Difference between Band Width and Response time for Signal Conditioners

[David_Troike]

I have a USB-6251 and a program to read 8 differental analog inputs.   I am collecting voltage and current over a 20 second time period.  

Channel 0 has a Narrow Band (4HZ) ±40V DC input to ±10V DC output signal condition module. 

Channel 1 has a Narrow Band (4HZ) ±100mV DC input to ±10V DC output signal condition module.  This channel reads a 50A/50mV shunt for current.

The sample time for these two channels was set to 500HZ. 

I did not get the expected waveform as it looks "filtered".  It completely missed in the inrush spike of the motor .

I believe I need to change the module to a Wide Band (10KHZ) signal condition module but I am confusing my self between bandwidth and response time.

The red waveform was gathered by applying the 0-50mV signal dirrectly to AI1/AI9. It appears to be noisy but maybe that is what it is suppose to see.  The blue waveform is the ±10V DC output of  the ±50mV waveform through a narrow band signal conditioner (4hz).

[johnsold]

Thank you for providing enough information that we can see what your issue is.

The filtered waveform is about what I would expect from a 4 Hz filter. It did not miss the spike at the leading edge - it filtered it.

In fact I wonder if you do not have a spike at all but have AC coupling somewhere in the system. The rising spike at the beginning and the falling spike at the end look more like AC coupling of a square pulse.  The inrush current waveform for a motor does not usually rise vertically like your data - the inductance of the motor windings slows the rise time.

The part you are calling "noisy" appears to be quite periodic - possibly 50 Hz. Are you in a part of the world where the power line frequency is 50 Hz?

What is the rating of the motor and applied voltage?

To answer the implied question: Bandwidth and response time have an inverse relationship. The exact relationship depends on the type and order of the filter so it is not simply a reciprocal relationship.  

The Nyquist criterion applies to any sampled data system. The sampling rate needs to be greater than twice the bandwidth of the signal (including noise and interferring signals). So, you could sample the output of the 4 Hz filter at any rate greater than 8 Hz. To properly measure the output of the wideband filter at 10 kHz you should sample at a rate of more than 20 kHz.

Lynn

[David_Troike]

Thank you for the very helpful explanation.

The motor is an 18V DC permanent magnet motor from a portable power tool.   The power source is an 18V Li-Ion battery back.  The motor does have some current limiting protection from the tool controller.  There is no speed control associated with this motor.  When the trigger is pulled it applies full voltage to the motor.

I never noticed the "noise" has a constant period to you pointed it out.   The testing is done in the US with a 60Hz power line frequency. The signal conditioning modules are Dataforth SCM5B that provides full isolation between the measured waveform and the output waveform.

I would assume some noise on the waveform is created by the switching of the poles on the commutator of the motor but at a much higher frequency.

I am not sure why the negative spike at the end of the waveform because there was counter EMF protection on the motor.   I did not open up the tool to verify it was actually present yet.

This waveform does represent the tool performing work.   When the tool is first started there is very little load on the motor.  The inrush spike is typical as to what we have seen over the last few years for this family of tools.

Towards the end of the cycle the tool current will rise until the tool's applied force is reached.  When the applied force is reached the motor controller removes the power applied to the motor.  The cycle is reset when the user releases the tool trigger. 

 Thank You again for you assistance!

[Henrik_Volkers]

But the red and the blue signal are from two different runs?

and the negative current peak is due to active break.... or a shortcut of the motor protection circuit and the back EMF of the motor attached inertial masses ??

[David_Troike]

Yes the red and blue are from two different runs.   There should be a snubber accross the terminals of the motor to surpress the back EMF.  On some of the tools the BACK EMF is applied to the motor to "brake" it to a stop.  If this is the case then I would expect back EMF causing the negative spike.

[johnsold]

Your periodic fluctuations could be due to torque variations.  The frequency could be related to gears. Many DC motor driven tools use gears for speed reduction and a variation in the instantaneous torque at the gear frequency is not uncommon. 

Without seeing the raw data it is hard to determine what is going on but my guess would be that looking at the inrush with a substantially higher sampling rate would show that the current does not instantaneously rise to 60 A. At 500 Hz sampling it has 2 milliseconds to rise.

You started by asking about the differences in the measurements.  Do you have other questions now?

Lynn