MCP39F511 Datasheet, PDF (47/62 Pages) Microchip Technology – Power-Monitoring IC with Calculation and Energy Accumulation

9.0 MCP39F511 CALIBRATION
9.1 Overview
Calibration compensates for ADC gain error,
component tolerances and overall noise in the system.
The device provides an on-chip calibration algorithm
that allows simple system calibration to be performed
quickly. The excellent analog performance of the
A/D converters on the MCP39F511 allows for a
single-point calibration and a single calibration
command to achieve accurate measurements.
Calibration can be done by either using the predefined
auto-calibration commands, or by writing directly to the
calibration registers. If additional calibration points are
required (AC offset, Phase Compensation, DC offset),
the corresponding calibration registers are available to
the user and will be described separately in this
section.
9.2 Calibration Order
The proper steps for calibration need to be maintained.
If the device has an external temperature sensor
attached, temperature calibration should be done first
by reading the value from the Thermistor Voltage
register and copying the value by writing to the Ambient
Temperature Reference Voltage register.
The single-point gain calibration at unity power factor
should be performed next.
If nonunity displacement power factor measurements
are a concern, then the next step should be Phase
calibration, followed by Reactive Power Gain
calibration.
Here is a summary on the order of calibration steps:
1. Transfer of the ambient temperature value
during calibration to the Ambient Temperature
Calibration register for temperature
compensation (optional)
2. Line Frequency Calibration
3. Gain Calibration at PF = 1
4. Phase Calibration at PF  1 (optional)
5. Reactive Gain Calibration at PF  1(optional)
Note:
Only needed if temperature compensation
is required. On system versions 0xFA14
and later this register transfer is done auto-
matically.
9.3 Single-Point Gain Calibrations at
Unity Power Factor
When using the device in AC mode with the high-pass
filters turned on, most offset errors are removed and
only a single-point gain calibration is required.
MCP39F511
Setting the gain registers to properly produce the
desired outputs can be done manually by writing to the
appropriate register. The alternative method is to use
the auto-calibration commands described in this
section.
9.3.1
USING THE AUTO-CALIBRATION
GAIN COMMAND
By applying stable reference voltages and currents that
are equivalent to the values that reside in the target
Calibration Current, Calibration Voltage and Calibration
Active Power registers, the Auto-Calibration
Gain command can then be issued to the device.
After a successful calibration (response = ACK), a
Save Registers to Flash command can then be
issued to save the calibration constants calculated by
the device.
The following registers are set when the
Auto-Calibration Gain command is issued:
• Gain Current RMS
• Gain Voltage RMS
• Gain Active Power
When this command is issued, the MCP39F511
attempts to match the expected values to the mea-
sured values for all three output quantities by changing
the gain register based on the following formula:
EQUATION 9-1:
GAINNEW
=
GAINOLD 
-E-----x---p---e---c---t--e---d--
Measured
The same formula applies for Voltage RMS, Current
RMS and Active power. Since the gain registers for all
three quantities are 16-bit numbers, the ratio of the
expected value to the measured value (which can be
modified by changing the Range register) and the
previous gain must be such that the equation yields a
valid number. Here the limits are set to be from 25,000
to 65,535. A new gain within this range for all three
limits will return an ACK for a successful calibration,
otherwise the command returns a NAK for a failed
calibration attempt.
It is the user’s responsibility to ensure that the proper
range settings, PGA settings and hardware design
settings are correct to allow for successful calibration
using this command.
 2015 Microchip Technology Inc.
DS20005393B-page 47