http://www.sames.co.za

7/16

PRELIMINARY

RM4104ASEB

between the SA4104A and the power supply (in the place

of J9, J10 and J11) and on the two shunt connections to

the PCB. The voltage input network cannot be protected

with a ferrite bead but this is typically not necessary

because the attenuation is very high.

Further protection mechanisms are related to the PCB

layout and are described in the next section.

PCB DESIGN CONSIDERATIONS

There are numerous PCB design aspects to consider when

designing

an

energy

meter

using

SA4104A.

These

principles have all been incorporated in the sample PCB

layout given in the “PCB Layout” section.

The first is the location of critical components. The current

and voltage sensing input resistors (R17 to R21) with their

associated low pass filtering capacitors (C11 to C13)

should be located as close to the device pins as possible.

The same holds for the reference resistor (R22) with its

associated filtering capacitor (C14) and the supply bypass

capacitors (C8 to C10).

The SA4104A should be placed on a solid ground plane

that is connected to the AGND pin of the device. This

ground plane should be kept clear of noise by only

connecting it to the ground plane of the power supply and

the LIVE IN input at a single point. It should also be kept

away from any high frequency, high voltage or high current

signals that may induce noise. For example, the first

section of the voltage input attenuation network (R6 and

R7) should be placed far away from this ground plane. If a

ferrite bead is used to connect the rest of the meter's

ground to this ground plane then identical ferrite beads

must be placed into the power supply lines (VDD and VSS)

and into the current input lines from the shunt. If a single

ferrite bead is placed some signals are filtered and others

are not, which will create differential noise between the

unfiltered and the filtered signals. This will affect the

performance

of

the

SA4104A

in

the

presence

of

electromagnetic disturbance.

As far as the immunity to electromagnetic interference is

concerned the guideline is simply to minimize the parasitic

inductance. Each PCB net has a parasitic inductance and if

this is not sufficiently small it could cause resonance with

the parasitic capacitance at low enough frequencies to

affect the performance on the HF interference test or the

FTB test. Keeping the PCB tracks as short as possible is

one method to avoid this scenario. Parasitic inductance is

also a factor that can render the MOV almost useless

because a voltage spike can be amplified in both

magnitude

and

duration

by

series

inductance.

The

capacitor in parallel with the MOV cancels some of this

inductance but still all measures to avoid parasitic

inductance should be adhered to.

EXAMPLE DESIGNS

Example1

Nominal voltage: 220V

Maximum current: 40A

Basic current:

10A

Shunt:

320

μ

Ω ; 12,5mV@ 40A

Pulse constant:

1600imp/kWh

Motor constant:

100imp/kWh

Using the design equations derived earlier:

equation(1):

R18, R19, R20, R21=200

Ω

equation(2):

choose C11, C12, C13=100nF to obtain

equation(4):

R16=100

Ω

equation(5):

R17=10k

Ω

equation(8):

Ω

equation(9):

choose R6=100k

Ω and obtain R7=75kΩ

equation(10):

R8=20k

Ω

equation(11):

R9 = 10k

Ω, R10 = 4.7kΩ, R11 = 2.4kΩ,

R12 = 1.2k

Ω, R13 = 620Ω, R14 = 300Ω

and R15 = 150

Ω. Some ratios are not

entirely accurate, but to ensure low cost

it is important to use only standard

resistor values.

equation(12):

using IVP=11 obtain DF_LED=879

equation(13):

DF_MO=16

Using Table 2 set FMS = '0', R3 = '1', R2 = '0',

R1 = '1' and R0 = '0'.

Example 2

Nominal voltage:

220V

Maximum current: 10A

Basic current:

2.5A

Shunt:

1m

Ω ;10mV @ 10A

Pulse constant:

3200imp/kWh

Motor constant:

100imp/kWh

Using the design equations derived earlier:

equation(1):

R18, R19, R20, R21 = 156 so set the

will be 16.7

μA which is still sufficiently

below the saturation point of the current

inputs

equation(2):

choose C11, C12, C13=100nF to obtain

equation(4):

R16=75

equation(5):

R17=7.5k

The

calibration

network

remains

unchanged

from

example1.

equation(12):

using IVP=11 obtain DF_LED=1758

equation(13):

DF_MO=32

Using Table 2 set FMS = '0', R3 = '0', R2 = '0',

R1 = '1' and R0 = '1'.