DRV421EVM: High DC accuracy ratio error

Hi Yutong,

Thanks for the question and using the E2E forum.

We're looking into this and will give you an update when we return from the holiday week.

Louis

Hello Yutong,

I have a couple of questions.  Where are the primary winding located?  I see your calculations don't take the 36 windings into account?  Are your 2000 windings distributed evenly around the core?  What is the tolerance of the 1Ω resistor you are using?  Why is your core so large?  Is your primary current a large conductors.  85mm seems large but you may have a reason for it.  

The error in placement of the core should not cause much error but I would assume about 0.33% based on my simulations using permeability of 14000.  I could not find the your material's permeability but my simulation came close to your gain and inductance.  So this should not account for much error.

 Regards,

Javier

Hello Javier,

thank you for quick respond. Here are my answers to your questions.

• The primary windings are wound homogenically around the core, like in this picture. We would like to have a rated primary current round about 300A ~ 500A.
• I do take the 36 Windings into account, because our Amplifier can only generate a max. current of 25A, and the currents in my table (from -360A and 360A) are the currents already multiplied by the number of windings.
• The resistor is not the problem, it is very precise.
• The core size is defined by our core calculate engineer.

It seems that the primary windings have big impact on the measurement and could be the reason that causing the error. Can you tell me if it is allowed to build the primary conductor as Ampere Windings like in this picture? Does the chip have problem to detect or work with magnetfeld from this kind of windings?

Thank you in advance!

Best regards

Yutong

Hello Yutong,

Your windings look good and that is how I would have done it.  I do not think that the error is coming from the primary windings.  

Yutong Ren said:

Can you tell me if it is allowed to build the primary conductor as Ampere Windings like in this picture? Does the chip have problem to detect or work with magnetfeld from this kind of windings?

Because you have the winding evenly distributed and a core that has high permeability the fields would be coupled through the core and should not be an issue.  Therefore the coupling to the DRV421 should go through the core as well.

I don't see why you should have such a larger error. Below are some other things to look at.

• Maybe look at how you are measuring your primary current.  Normally sources do not have good accuracy.  If you are using an external shunt or another measurement then this should be fine.  
• Another thing is the magnetic coupling to the DRV421, could you show an image on the DRV421 inserted to your airgap?

Regards,

Javier

Hello Javier,

yes, we used an external primary shunt to measure the current. We believe that the problem is caused by the primary windings. Yesterday we managed another Amplifier which can produce over 300A, and thus we don't need 36 Windings, instead, only one primary conductor, and the test results show the output Voltage from the Chip was without error, verified.

However we still cannot understand the problem of primary windings. 

What's the difference between producing 10A in 36 Windings, and producing 360A in one single primary conductor? Could that be different kind of magnetic fields near the air gap and the chip area which cause the problem?

Because in the application note of the DRV421EVM, primary current is also through only one conductor, we guess maybe this chip is not working with ampere windings?

Thanks and regards

Yutong

Hello Yutong,

This is the magnetic coupling should not matter either with windings or single primary.  

Yutong Ren said:

What's the difference between producing 10A in 36 Windings, and producing 360A in one single primary conductor?

There should not be a difference in this case and the magnetics should be the same.

Yutong Ren said:

Could that be different kind of magnetic fields near the air gap and the chip area which cause the problem?

This could be but with a decent permeability and your pictured setup I don't think it will be much.  The only issue that is possible is the Rshunt being placed close to the inside of the core.  

Yutong Ren said:

Because in the application note of the DRV421EVM, primary current is also through only one conductor, we guess maybe this chip is not working with ampere windings?

This should not be the case.  I don't understand why you are having issues and I have done this exact experiment when testing about 1000A.  

Regards,

Javier

Hello Javier,

thanks a lot for the answer. For further tests we only used one single primary conductor and with the original DRV421EVM Board, without any changes. We got some good results with the ratio, but only in one single current direction, the results in other direction were still bad. The error is calculated between the measured value of primary and measured output voltage, which can be converted into the secondary current. It is weird because only one direction is good, and the error is always bigger at small current area.

We would like to send you some of our cores samples with compensation windings to your address, kindly request that you can help us test out our cores with your DRV421-Board, equippment and settings. It is hard to believe that this original board and chip cannot work properly with our cores, because our next step is actually develop a new product for the company with the board. Could you kindly give us a shipping address with your cost estimation on us? 

Best regards

Yutong

Hello Yutong,

Following up offline.

Please see private message sent.