Causes and Cures for Tension “Zero-Drift”

When you have no material in your machine and the tension reading is indicating something other than zero you are most likely experiencing “zero-drift”.

There are several possible culprits, and one needs to examine the entire system to best provide the solution.  The components involved include the load cells, cables, the idler roll, bearings and shafts, and the tension signal conditioner or amplifier.

Many people wonder since calibration affects zero, and they calibrate at the amplifier or controller, that this is the culprit.

The logical place to look is the load cells. In order to provide precise and accurate readings, the load cells must be very sensitive to small changes in load.  There are basically two types of loads applied to the load cells.  The force that we are interested in is the force that is applied by the web, which is directly proportional to the tension.  Then there is what we call pre-loading.  These are forces that the load cells see that have nothing to do with tension.  They are due to the weight of the roller assembly supported by the load cells, plus forces exerted on the load cells during the mounting process.  The roller assembly weight is a constant and if the mounting procedure is followed precisely these forces can be minimized to be as constant as possible and, while they are still seen by the load cells, they can be electronically “zeroed” and not “seen” at the amplifier output.  Unfortunately, proper mounting procedures are often not followed.

The following is a troubleshooting guide:

Mounting Surface
It is also important that the mounting surface is very rigid.  Any deflection in the structure can transmit to the load cell assembly, causing a minute amount of preloading.  Remember, most load cells need to deflect only .005 inches (.127mm) to provide full output.  This means that if there is deflection of .0005 inches .0127mm), the load cell will see this as a 10% change.

Temperature Variation

Temperature variations both in the plant and on the machine can also be a cause of zero drift.  10-15 -degree C changes could cause this.  While there may not be much done about changing this fact, calibrating when the plant is at its most typical temperature is advised.  Recalibration may be required seasonally if temperature fluctuations are severe.

If Using Idler Roll
Look at what style of load cells are being used.  The trouble shooting methods are different depending upon the type of load cell.  Load cells with a dead-shaft idler roll commonly have the shaft clamped in the load cell with some twist on the shaft, meaning that one load cell head is rotated in one direction, and the other load cell head is rotated in the opposite direction.  This can impart a force on the beams that may show up as zero drifting.  To check for this, grasp the clamp on the load cell on one side of the roll and twist it.  The idler roll should rotate slightly against the anti-rotation stops.  You should hear a slight “click-click” as the head hits the stops.  If the head on the load cell does not move, then the shaft needs to be re-set in the load cells.  This is done by loosening both idler roll clamps and carefully tightening them evenly on each side while supporting the idler roll.  If bushings are being used, make sure they are split bushings and that the split in the bushing is aligned with the split on the clamp. Check that there is slight rotational movement of the head.

Another possibility is that the idler roll is not centered between the load cells and the shaft pushes up against one load cell.  This may not be apparent during initial installation, but after running some time, it can show up as zero drifting. Shaft expansion during operation can push the load cell head up against the beam and prevent it from returning to its zero position.  To make sure this is prevented, make sure you can pass a business card (or something of similar thickness) between the end of the dead shaft and the back of the load cell clamp.  There is also a wave spring in the load cells behind the clamp that compensates for axial shaft expansion. Make sure there is axial movement with the idler roll installed.  You should be able to compress the wave spring slightly by pushing axially on the idler roll, and noting if the shaft can move (perhaps 0.125” to 0.0625”).  If it does not move, or moves in one direction and not the other, re-centering of the dead-shaft will eliminate that as a potential drift cause.  In some cases it may be necessary to shorten the dead shaft.  In the case of the L/ER style load cells, the roller may need to have the bearing bore made deeper, or the overall shaft length shortened.

If Using Pillow Blocks

With under pillow block load cells if the mounting surface is not perfectly flat, tightening of the mounting bolts could cause zero drift.  If there is old paint or debris between the base of the load cell and machine frame where the load cells are mounted, problems can occur.  To check this, zero the circuit and then loosen the mounting bolts one at a time.  If the zero value changes, then you must ensure the mounting surface is flat.  Sometimes the bearings and the way they are mounted can cause this as well.  Some pillow-block bearings do not properly allow for expansion due to heat.  This can impart a side-load on the load cells and cause zero drifting.  This can be addressed by either replacing the bearings with self-aligning bearings or proper maintenance of the existing ones.

Conclusion

As you can see, there may be many causes, but typically it is not the amplifier.  Just to be sure, the +/-2.5Vdc power supplies out to the load cells need to be very stable.  You can measure these easily.  It does not need to be exactly 2.5 vdc, but it should not be changing at all from the value you do measure.

If you are experiencing zero drift and are unable to troubleshoot contact a Montalvo Applications Specialist for further assistance.