60Hz, low level signal (<100mV) rectifier

Thread Starter

Haasenfefer

Joined Feb 7, 2014
2
I need help rectifying a very small signal, 10 - 80 mVac @ 60Hz. We are trying to measure and datalog an AC current. The current sensor is a Fluke i400s which outputs an AC mV signal proportional to the input current. Our data logger only accepts a DC signal. I've built a few precision rectifier circuits, which I found on the web, but nothing was stable at 60Hz.

I've tried using a LM324N op-amp and 1N5818 type diodes, with a wide range of resistors (1%) for the gain/feedback portion of the circuit.

Any direction, such as which op-amps, diodes to use/or avoid for this application and a schematic would be helpful.

Thanks!
 

t06afre

Joined May 11, 2009
5,934
Try to amplify the signal before using ideal rectifier circuit. You will need a dual power supply for all ideal rectifier circuits I have seen. Do you use this in your setup?
It could also be that the LM324 is not the best option in this case. It is much better opamps out there
 

crutschow

Joined Mar 14, 2008
34,281
Here's a precision full-wave rectifier that not only operates from a single supply but also uses no diodes.

What do you mean "nothing was stable at 60Hz"?
 

THE_RB

Joined Feb 11, 2008
5,438
Use two 1 meg resistors as a voltage divider on your regulated 5v rail etc, and connect that to the input of your DC datalogger.

Then just AC couple the signal to that point with a capacitor. I do this all the time for recording AC signals with DC ADC modules.
 

ScottWang

Joined Aug 23, 2012
7,397
Use two 1 meg resistors as a voltage divider on your regulated 5v rail etc, and connect that to the input of your DC datalogger.

Then just AC couple the signal to that point with a capacitor. I do this all the time for recording AC signals with DC ADC modules.
So as your method, do you need to divided by 1.414 to converting to the AC voltage in your progam?
 

crutschow

Joined Mar 14, 2008
34,281
Use two 1 meg resistors as a voltage divider on your regulated 5v rail etc, and connect that to the input of your DC datalogger.

Then just AC couple the signal to that point with a capacitor. I do this all the time for recording AC signals with DC ADC modules.
But if the data logger is designed for DC won't it give a varying (and undefined) output with an AC input as determined by its input circuitry and sample rate? :confused:
 

THE_RB

Joined Feb 11, 2008
5,438
So as your method, do you need to divided by 1.414 to converting to the AC voltage in your progam?
Generally the datalogger software is very "AC capable" even though the hardware expects a voltage in a "DC" range of 0v to ?v.

So you just log the waveform, and the software displays the waveform and will show stats like freq, p/p voltages, RMS voltage etc.


Crutschow said:
But if the data logger is designed for DC won't it give a varying (and undefined) output with an AC input as determined by its input circuitry and sample rate?
I'm not sure I understand your question?

The "DC" data logger records the DC voltage at instants in time.

So provided the AC waveform has been shifted by my suggested circuit to be within (say) +2v to +3v the datalogger will record lots of "DC" points between 2v and 3v. Then the software does all the math and displays the wave and all the data.

Obviously the sample rate needs to be 10 times (or more) faster than the AC waveform frequency to reproduce the wave with any reasonable accuracy.
 

ScottWang

Joined Aug 23, 2012
7,397
Generally the datalogger software is very "AC capable" even though the hardware expects a voltage in a "DC" range of 0v to ?v.

So you just log the waveform, and the software displays the waveform and will show stats like freq, p/p voltages, RMS voltage etc.
Thanks.
What you said about the input circuit, does it looks the same as the attched?
 

Attachments

crutschow

Joined Mar 14, 2008
34,281
............................................

The "DC" data logger records the DC voltage at instants in time.

So provided the AC waveform has been shifted by my suggested circuit to be within (say) +2v to +3v the datalogger will record lots of "DC" points between 2v and 3v. Then the software does all the math and displays the wave and all the data.

Obviously the sample rate needs to be 10 times (or more) faster than the AC waveform frequency to reproduce the wave with any reasonable accuracy.
I guess that was my confusion. If it's a DC data logger, how often does it take a sample and what is the AC frequency response of this "DC input". It would think that some of them may have a filter to suppress line frequency interference, thus they would not respond accurately to such an AC input. :confused:
 

THE_RB

Joined Feb 11, 2008
5,438
That's a good point and is information the OP didn't supply.

Most dataloggers I have seen are capable of thousands of data points per second, and I made an assumption there, and another assumption that the OP would have chosen a datalogger that samples much faster than the 60Hz mains freq if he wants to sample that waveform. But you're right, that's a lot of guessing.
 

Thread Starter

Haasenfefer

Joined Feb 7, 2014
2
A little more detail on the project at hand. We have a suspected faulty inverter (AC to DC, 125A output). The datalogger (500Vdc input rated) I'm using is sampling at 500ms, or 2 points per second. The reason for such a slow sampling rate is that we don't need to trace the waveform but just track the instance of any rise in; input AC current, output DC current and/or internal temperature of inverter. Currently the DC output current and temperature are being sensed without any issues by our datalogger.

We're limited to the Fluke i400s clamp for the input AC current measurement, so our secondary signal will be very low: 1Aac=1mVac (400A setting). Our nominal primary current is 25A so that =25mV @ 60Hz (line frequency).

My idea was;
Convert the AC primary current to AC secondary mV signal (via i400s clamp),
convert AC secondary mV signal to a DC mV signal (via precision rectifier),
then input DC mV signal into datalogger (read mVdc),
measure AC primary current (with clamp-on meter),
calculate ratio between AC primary current to DC mV signal,
have datalogger do the math and display/log the input AC primary current, when triggered.

I've just received some sample TI op-amps that were suggested in one of the earlier posts, so i'm going to construct that circuit tonight.

Thanks for all the help so far everyone!
 
Top