Why are we getting an A/D error after we replace our old controller?

We are replacing our obsolete weight controller with a new Hardy controller and want to reuse the same load cells. Our traditional calibration methods are giving us problems when we try to set the span. I think it's because our old unit has 10 volts excitation and the Hardy unit has 5 volts excitation. When we run the mV calibration signal above 15, we get an A/D Conversion Error on our Hardy unit. Do we need to take our old calibration values and divide by 2?

The excitation voltage on all Hardy controllers designed after 1999 is 5vdc.  Other units and other manufacturers may use another voltage for excitation.  The voltage used does not matter, as load cells are analog devices.  Load cells can and will work with any voltage supplied to them, as long as this voltage is less than the manufactures maximum voltage specification.  The A to D converter detector's ability to filter noise is the key to a successful weight controller design.

In your example it looks like the load cells are over ranging.  Where some instruments do not let you know the load cells are out of range it is an important bit of information. A load cell manufactured to a specific mV/V range is said to be linear to it's certificate.  If the load cell is over ranged/damaged and your instrument lets it be calibrated.  You are using load cells that may no longer be linear.

The signal returned from the load cell will be based on its mV/V rating. This means that if you have a 2 mV/V rated load cell, and hook it up to a controller with 10 volts excitation, the load cell will return from 0 to 20 mV over its rated range. If you hook up the same mV/V load cell to a Hardy unit with 5 volts excitation, the load cell will return from 0 to 10 mV over its rated range. The same load cell with the same load applied will return twice the mV with a controller providing 10 volts excitation as compared to a controller providing 5 volts excitation.  Different excitation voltages supplied, but the load cell will work equally well on either unit.

The input range that a controller can accept is based on its sense voltage limits. A controller using 10 volt excitation is probably designed to accept an input range of 0 to 31 mv.  This is the full range of a 3 mv/v rated load cell using a 10 volt excitation. Most Hardy units use 5 volts excitation and are designed to accept an input range of 0 to 15 mV, which is the full range of a 3 mV/V rated load cell with a 5 volt excitation.

If you are exceeding 15 mV on the load cell input to your new Hardy controller you will get an A/D error. This may be a faulty load cell or if you are using a simulator,  limit the simulator output to stay within that specified range.

The sense voltage actually establishes the limit by figuring the maximum range.  This sense voltage reflects any line loss for excitation voltage and ranges the controller.  If your excitation voltage is reduced by ten percent due to line loss or protective devices, the sense line over range limit will also be reduced ten percent.  Using 5 volts excitation reduced to 4.25vdc reflects a maximum range reduction of 15 mv to 13.5mv.

This differences in the mV range of the units are transparent unless you are using a simulator or are actually measuring the mV with a meter. You will need to use the sense line value times the load cell rating to calculate the milli volt change when a known weight is applied. In the example above the 13.5mv signal represents full scale.