Current sensing circuit - shunt into amplifier into ADC

Thread Starter

damirault

Joined Sep 7, 2016
2
Hi all
I'm into a circuit that's new to me.
Basically I need to monitor current (most likely through a shunt), 8-30VDC, max load current 4A. Repeated 10 times on a board. I need a high signal when the load current is above 40mA and a low when below. I'm looking at the LTC6101 and the LTC2460 ADC. My questions are, 1) will this work and 2) Can this be achieved in a simpler way?
Thank you!
-Dan
 
Hi all
I'm into a circuit that's new to me.
Basically I need to monitor current (most likely through a shunt), 8-30VDC, max load current 4A. Repeated 10 times on a board. I need a high signal when the load current is above 40mA and a low when below. I'm looking at the LTC6101 and the LTC2460 ADC. My questions are, 1) will this work and 2) Can this be achieved in a simpler way?
Thank you!
-Dan
If you just want a high/low threshold then you don't really need an ADC. And if that's the case, what is the threshold? The difference between 40mA and 4A is pretty significant. Could you get by with, for example, above or below 2 amps? In that example, you could use a 100 milliohm shunt and get a big enough signal to measure directly with a comparator. Like a
Hi all
I'm into a circuit that's new to me.
Basically I need to monitor current (most likely through a shunt), 8-30VDC, max load current 4A. Repeated 10 times on a board. I need a high signal when the load current is above 40mA and a low when below. I'm looking at the LTC6101 and the LTC2460 ADC. My questions are, 1) will this work and 2) Can this be achieved in a simpler way?
Thank you!
-Dan
If you just want a simple threshold, above or below 40 mA, then you don't need an ADC at all. Just a comparator. Look at the LT6108. But since you need 10 channels, you probably dont want to pay for 10 references. Look at TIs current send amps INA181 et al. Then you can use a commodity quad comparator like an LM339
 

Ian0

Joined Aug 7, 2020
9,667
First, use an INA180 or INA181 I'd recommend the one with a gain of 200 - the A4 version. (These are also available in duals and quads - INA2180/INA2181/INA4180/INA4181) and a 0.15Ω 1W current sense resistor.
That would give 6mV across the resistor at the trip point and 1.2V at the output.
Then you have a variety of possibilities:
If you already had a microcontroller with enough spare A/Ds then you could simply measure the output voltage.
If you have a microcontroller with spare comparators, you could use those.
If neither, use some LM339 comparators. Connect the + inputs to the outputs of the INA
Connect the - inputs together and connect them all to a voltage divider from your 5V or 3.3V supply. (1.6k/5.1k for a 5V supply, 4.7k/8.2k for a 3.3V supply)

If you need a very accurate threshold, connect the - inputs to a 1.2V reference and use a 0.15Ω current sense resistor.

If you don't need great accuracy, connect the INA outputs straight to a logic input, then use a 0.2Ω sense resistor for a 3.3V supply or a 0.33Ω resistor for a 5V supply.
 

Ian0

Joined Aug 7, 2020
9,667
As an alternative, placing a schottky diode across the sense resistor allows a larger value sense resistor, which eliminates the INA current sense amplifier; although the shottky diode will need a heatsink at 4A.
The blue trace is the voltage across the sense-resistor and diode.Screenshot.png
 

BobTPH

Joined Jun 5, 2013
8,806
A small point. Shunt is the wrong term for ehat you are doing. A shunt is a resistor you put in parallel with s sensitive meter to read high currents. You will be using a series resistor, which is called a current sense resistor.

Bob
 

Ian0

Joined Aug 7, 2020
9,667
A small point. Shunt is the wrong term for ehat you are doing. A shunt is a resistor you put in parallel with s sensitive meter to read high currents. You will be using a series resistor, which is called a current sense resistor.

Bob
Interesting point - but doesn't the INA180 (or whatever) take the place of the "sensitive meter" so that the low value resistor is in parallel with it?
If I were actually using a meter, then the "shunt" would still be in series with the load.
 

BobTPH

Joined Jun 5, 2013
8,806
I guess you could look at it that way, but the ina is a voltage measuring device, not a current measuring device. I am only stating what I believe to be the accepted usage of the two terms.

Bob
 

crutschow

Joined Mar 14, 2008
34,280
I agree that a shunt resistor is the correct term for measuring current in either circumstance.
How you measure the voltage across the shunt makes no difference.
The "sensitive meter" that's across the shunt measures the shunt voltage not current.
 

Ian0

Joined Aug 7, 2020
9,667
Whenever I buy one of these
https://uk.farnell.com/productimages/standard/en_US/72T0825-40.jpg
I have only ever seen it called a shunt.
And whenever I buy one of these
https://uk.farnell.com/panasonic/erj8cwfr010v/current-sense-res-0r01-1-1-w-1206/dp/2145283
I have only ever seen it called a current sense resistor
And it seems that the terms are mutually exclusive, yet the devices perform exactly the same function.

I would agree that "current sense resistor" is a better descriptive term, but since when has the electronics industry preferred the most logical name for anything. In fact, I think it's getting worse - everyone knew what a fuse box was, but "consumer unit" . . . .?
 
Tomato, tomaaaaato. We all know what it is by either name.

But "shunt" is ambiguous. The first hit on Google is the medical device. Also, there's "hyperspace shunt". Whereas the first hit on Google of "shunt resistor" or "current sense resistor" are the kind we are talking about.
 

Ian0

Joined Aug 7, 2020
9,667
OED gives the first usage in 1440 from Sir Gawain and the Green Knight "Þat oþer schalk wyth a schunt þe schene wyth-haldeȝ." where it means to push or step aside, from "shove". Any middle-English scholars out there? The railway usage was a year earlier than the first electrical usage (1863) where it meant "The shunt is a wire connecting the two ends of the galvanometer coil." The term was first used in medicine in 1923.
"Choke" is fairly ambiguous as well!
 
Dan,

I think we need some more information from you, in order to provide useful suggestions. This seems like a long list of questions, but answers will help us better understand what you need. I'm sure even more questions will arise. Indeed, I keep thinking of one more, but best to see what you say before posing any more.

Is this project for volume production, or for one or two units?

Can a microcontroller be used, or do you need a firmware free solution? What microcontroller family would be best for you?

Does your system have a single common "ground" reference for all 10 DC supplies and the logic output? How is this arranged?

Are all ten DC outputs possibly active simultaneously? Each one needs its own separate 40mA threshold detector?

If shunts are used, how much drop at the 4A maximum is acceptable?

Can each shunt go on either the positive or negative side of the DC output it is monitoring?

Can the shunt be located on the input side of the circuit that sets the output voltage? So all of the shunts might have one end on a common positive or negative bus?

Someone has already asked if you actually need a measurement of the current, not just a high / low indication.

How accurate must the 40mA threshold be? Temperature stability? Is the 40mA value fixed, or might it need to be adjustable?

Do you need 10 separate logic outputs?

How quickly does the logic output have to react to a change in current?

Ted

PS: In this context, "Shunt" is exactly the correct term in present day EE jargon. In the USA. It is really a contraction of "Current Shunt". Each trade's jargon often is at odds with the civilian dictionary, as well as the jargons for other lines of work.
 

BobTPH

Joined Jun 5, 2013
8,806
Strange that I have never heard it called that when used the way you are. Always current sense resistor or just sense resistor. And always shunt when used across a current meter to multiply it's range.

Since the English language meaning has to do with diverting something, this makes sense to me. We use as small a current sense resistor as will work in order to avoid diverting a significant amount of current. Whereas, in the other case, the purpose of the resistor is to divert enough current to keep the meter within it's allowable range.

You can also make the case there there is not such thing as a voltmeter, all of them actually measure current. Or the opposite, all current meters actually measure a voltage across a resistor.

Edit: Okay, I did some searching of application notes from manufacturers and "shunt" and "current sense" seem to be used interchangeably. I stand corrected.

Bob
 
Last edited:

Ian0

Joined Aug 7, 2020
9,667
You can also make the case there there is not such thing as a voltmeter, all of them actually measure current. Or the opposite, all current meters actually measure a voltage across a resistor.
I'd say that they are all current meters. The movement of the pointer is due to the force created by a current in a wire in a magnetic field. Replace the wire with one of different thickness or of a different material and the deflection will remain the same. If the voltage remained the same the current would change and so would the deflection of the pointer.
 
Top