Using ACS712 in parallel

Hello everyone. I am trying to measure current with a ACS712 (30A) sensor. However, I need to have a good precision, which the 30 A model can't deliver. If I use 6 ACS712 of 5A in parallel will the accuracy increase or not? I have tried this with 2 30 A sensors. However, due to a faulty sensor it doesn't add to actual current in the circuit (5.14 A) as you can see below. The faulty sensor when measuring 0 A reads average of 514 instead of 512/511. Do you know any way to solve this?

Green trace looks like 5.7+/-.2

you can use a shunt and ADS1115 in differential mode

1 Like

Minor differences in impedance will result in significant differences in current distribution across the sensor array. This may also explain the 0 reading on your "faulty" sensor. On top of this, your setup will not compensate for systematic errors, and may even induce new, unanticipated ones.

So no, not a particularly good idea, although creative.

2 Likes

Post a drawing of how you have this connected.

Dangerous quote.

That won't work most of the time,
because the two inputs of the ADS1115 must stay within VCC/GND of the ADS supply.

A shunt in the +line could be higher than ADS VCC, a shunt in the ground path of the load could produce a negative voltage, depending on current direction.

A shunt with a chip of the INA family could work,
assuming OP wants to measure low voltage DC, not mains AC.
Leo..

1 Like

Yes but it should have been 5.14 A according to a multimeter that I have

Idea:

Instead of putting multiple 5A sensors in parallel, put five 30A sensors in series.
Average them, or take the median. The median protects to some extend to outliers.

e.g.
Suppose you read 11.5, 25,3, 25.1, 25,4 and 25,2 (note there is one outlier)

Average = 22.5 (affected by outlier)
Median = 25.2 (not affected by outlier)
Average of middle 3 sensors = 25.2

1 Like

No it will not. As a matter of fact it may be worse.
What accuracy do you need?

1 Like

I would need like 0.2 A of accuracy

The ACS712 has an accuracy of +/-1.5% at 25ºC
So for the 30A device that's +/- 0.45A

1 Like

They make inexpensive current shunts that are accurate and temperature compensated for not a lot of $$$. More cost effective less accurate are also available for a few $$$. Its accuracy will depend on your hardware and temperature compensation if needed.

I did not see your required accuracy or if it was AC/DC.

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.