Amplifier for DC voltage - suggestions please

Thread Starter

stdave

Joined Jun 25, 2017
6
I need to increase the voltage of a signal (little current here as the voltage is just being recorded) from around 70mv to around 1 or 2 volts. I see there is a ton of op-amps at digikey - but I have no idea which one to use. Of course it must amplify DC - any suggestions would be helpful. If it helps, my application follows.

I'm monitoring all three phases of an electrical service (maybe even the neutral - so that would be four). I want to monitor current in the range of 5a - 100a. I'm using current clamps that I purchased on eBay. I want to read it with a small device (DataQ DI-149) that connects to my computer so that I have a constant log - but the DataQ reads only from 0 to 10 volts. I would change the signal from the clamp to DC (via bridge) so that it is easier to read with my Dataq voltage reader. The problem is that the current clamp only outputs a maximum of 10mv per amp and that will barely register on a device designed for 0-10 volts. If I'm reading say 8 amps with the clamp then the clamp output is only 80mv - less than a tenth of a volt. I would also like to vary the gain so that I can more accurately monitor higher or lower current (so the display in my monitoring graph doesn't top-out or bottom out.

I know there are very popular versions of certain products (ie 555 timer) that people have used for years - maybe that's what I'm looking for... but of course the op-amp standard of the same idea? Some product or circuit where I can get started...
 

AnalogKid

Joined Aug 1, 2013
10,987
A 555 will not work in this application.

You mention a bridge. Forgetting gain for the moment, is the output from the current clamp a DC signal proportional to the AC current, or an AC signal that needs to be converted?

Link to the clamp, datasheet, photos, anything.

ak
 

Thread Starter

stdave

Joined Jun 25, 2017
6
A 555 will not work in this application.

You mention a bridge. Forgetting gain for the moment, is the output from the current clamp a DC signal proportional to the AC current, or an AC signal that needs to be converted?

Link to the clamp, datasheet, photos, anything.

ak
Thanks for your response - The output of the clamp is AC voltage - with the small voltage output relative in amplitude to the current being measured. About the bridge - I'm thinking that I would use a diode bridge to convert the AC to DC - but that may not be absolutely necessary - the DataQ actually reads both +10v and -10v. It creates a real time display with a 0V center line graphing time vs amplitude while you watch and records it to a file. I just thought converting the AC clamp signal to DC would create a more pleasant read on the graph.
 

MrChips

Joined Oct 2, 2009
30,712
TS was using reference to 555 only as an example of a ubiquitous circuit.

Is there a similar part in the amplifier function? Yes, there is, and it is known as the μA741, LM741, or just plain 741, and various other aliases.
This was one of the first frequency compensated and stable IC opamps that appeared in the mid 1960s. This is a model IC that is used in just about every textbook on operational amplifiers. So we are looking at an IC that was designed over 50 years ago!

If you follow the various discussions on this site you will be reminded not to use a 741 in a new design.

Before you select an op amp, you need to learn about op amp basics and common op amp circuit topology.

Here is your basic non-inverting op amp circuit:


What IC to choose? There are tens of thousands of op amp ics available. You can get one, two or four opamps in a single package.
One can write a 200 page manual on the criteria on choosing an opamp.

LMC6081 is a single op amp in an 8-pin package. Why did I select this one? Because it is the first hit that appeared on a google search. What other ways do you know on how to choose an op amp? I can list about 20 ways!
 

#12

Joined Nov 30, 2010
18,224
You can use a 741, but the results will be fairly bad. First, you need a dual power supply of about +7 volts and -3V (minimum) to get a +2 volt output. So why make a 10 volt supply when you can get a chip that will work on 5 volts? Then the input offset voltage of (up to) 6 mv is an 8.6% error on your 70 mv input, and a lot worse if you are starting with less than 70 mv. You can add a pot to null that offset, but why bother when you can get a chip that doesn't need to be corrected?

If you already have a fairly wide dual power supply on board and don't need better than 10% accuracy, a 741 is a pretty good DC amplifier that can almost get to audio range frequencies if it only has to produce 2 volts peak (slew rate limited to .25 V/usec).

With that level of performance, it's no wonder why discrete audio amplifiers were still being made in the 1970's.
 

AnalogKid

Joined Aug 1, 2013
10,987
Since the output from the current sensor is AC, not DC, you actually are looking for a low frequency AC amplifier circuit. The 741 has adequate bandwidth for this and the effects of its offset voltage are removed by the circuit, but it has a relatively high noise component. There are many more contemporary opamps with 1% of the noise.

Depending on your country, the signal out of the current clamp will be 50 Hz or 60 Hz AC. I doubt that the real-time display will show this distorted sine wave accurately, and even if it did I don't think the result will tell you what you want. So part of the interface circuit should be some form of an AC to DC converter.

A simple bridge rectifier arrangement will not work for this with a 70 mV signal because the minimum voltage for conduction is around 0.6 V for regular small signal diodes and 0.2 V for Shottkey diodes. Also, if the bridge were after the gain stage it would introduce a negative offset into the results (two forward voltage drops of the diodes) that changes significantly with temperature.

Since you are not worried about half-cycle transients, an active rectifier circuit acting as a peak detector should be ok. You might not be familiar with any of these terms. If not, read up on them and we'll answer your questions.

An alternate approach is to digitize the AC output from the gain stage, extract the peak value in software, and display that.

ak
 
Last edited:

JWHassler

Joined Sep 25, 2013
306
Depending on your programming expertise/backup, some of the work could be pushed onto the DAQ DI-149 (AC-DC conversion, some of the scaling, some of the filtering)
An amplifier tuned to 50/60 Hz is needed, though
 

AnalogKid

Joined Aug 1, 2013
10,987
A classic circuit. Add a series feedback capacitor for high frequency noise rolloff, and an output holding capacitor, set the forward gain to whatever is appropriate for the range, and everything is done in one stage.

ak
 
Top