# DC Voltage & Current Sensor Help Needed

#### ElectEng

Joined Sep 3, 2019
42
Requesting help on reading the DC voltages and currents on my project. Values should not exceed 35V and 5A.

I am going to make my own circuit card regardless but would like some ideas. I could try making a board based off the ACS712-20A Module for Arduino and add a voltage divider onto it.

Is this reasonable? Any advice?

For the voltage sensor do I need to worry about the current? If I make the current sensor how big should I make the traces?

#### Sensacell

Joined Jun 19, 2012
3,011
Your voltage sensor needs to share a common ground with the voltage being measured, as drawn, it doesn't.

The ACS712 provides isolation- totally

#### ericgibbs

Joined Jan 29, 2010
15,734
hi EE,
Also check your resistive divider, using the values on the diagram, the Voltage limit is 25V, to give an ADC in of 5V.
At 35V the input to the ADC will try to over drive it at ~7V.

Reduce the 7k5 to 5k

E

#### ElectEng

Joined Sep 3, 2019
42
Your voltage sensor needs to share a common ground with the voltage being measured, as drawn, it doesn't.

The ACS712 provides isolation- totally
Good catch! Thanks.

hi EE,
Also check your resistive divider, using the values on the diagram, the Voltage limit is 25V, to give an ADC in of 5V.
At 35V the input to the ADC will try to over drive it at ~7V.

Reduce the 7k5 to 5k

E
Thanks for the heads up. I actually just grabbed a photo on google images to show the hook-up and disregarded the values.

I am slightly confused. Let us say I want to read the voltage, 12V - 18V, and current, 1A - 2.5 A, coming off the solar panel. Do I need to design a circuit card that can withstand 2A on the voltage divider side as well? In other words, can I do this with 0805 SMD Resistors on the voltage sensor part? If not, do I need to place several resistors in parallel?

#### BobaMosfet

Joined Jul 1, 2009
2,002
@ElectEng-

Your voltage divider is useful solely for measuring voltage. As such, current is not a concern so long as you use high-enough value resistors. The higher the value, the less current they will pass, and the less energy they will dissipate as heat.

If your voltage divider is 30K + 5K, for a total of 35K, then:

I = E/R
I = 35 / 35000
I = 0.001A (1mA of current flowing through the divider, total)

P = IE
P = 0.001 * 35
P = 35mW

1/8th Watt (which is 125mW) Resistors are more than enough to safeily handle the power involved.

#### KeepItSimpleStupid

Joined Mar 4, 2014
5,088
Comment:

Dividers are one thing. Max input voltage with the gizmo "powered off" can be a real problem.

The AMP04 https://www.analog.com/media/en/technical-documentation/data-sheets/AMP04.pdf would allow differential measurements. Protections to something in range of much less than +-18V is easy to do. With a single supply , you get almost 0V. You always have to worry about that to. You can ALMOST reach the supply voltage too.

Let's assume that the aquisition will be "close" to the panel, You don't want to run lead 100'. Voltage generally needs twisted-pair shieleded cables, ideally with the shield connected at the sourcing end.

Shielding protects against RFI and twisting EMI. The EMI protection occurs when the inputs are differential.

If these are power supplies that you were measuring, other stuff comes into play because the reference is usually the positive terminal,

See if you can use any of the power supply supervisory IC;s. Some can do voltage, current and power.

Current (0-20mA) and (4-20 mA) is used for log distance process signals. The 4mA is used to power the transmitter.
Converting to a voltage requires a 250 ohm resistor on the measuring side,

If your changing the setpoint (0-5V), again a resistor can be used. Isolated or non-isolated current out can be used. The 250n ohms at the device ensures the device gets a good reference.

When I did my dividers from about 40V to 0-5V, would set in the software the nominal voltages. e.g. 40V = 5V, then to calibrate, the initial calibrations would be 1.000. You don't want 5*7.5 = 40V. When selecting resistor ratios, you do need to check the absolute possible output and make sure it's not exceeded when tolerances are applied.