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.

DRV421EVM: Expected noise floor for this chip?

Part Number: DRV421EVM
Other Parts Discussed in Thread: DRV421, DRV425

I'm working with the DRV421 evaluation module, attempting to measure a relatively low current  of ~50 mA. I'm able to read the signal, but there is also significant noise in the system. I have an overall gain of approximately 1, so the 50 mA current results in a 50 mV shift in the output signal. Unfortunately there is also about 50 mV of periodic noise on the signal, so the S/N ratio is rather poor. 

I have read through several other forum posts for this module and I'm aware of the 8 uT offset, and have also developed magnetic shielding for the toroid using a triple layer of MuMetal. I have also built a minimal interface board that brings the output signal into coax within a few mm of the connector to the eval board while still inside the magnetic shielding.

Following is an image of my core and compensation coil inside the magnetic shield. When fully assembled the core and coil are axially centered within the shielding material. 

Since the current I'm trying to measure is small I am using a smaller number (~350) turns on the compensation coil to increase the gain. 

Is it typical to see so much noise from this sensor, or is this a result of a low number of turns that might be marginal on overall inductance? 

Can you suggest any methods for reducing noise on the signal?

Thank you,


Core with coil

  • Hello Nigel,

    For the noise of the flux gate you can expect about 6µT for the full bandwidth.  If you limit the bandwidth you will get better results.  This calculation can be done by using the same method in this post.  The reason that the noise is more than the DRV425 is because the DRV421 has much higher bandwidth than that of the DRV425.  This is also dependent on your Rshunt and compensation coil as this could limit your bandwidth.  Limit your bandwidth could help but if the noise you see is larger than the 6µT peak to peak mentioned above I would look at the inductance.

    The inductance will impact the stability of the feedback and we have some information in this Application Note.  I would suggest the impedance be larger than 0.1H.  If you do not know you inductance you can use the tools provided in this post to simulate your inductance.

    Please let me know if this does not explain what you are seeing or if you need further assistance.