The Internet is full of troubleshooting guides for strain gauge load cells; all you have to do is search. While most of these guides contain useful information, some go into more detail than is necessary while others gloss over some important areas. This guide is intended to help you diagnose problems with strain gauge load cell systems. It is not intended to help you isolate the problem to an internal component within the load cell or explain exactly how the load cell operates. After all, you most likely are not interested in repairing or rebuilding the load cell; you just want to make sure the load cell is at fault and return your scale to proper operation. This guide is not written around any one manufacturer but is general in nature. Let’s get started.
Define the Problem
The first step is simply to define the problem. Here are some typical symptoms with corresponding diagnostic steps to take. Your scale may exhibit more than one.
1) The weight display is not stable. It moves up and down continuously. (go to section A)
2) The weight display shows an error message indicating that the load cell output is too high or too low. (go to section B)
3) I can’t zero the weight display. (go to section C)
4) The weight display shows weight, but it does not change with the addition or removal of load. (go to section D)
5) The weight display is blank. (go to section E)
6) The weight display shows the scale to be over capacity when it isn’t loaded at all. (go to section F)
7) The weight display functions correctly, but the weight value shown is wrong. (go to section G)
8) The weight display will not return to zero with the removal of load. (go to section H)
Before you Begin
Before moving to the sections containing the diagnostic steps, first make sure that the scale has been installed correctly. Some installation problems can have the appearance of a load cell problem when, in fact, the scale has simply been installed incorrectly or not been calibrated. Has the scale been operating correctly then, for some unknown reason, exhibited one of the aforementioned problems or has it never performed correctly? If the answer is the latter, return to the installation and calibration instructions that accompanied your scale and make absolutely certain that it has been correctly installed and calibrated.
Most of the diagnostic steps listed below require the use of a digital volt ohmmeter. While an inexpensive model will work for some of the tests, it is best to invest in a quality meter. If you can find a meter that has a 2000 Megohm or greater range in the resistance mode, use it for these steps. Otherwise you will also need a Megohmmeter (a meter used to measure high resistances) to perform leakage tests. Again, invest in a quality instrument; it will make the job much easier.
A changing weight display is generally a sign of moisture in the load cell, load cell cable, load cell connector or terminals, but not always. Begin by identifying the signal or output leads of the load cell. Your weight indicator may have either a multi-pin connector for connection of the load cell or screw or spring terminals for connection of the load cell cable. Refer to the instruction manual for your weight indicator to determine which pins or terminals are the input or signal leads from the load cell circuit. Even if your scale uses more than one load cell, there will only be two signal leads from the load cells that connect to the weight indicator. Once you have determined which pins or terminals are the + input or signal and – input or signal, short them together with a piece of bare wire and observe the weight display. If the display is now steady, the problem is most likely in the load cell or load cells but, if the weight display continues to move up and down, the weight indicator may be at fault.
- Assuming that the weight indicator became stable when the signal was shorted, continue by looking for signs of moisture or water in the load cell junction box or boxes, along the load cell cables and at load cell connectors. If none is found, continue to step 2. If you do find moisture in the junction box, dry it out completely and seal any entry points where the moisture entered.
- If your scale has more than one load cell, disconnect one load cell at a time and observe the weight display. If the weight display becomes stable when a cell is disconnected, you have isolated the problem load cell, now go to step 3. If your scale only has one load cell, continue to step 3.
- Disconnect all of the wires of the problem load cell from the weight indicator or load cell junction box. Twist the wires together. Using either a digital multimeter with a 2000 Megohm scale or a megohmeter, set it to a resistance range of at least 2000 megohms or more. Connect one of the test leads from the meter to the load cell body making certain that it makes good contact. Touch the other test lead to the twisted wires coming from the load cell. Do not touch the test lead or the load cell body during the test because your body resistance will adversely affect the reading.
- Observe the meter display. If the meter display shows 2000 megohms or more, there is no moisture in the load cell or load cell cable. If, however, the reading on the meter shows something less, it is likely that there is moisture in the load cell or load cell cable.
- Sometimes you can drive the moisture out of the load cell by placing it in an oven set to 150 degrees F and leaving it for several hours. Repeat steps 3 and 4 to make sure the moisture is gone before reinstalling the load cell. While baking a load cell may allow you to return it to temporary service, it will not eliminate the moisture problem. If you find a seal or potting on the load cell that is cracked or broken, it may be the entry point for the moisture and the cell must be replaced or repaired.
Most modern weight indicators will display an error message when the load cell output is too high or too low. The level at which these output error messages are triggered should be in the owner’s manual. A load cell with an open in one or both signal or output leads will produce this error as will those that are simply too high or too low. High outputs can be the result of shock loading where the load cell spring element is actually permanently deformed producing the high output. If this is the problem, it is sometimes possible, when the scale uses a single load cell, to repeat the calibration procedure establishing a new zero reference point. This technique may get you out of the woods sometime but should only be considered temporary until the cell can be replaced. Perform the following steps to determine if the cell is at fault.
- Set the digital multimeter to read dc volts in a range of 20 v. With the indicator on, measure the voltage across the input or excitation terminals of the load cell circuit. Normally, you should see something between 5.0 and 15 volts although there may be other excitation voltages that are higher or lower. Record this voltage reading on a piece of paper.
- With the scale unloaded, set the digital multimeter to read a range of 20 mv. Note that if your meter is auto-ranging, all you need do is select the DC voltage setting. Read the voltage between the output or signal terminals of the load cell. A load cell with an output of 2 mv/v should read somewhere between 2 and 15 percent of its rated output. (This, of course, depends on the dead load of the weighbridge but is a place to start.) For example, if you recorded an excitation voltage of 8.0 volts in step 1 and you have a 2 mv/v load cell (you can determine this by looking at the nameplate or data sheet that accompanied the load cell) the maximum output is 8 volts x 0.002 volts per volt (this is the same as 2 mv/v) = 0.016 volts or 16 mv at the capacity of the load cell. If the cell is loaded to 10 percent of its capacity with the platter or weighbridge, the voltage read on the output terminals should be around 1.6 mv. Note that it could be somewhat higher or lower depending on the dead load. If, for example, the reading shows a -4 volts, then the load cell is at fault and should be replaced. Or a reading of +20 mv also indicates a faulty load cell.
If you are unable to zero the display on the weight indicator, it may be due to one of several things. First, make certain that the weight indicator has been set up to do a semi-automatic zero to scale capacity. Second, if the weight indicator is not at fault, the load cell may have suffered a shift in its no-load output. If the shift is in the positive direction, it may be the result of a shock load. Refer to Section B on how to check for a shift in no-load output due to a shock load. If the weight display is off set to a negative direction refer to Section B. Depending on the magnitude of the offset, it may be possible to recalibrate the scale establishing a new zero reference point and return to normal operation. As previously mentioned, this is a temporary fix and should be used to get you back in operation while a new cell is ordered.
With some weight indicators, it is possible to display a weight value but not respond to changes in load on the scale platform. In any commercial application, this should not happen since the weight indicator is forbidden to display a valid weight reading if the load cell is not operating. Before checking the load cell, cycle the power to the weight indicator to see if the problem is corrected. If it isn’t, turn the weight indicator off and follow the steps below:
- Begin by disconnecting the load cell from the weight indicator. Identify each of the load cell leads. There should be four or six leads plus a shield. Refer to the data sheet accompanying the load cell for the wire color for each function. If you don’t have the data sheet, you should be able to obtain the information from the Internet.
- Set the digital multimeter to resistance and a range of 2000 ohms if the meter is not auto-ranging. Measure the resistance between the + and – signal leads. Normally, the output or signal lead resistance is very close to the bridge resistance and, in most cases, will be either 120 ohms, 350 ohms or 700 ohms. If you measure a resistance that is within 10 ohms of one of these values, the input resistance is probably acceptable.
- Next, measure the resistance of the input or excitation leads of the cell. This resistance will normally be somewhat greater than the output resistance. Typically, if the measured resistance is no greater than say 120 ohms more than what was measured in step 2, the reading should be acceptable.
- If either the input or output resistance of the cell is substantially (>50 ohms) the cell should be replaced.
If the weight indicator display is blank, check to see if there is an over capacity annunciator that is also on. If not, make certain that the weight indicator is turned on and is plugged into a power outlet that is active. If everything looks correct, check to make sure that a breaker is not tripped or plug a light into the same outlet to see if it works. If everything checks out, toggle the power switch on the indicator to see if it goes through a display test. If nothing happens, have the weight indicator repaired before proceeding. If the over capacity annunciator is turned on, refer to Section B.
If the weight indicator shows an over capacity message with the platform or weighbridge unloaded or only slightly loaded, refer to Section B. Over loads are indicative of a large positive offset of the load cell(s).
If the weight display is functioning correctly but giving a weight that is obviously wrong, it could be the result of one of several problems. Obviously, the first thing to check is the calibration of the weight indicator. If you are absolutely certain that the indicator has been properly calibrated, then take a look at the scale itself. If the indicator is reading light, look to see if there an obstruction that is causing a bind in the movement of the weighbridge. If there is more than one load cell, are they all taking weight? Is the checking system out of adjustment and sharing some of the load with the platform? If, after checking all of these things, the problem still persists, disconnect the load cells one at a time observing the weight display. If you find that the display changes very little when one particular cell is disconnected, use the previously described techniques to check the load cell. It is also a good idea to check the condition of the load cell junction box and trimming circuit to ensure that they have not been damaged or contaminated with moisture.
If a weight indicator fails to return to zero after removal of a load, it is almost always some sort of mechanical obstruction. Observe the weight display after removal of the load. If the display gradually decreases over a period of ten to fifteen minutes to zero, the problem may be creep in the load cells. Note, however, that this is affected by the length of time the load has remained on the platform. In other words, leaving a heavy load on the platform for ten minutes or more may cause a delay in the return to zero due to creep in the load cells. Also, a change in temperature from when the load is applied to when it is removed can also cause a shift in zero and should be taken into consideration. If neither of these are the problem, look for binds in the weighbridge or material like rocks or snow lodged between the weighbridge and pit wall. Any connection between the live portion of the scale (the weighbridge) and the dead portion (pit wall, lower frame, floor, etc.) will affect the reading and must be removed. Another place to look is the restraint system. An incorrectly adjusted or installed restraint system can place a force on the weighbridge preventing a return to zero.
Finally, for those of you interested in digging a little further into the anatomy of a load cell, refer to the figure below. This is a typical schematic of a strain gauge load cell. The strain gauges are wired in a bridge formation commonly referred to as a Wheatstone bridge. Normally the resistance of each strain gauge is the same with values of 120, 350 and 700 ohms being common but there could be others. The point is that all four strain gauges have the same resistance. Sometimes manufacturers may place more than one gauge in the arms of the bridge so you could have two 350 ohm gauges in each arm which would give you a 700 ohm bridge.
There are resistors in series with the excitation or input arms of the bridge. These resistors are typically constructed from Balco and are used to compensate for changes in span or output with temperature. As the temperature goes up, the gauge spring element gets softer and deforms more to a given load. To make things worse, the gage factor also increases with temperature causing the output of the cell to go even higher with temperature. As the temperature increases, the resistance of the Balco resistors increase thus decreasing the excitation voltage across the strain gauge bridge itself lowering its output. With the proper selection of the Balco resistors, the output of the load cell remains relatively constant over its range of operating temperatures. Normally a resistor is wired in parallel with each Balco resistor. This higher value resistor is called a shunt resistor and used to linearize the response of the Balco resistor. The values of these resistors remain the same from cell to cell provided they are the same model load cell from the same manufacturer and of the same lot. Other lots of load cells may have slightly different values of Balco and shunt resistors.
Additional resistors are found in the output of the strain gauge bridge. These resistors are used to compensate for changes in the no-load output of the load cell with changes in temperature. Manufacturers check each individual load cell for changes in no-load output with temperature then select the proper sized resistors to compensate for this change. For that reason, there is greater variance in the output resistance than there is in the input resistance since each load cell requires a different amount of no-load temperature compensation.
Finally, some load cells have one or more resistors across their output terminals to trim the output to an acceptable level.
Knowing this, it is possible to determine if there is a break in the internal load cell circuitry and the possible locations of that break. While you probably will not repair the load cell yourself, you may nevertheless, find it interesting to know just what likely happened within your load cell.
The table above is based on a typical 350 ohm strain gauge load cell and is for illustrative purposes only. Do Not use these values in troubleshooting your load cell.
1 Normally an open Balco is the result of a voltage surge on the excitation or input lines to the load cell.
For Load Cell Product Info: