DRV5056: Power-on time

Hello M.H.,

Thanks for considering to use Texas Instruments.  The conditions are provided to let customers know how we tested our devices for deriving the specification bounds we list.  You can power on with a non zero field present.  I suspect they used 0mT as the condition to isolate the affects of power on and proper biasing of internal stages from any measurement and possible internal filtering of the transduced signals. I will check with the design and validation teams to see if they can provide any more insight on this.

Hello M.H.,

I have not been able to get much insight on why our team defined this specification at 0mT.  However, the design engineer I spoke with does not think a field should change the power-on time.

Hello.

Thank you for answering my question.
I understand that magnetic field does not effect power-on time.

Using DRV5056A4 at 5V Vcc, I am tying to measure the change in the bias magnetic field (75mT bias) caused by the approach of a soft magnetic material.

First, I measured the bias magnetic field without any soft magnetic material.

After power on, the output voltage of the hall sensor increases by approximately 10mV over a few seconds and then stabilizes.
The bias magnetic field is provided by a nearby neodymium magnet, so it is unlikely that the magnet field has changed within a few seconds.

One DRV5056A4DBZR and 0.1μF ceramic bypass capacitor are mounted on a 2-layer FR-4 PCB , no other components.

Vcc is provided by an external regulated power supply.

Excluding the change of power-on time, what could be the possible cause?

Hey M.H.

Do you have a filter on the output?  That could potentially slow down the output.  Also do you current limit the supply?  It is possible that a little more current than the typical operating supply current might be needed at power-up. Yet another possibility though I think very unlikely is a nearby high current trace producing an opposing field. I will check with my team to see if they have any other ideas.

Hello Patrick Simmons.
Thank you for suggesting possible causes.

No output filters were used.
The current limit of the voltage source is 80mA, while the maximum operating supply current of DRV5056 is 10mA.
There would be no high current traces nearby.

I speculated about other causes.
Is it possible that the temperature of the hall sensor slightly increased after it was powered on, lead to its sensitivity to change due to the sensitivity temperature compensation?
If the temperature of the hall sensor takes a few seconds to stabilize, that might the cause of this behavior of the output voltage.

Thank you.

Hello M.H.,

In a scenario where you are heating up both the magnet and device to a significantly higher temperature, and they are not changing temperature at the same rate, then I would expect a delay like this.  However, for power-on at a stable ambient temperature I would not expect such behavior.  How close are your test points to the device?  Could you possibly provide  a picture of the probe connections? 

Hello.

Here is the probe connections.

The blue wire on the right is soldered directly to the output pin and is approximately 90mm long.
The VDD and GND pins are connected to wires through the PCB traces.

The bias magnet is mounted on the back side of the PCB(1mm thick) with a gap of 0.45 mm.

Hey M.H.,

Thank you for providing that.  We are going to try to do some measurements of our own tomorrow and get back to you.

Hello M.H.,

We are having difficulties reproducing what you observed.  In our first test case we power up the device to 5V with no field present.  From that you can see that after VCC reaches 3V, it takes less than 400us to reach a stable vq.

When we provide a field, it does seem to take longer, however it does seem to reach the output value around 400us.

Have you observed this behavior on more than one device?  Could you provide a picture of the part top markings?

Hello M.H.,

Here are some additional captures we were able to collect.

Please let us know if there are any questions regarding the captures.

Best,

Isaac

Hello Patrick and Isaac.

Thank you for trying to reproduce.

I think the problem is that you are measuring the output voltage at 1V/div and I am measuring it at 10mV/div.

I also measured at 1V/div 200μs/div, and like your results, the output voltage appeared to settle within 400μs.

Since I intend to measure small magnetic field changes with a hall sensor, I measured the output voltage near its final value at 10mV/div .

And with that method, it seemed to take several seconds for the output voltage to settle to its final value.

Here is another result with 10mV/div.

A difference of 10mV with an output of 2.5V corresponds to 0.4%. Actually, should such a small difference be considered an error?

In other words, is DRV5056A4 not suitable to measure small changes in a large bias magnetic field(e.g. 0.5mT out of 75mT)?

In addition, I also measured boards with DRV5056A4 and peripheral components such as a LDO and a microcontroller, and they also had similar results.

The part top marking is 56A4.

Hello M.H.

You raise a valid point.  I will see if I can get some more measurements done.  I most likely will not have a chance to check these until sometime tomorrow.

Hi M.H.,

We were able to reproduce the problem you observed, and we're passing it on to the relevant teams. Thanks for your patience.