Differential Amplifier-Isolations voltages

Dear All

I have an basic doubt of the Differential Amplifier.

What exactly it works..??

how much isolation it gives..?

Can i use the AC Voltage measurement by using this..??

Kindly Reply
Marx

- - - Updated - - -

https://obrazki.elektroda.pl/3692871600_1467007943.png



My circuit is like this.

isolation vs Difference Amplifier

Dear Friends

I had gone thru the Both circuits Of Isolation and Diff amplifiers.

Isolation: Where i am getting optically and Pure Isolatated signal with the help of Optocouplers for the sensing input voltages and to Measures by controllers.

Difference amplifiers: Where i am taking the Live and Neutral signal where my neutral and diff amp Ground point is not attached anywhere but i am getting clean signal at the Output of Diff amplifier.

Now i am getting confused what to use and why..??
Diff amplifiers is cheap but optically isolated is Costly..

But i am not confident in using Diff Amplifiers..
Kindly let me know can i use the Diff Amplifier to sense the line voltage and give it to the Controller fro processing.

yours Feedback is appreciated.

Marx

Re: isolation vs Difference Amplifier

0.25 W style resistors are often rated for 250 or 500V thus two series.
This diff. amp scales down the AC voltage but your + input reference for 0V in , is 5V and it's DC gain is 2 while the scale down AC differential amp gain is unity (1), thus the null input gives 10 V out with a 12V unipolar supply. Not good.. Better is use equivalent 22k using Vcc/2 out . Thus requires the input reference of Vcc/4 with your gain if 2 on (+) input so 12/4=3 V. and Req with R divider must be still 22K for unity AC gain with bakanced inputs. OK still with me?

... for the next stage precision AC to DC full wave rectifier Op Amp. to get 0 to x V DC for AC Peak 0 to root(2) * Vac RMS input. If you want to measure average rectified AC and convert to RMS use correction factor ( I recall, check) pi/(2root(2) using R ratios or software, or measure both peak smooth, peak and hold or average, you decide.

There may be better ways to do it as well. Start by a table of precision AC inlut values and DC precision values and ensure Vcc does not affect the output much by regulation and gain matching.

sorry no more time...

Isolation requires special matching to get accuracy differential matched gain to get 3kV or more isolation, unless your DC is isolated. Also RF noise requires an input Balun to reject RF noise.

Sunny:
Looking out basic difference of both.

Isolation vs differencial Amplifier for ac voltage sensing with PIC controller..?

Thanks
Marx

Hi,

General descriptions of differential amplifiers you may find a lot in the internet. Therefore I just comment the given circuit.

***

The differential solution is not truely "isolated", but it uses high ohmic resistors limiting the current to a safe value.
But you need to keep on local safety regulations about what devices to use and what clearance (creepage, air gap) you need..

I like the differential solution (and used it many times) because it is more precise. The drawback is the limited common mode voltage. But with your resistor values it should be no problem.

****
Against SunnySkyguy I don´t think it is gain of 2.
I think the OPAMP output is shifted by 5V DC.
Then it´s divided by 2 because of the two resistors.

This is the point I don´t like that much.
The output of the opamp is clean and low impedance - wich is good - but the two resistors generate an 5k impedance.

With the given values: 109V differential input voltage give 1V OPAMP output voltage.

Usually one wants to avoid noise, especially higher frequency noise. Therefore one uses low pass filters.
But in your case the capacitors with the 2k2 and the 22k form a high pass filter. Noise (above fc = 30kHz) generated by the OPAMP and those three resistors is multiplied with gain of 11.

Klaus

There are at last two meanings of isolation here. One for safety from lightning transients and one for common mode noise isolation between the AC and DC ground.

Normally AC transformer line is earth grounded to what is called Neutral (N) while the other line (L) is hot (H). There can be considerable current pulses on ground from various household electronic line filters so the difference can interfere. We call this a common mode signal (CM) because it is common to 5V , yet it can create a differential voltage with the AC connected to the external ground. Therefore Common Mode rejection over 1MHz bandwidth is important. A differential Amplifier has excellent (>100dB) rejection ratio or CMRR up to 5Hz or so then reduces as the internal gain of the Op Amp and thus its feedback is reduced to 1 at the BW limit and further. The safety isolation is reduced by attenuation but is limited by the dust + humidity on the 1M resistors and voltage breakdown rating of a clean part often 500V. Therefore it has relatively poor safety isolation compared to an optical isolated rated for 3kV. Meanwhile hydro meters only have arc suppression for 6kV so some line filtering is often assumed to be part of the design to reduce it's potential from exceeding 3kV. So the Differential Amp has worse safety isolation and CM noise rejection at 1MHz for AM Radio interference than the optical isolator.

However the optical isolator is not a perfectly linear device with the transfer function of LED current to photo diode current having a limited linear dynamic range. It depends on having a large resistor to linearize the diode nonlinear V vs I characteristic.

In both cases you want to full wave rectify the AC input to a DC level to measure the AC level unless you are planning on a precision AC rectifier using an Op Amp after the AC differential amp.

Often DMM users get false readings for AC voltage or current because AC inverters and pulsed signals are measured in the meter by Vpp then scaled by assuming it is a sine wave to RMS. You must decide if you wish to measure Vpp x k where k is a scale factor for either average to RMS or peak to RMS and choose a time constant to average either value. RMS is truly only a measured in heat due to pulse duty cycle or sinusoidal phase shift in motors causing reactive power factor errors in tryue power measurements.

I agree Klaus on the 5V. It is a unity follower from + input (with 1% offset) if line is grounded.

It is hard to explain both non-ideal design flaws and fundamental differences in isolation in one answer.

Look for Instrument Amp design for the best CMMR and also Balun design for CM + HF noise rejection on AC lines. It is basically a differential low impedance 1:1 choke that raises the Common mode impedance to isolate noise from getting converted to a differential signal by high inductive impedance to lower capacitive ratios. THus improving CMRR more than an Op Amp can do. The Opto meanwhile has excellent common mode (CM) frequency isolation and thus high CMMR, since there coupling capacitance is very small.

One must learn to find the critical and important specs for any design requirements before making any choices in parts by writing a list of all assumptions or write a spec. to define what is expected for inputs and outputs and what conditions must not interfere with the results. ( like 3kV or 6kV lightning impulse noise for both the safety of the parts and the human touching it.)