is pointed out in the specification,if  the core is saturated, The internal flux-gate sensor is also  saturated,  the device will

    skip demagnetization. 

    I use this demag circuit as follows.   The degaussing current is adjusted by changing the value of R32. 

    when R32 is 20ohm  ,Rshunt=R6=200ohm,N=200,the ICOMP waveform is as follows,  

    But the final output voltage drift  is still exist ,  about  20mV  , .

  • Hello,

    With 200 turns you may not have the recommend inductance on the compensation core. Also keep in mind the specification of 8µT is something you will not be able to completely remove.

    If your core gain is 800µT/A at our sensor location. The error on the primary current measurement will calculate as follows.

    8µT/(800µT/A) = 10mA (Primary current error)
    This they calculates as 10mA/200 = 50µA on compensation coil
    With Rshunt = 200Ω there is a 10mV at the input of the differential amplifier
    Therefore after the differential amplifier you can get 40mV.

    Another note:be careful using Schottky diodes as they will limit the measurement range of your and create errors at larger voltages. If the diodes start to conduct at 200mV and because you have two across your Rshunt you will be limited at or below 400mV at the input of the differential amplifier and then limited to 1.6V at the output. Any leakage through diodes will result in an error.
  • Hello user4308845,

    We have not heard back from you in awhile. As such we will close this thread. However, if you need further support please respond below or start a new thread.
  • when N=1000, Rshunt=1k  ,G=800uT/A,  the differential amplifier  can  also get 40mV theoretically ,but  actually about 200mv .

    I doubt  the demag cycle is still error . 

  • Hello user 4308845,

    Can you please check if the degauss is the issue with your core by running the demag with 10Ω or 50Ω as I believe from you previous post that it worked some where in that reagion.  Then change the Rshunt back to 1KΩ.  You can do this by placing an alternate path above your 1k then removing it after the demag function.  For the diagrams you had sent the demag seems to work only with lower Rshunts and the bypass with diodes is not working with the 1K. 

     

  • Hello user4308845,

    Did you have a chance to perform the tests Javier suggested above?
  • Hi

    when I use 1000 turns ,and Rshunt=1k, if the core gain is 800uT/A according to the above circuit.   so the differential amplifier I can get is also 40mV.

    but actually the error voltage is close to 200mV .     Why?

  • Hello,

    The degauss operation provides pulses of current through the secondary, with the intention that each pulse will magnetize the core to the opposite polarity and that each polarity reversal will bring the offset closer to 0. In reality it will never reach 0, but it will converge to a much smaller value than what would be present without the degauss.

    In order for this degauss operation to function properly, each pulse from the DRV421 needs to be able to generate a sufficient magnetic field to overcome the residual core magnetization and hysteresis. The magnetic field generated from the core is related directly to the current. The current through the inductor does not instantly change. The rate of change is related to the voltage across the coil, the resistance in series with the coil, and the inductance of the coil. By choosing a large series resistance you effectively speed up the rate at which the current can change. However, you also reduce the maximum B-field the secondary can generate because you have further limited the current. As the device has a fixed degauss duration, it may never actually generate the required B-field necessary to reverse the polarity depending on the charge rate. Alternatively the charge rate may be sufficiently fast, however the current max may not be sufficient to affect the core magnetization.

    The diagram below illustrates the current profile through the inductor over 1ms. Note that the total degauss duration is .5ms with the degauss pulses much smaller than that. Also note that the current is far below the IIcomp typical peak current.

    So this is why Javier recommended you change the resistance. This potentially allows for you to achieve a higher current through the secondary that may be able to generate a magnetic field sufficiently high to overcome the magnetization of the core and the hysteresis.