DRV5053: change in sensitivity ?

Hello Akash,

We specify the DRV5053 with a linearity (Le) of 1% typical in datasheet section 6.7 . So technically sensitiviy can change with the magnetic field up to 1% error. The DRV5053 does have an offset. This offset can be realized by noting the quiescent output (VQ) specification of 1.02V typical with 0.9V minimum and 1.15V maximum (also in datasheet section 6.7). So even when no magnetic field is present, the output voltage can start anywhere in between 0.9V to 1.15V and this would be the device output offset.

Between two 5053 sensors the sensitivities can vary according to the the sensitivity maximum and minimum specifications. So two units of the DRV5053PA could potentially have sensitivities that differ by 25mV/mT. This is so since one unit can have a -35mV/mT and the other can have a -10mV/mT.

What is the exact error you are seeing for your experiments with just one single unit under same conditions? What are the testing conditions? How are you generating the magnetic field?

Overall, the DRV5053 is not a precision device. If you need tighter sensitivity specifications, then I recommend using the DRV5055.

Sincerely,
Peter Iliya
Current Sensing Applications

I am generating magnetic field in a high current AC environment ranging from 100A to 8kA. This device is linear which is good but it doesnt have a fixed sensitivity hence its not possible to calibrate. Sensitivity has a drift of 5-7% over time. Also the two hall effect sensors of same specification have an offset between them. It shows very good linearity for AC but not that good for DC. how precise is DRV5055 as compared to DRV5053?

Hi Peter,
As per the graphs in the datasheet, I see the drift in the sensitivity is less than 2% in the temperature I am working. I think its due to the quiescent output as you mentioned before but I dont know what causes it to change. Is it because of some magnetic offset? I dont think DRV425 can be used for very high current like 10kA and more. Also does Rshunt generate reactance in AC magnetic field.

Hey Akash,

I cannot really say why you are seeing this error without more information about your setup, current source, temperature, etc. A couple things I would look into are:
1. The drift of your current source (how much does it drift in amplitude and frequency?). For example is your load heating up and thus causing your current to drift? Is the frequency going past BW of DRV5053?
2. Temperature variations for the DRV5053 and for any other cabling or interfaces being used.
3. The power supply for DRV5053. Is voltage level kept constant throughout experiment? How noisey is the supply voltage?
4. How are you ensuring the DRV5053 is being placed at the exact same distance from current source for all of these measurements? There's a strong inverse polynomial relations (1/distance^3) between field, flux density, and distance.

The DRV425 is a current-output device and the shunt resistor is needed to measure this output current in order to complete feedback operation. Please see the "Theory of Operation" document I sent earlier for more information. Overall, I would recommend using the DRV5055 because this part should absolutely not see this 5-7% drift in sensitivity. Or I would use the DRV425 (which can easily be used to measure 10kA, you just need to place further away) since this will be the most accurate solution.

Best,
Peter Iliya

Hi Peter,

Thanks for your reply. There is no drift in amplitude or frequency. Also not much temperature variation. The position of the sensor is also fixed. Can you please give me more information about the following "There's a strong inverse polynomial relations (1/distance^3) between field, flux density, and distance." And for DRV425 as you say is accurate but it also has an sensing element having some sensitivity factor or gain factor for the detection of magnetic field. I wonder how did you manage to reduce the drift in DRV425 as compared to DRV5053. 

Akash

Hello Akash,

Does your current source have any drift or accuracy specifications? Are you measuring the current with another instrument, to show that is is not drifiting? How fast is the current changing? Over how long of a time is one, singular DRV5053 measuring magnetic field before the sensitivity begins to drift? What happens is you toggle power? Is the sensitivity the same before and after toggling DRV5053 power supply? Do you have any data or experiment set up pictures to share?

Consider using the DRV5055. If you still see 5-7% drift over time, then your set up may be drifting, if not then it is the DRV5053. If the sensitivity of the DRV5053 stops drifting after some amount of time, then your reset power, resume measurements, then you know something is drifting to its steady-state values. One debug test could be to determine the sensitivity of one DRV5053 with a magnet. With this unit and magnet in fixed position, turn the DRV5053 on and off repeatedly and ensure you keep getting the same output reading. This shows that unit's sensitivity does not change with on/off cycles because you are eliminating the effects of DRV5053 self-heating and drift of magnetic source. Now turn on your current source system and let it sit for a long time. Once at steady state, place the same DRV5053 at your current measuring location and turn on the 5053 and record some initial measurement immediately. Then, let DRV5053 sit for a reaonable amount of time and observe if the sensitivity is changing. Turn everything off and let sit for reasonable amount of time. Now turn everthing back on (DRV50553 and current source), take initial measurements and let everything run for reasonable amount of time. Observe if the DRV5053 sensitivity changes over this time similarly to what you have been seeing. This experiment assumes that you are seeing the sensitivity for one unit change over time, but eventually settle to a final value.

Technically I was referring to the magnetic field of a magnetic dipole, you can view a derivation of this here:

Overall, I just to want to make sure that the DRV5053 is placed at same distance every time you record your measurements because a slight change in distance will have large impact on your measurements. For the magnetic field of infinitely long current carrying wire, obviously B = uo*I/(2*pi*r).

The DRV425 works completely differently from the DRV5053 and our other Hall-effect sensors. It is more accurate because there is design work needed in order to use it.

Best,

Peter Iliya