I have to test speen sensors for work and have run into a problem. The sensors in question are hall effect based with a digital output and are used to monitor wheel speed using a toothed wheel. My specification requires me to measure several resistances between supply, ground and output. These resistances must fall within tolerence. I have a multimeter which will measure these fine. I have another meter the same make but newer model which is out by 9Mohm. Also the calibration engineer tried his meter and this was out by 15Mohm.
My question is; is it possible that each meter could be calculating the resistances using slightly different voltages and could these voltages account for the difference in resistance?
Sensor testing
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
Very interesting... Very, very etc etc.
Weeeeeeeeeeell... If the meters at at a different voltage then ... More spec needed. ( sensor spec)
How a bout some pics.
Is the wheel spinning when you take the readings?
Come an av a go if ya think ur are enough!
Tar&stuff.
Weeeeeeeeeeell... If the meters at at a different voltage then ... More spec needed. ( sensor spec)
How a bout some pics.
Is the wheel spinning when you take the readings?
Come an av a go if ya think ur are enough!
Tar&stuff.
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
There is more to it than the voltage of the meters, how far are the meters from the sensor, the impedance of the person holding the meter is a magnetic change, in ...struggle for words...strength and direction. How sensitive is the sensor. The meter causes change infield, lots of things could effect it, a change in room temperature changes the impedance of the atmosphere, ESD.
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
Try changing the test leads over in the meters.
Is to the meter or the test lead?
Is to the meter or the test lead?
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
I should learn to read before I post ...baly predictive text.
Is it the meter or the test lead?
Ie the impedance to both current and magnetic field, the plastic handles are a change in field.
Is it the meter or the test lead?
Ie the impedance to both current and magnetic field, the plastic handles are a change in field.
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
To get an accurate reading the sensor must be in a uniformed magnetic field of known strength, and known direction.
Put long wires on the sensor and include in wire resistance in calculations.
Try putting all 3 meters on a passive resistor and see if they are closer.
Put long wires on the sensor and include in wire resistance in calculations.
Try putting all 3 meters on a passive resistor and see if they are closer.
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
Put all 3 meters on a passive and see if they are the same as individual readings, if they are then it is not the voltage,
-
spanner_monkey
- Posts: 5
- Joined: Thu Aug 06, 2015 2:40 pm
Re: Sensor testing
Sorry for the late reply.
I managed to narrow the problem down to the relationship between the probe and the meter.
As an example; The probe is tested out of circuit and away from any object that could cause of effect a reading. One of the tests is to check and record the resistance between two of the terminals. The specified reading is 10Mohm. The meter we regularly use gives us a reading of 10Mohm within acceptable limits eg +/- 1Mohm. The meter we have as a backup is a later version of the same meter. This meter gives us a reading of 1Mohm on this particular test. Both of these meters have been calibrated to industry standard calibration standards and both give identical readings at both 10Mohm, 1Mohm and every reading inbetween. All test leads have been tested and conform to standard. All leads have been swapped between meters with no variation in test results. As a control we tried doing the test with the calibration engineers multimeter. This instrument is calibrated to a higher standard than ours and gives the same results as ours when checked against a proof resistance. This multimeter gave a reading of 25Mohm. This reading had everyone stumped, including the calibration engineer who said he had never seen anything like it.
The only thing I have come up with is that it must be some kind of interaction between the internal circuitry of the probe and the voltages the multimeters are testing at.
As I dont have a specification for the probe I cannot confirm or disprove my theory. I would just like to know if anyone else knows of a similar case, knows if my theory is even possible or can offer any alternate suggestions.
I managed to narrow the problem down to the relationship between the probe and the meter.
As an example; The probe is tested out of circuit and away from any object that could cause of effect a reading. One of the tests is to check and record the resistance between two of the terminals. The specified reading is 10Mohm. The meter we regularly use gives us a reading of 10Mohm within acceptable limits eg +/- 1Mohm. The meter we have as a backup is a later version of the same meter. This meter gives us a reading of 1Mohm on this particular test. Both of these meters have been calibrated to industry standard calibration standards and both give identical readings at both 10Mohm, 1Mohm and every reading inbetween. All test leads have been tested and conform to standard. All leads have been swapped between meters with no variation in test results. As a control we tried doing the test with the calibration engineers multimeter. This instrument is calibrated to a higher standard than ours and gives the same results as ours when checked against a proof resistance. This multimeter gave a reading of 25Mohm. This reading had everyone stumped, including the calibration engineer who said he had never seen anything like it.
The only thing I have come up with is that it must be some kind of interaction between the internal circuitry of the probe and the voltages the multimeters are testing at.
As I dont have a specification for the probe I cannot confirm or disprove my theory. I would just like to know if anyone else knows of a similar case, knows if my theory is even possible or can offer any alternate suggestions.
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
You might need 3 resistors in a star configuration, take 3 sets of readings, test each resistor with each meter, add the readings together and device by 3.
I am not sure how meters behave in parallel, current through a single resistor would be divided by 3, if you measure 1 resistor with 3 meters simultaneously, thus resistance reading would be roughly 1/3, That would give you the current through each meter in ratio, and from that you could work out the voltage across their outputs.
I will have to think about how to do that, may be 4 meters would be easier, a Wheatstone bridge configuration. You could have a star connection with 4 resistors instead of 3.
I am not sure how meters behave in parallel, current through a single resistor would be divided by 3, if you measure 1 resistor with 3 meters simultaneously, thus resistance reading would be roughly 1/3, That would give you the current through each meter in ratio, and from that you could work out the voltage across their outputs.
I will have to think about how to do that, may be 4 meters would be easier, a Wheatstone bridge configuration. You could have a star connection with 4 resistors instead of 3.
-
piratepaul
- Posts: 432
- Joined: Mon May 20, 2013 2:45 pm
Re: Sensor testing
Tell me what it is when you figure it out, the difference is baffling. Bang.
Impedance of the meters?
Impedance of the meters?