# Divide two different AC voltage

#### MrChips

Joined Oct 2, 2009
21,310
What are the frequencies of the two AC signals?

#### Romie77

Joined Jan 4, 2020
23
What are the frequencies of the two AC signals?
50HZ. Is it important if we do work on this IC?

#### MrChips

Joined Oct 2, 2009
21,310
50HZ. Is it important if we do work on this IC?
Just checking.
What happens when the denominator crosses zero?

#### RPLaJeunesse

Joined Jul 29, 2018
104
How fast do you need the result? Is it safe to assume always a sine wave? Might be better to gain up the signals a small bit (say to ~3V p-p) and use a simple micro to capture the signals via ADC, filter out the DC biases, determine the average AC values, and divide the results to get the ratio. An on-chip PWM or simple I2C DAC could make the result a DC voltage. (I'd try it with a "Bluepill" since the ADC will be faster than an those on an 8-bit Arduino, so better for "almost simultaneous" sampling.)

#### Romie77

Joined Jan 4, 2020
23
Both values varies from 0.1 to 3V AC.
AD532 is the only available IC, so I need to work on it.
I thought to simulate the result. But this IC is not present on 'multisim' (online) software.
Is there any free simultor available online?

#### crutschow

Joined Mar 14, 2008
25,114
What exactly is the output result you expect from this division?
Note that dividing by zero, when the sinewave goes through zero, is undefined and will likely saturate the chip.

I couldn't find an AD532 Spice model, but I do have an AD633 model, which is a similar device, for the free LTspice simulator program from Analog Devices/Linear Technology, which I use.

#### Romie77

Joined Jan 4, 2020
23
I think the way in which I had described the issue is not perfect. Sorry. Actually, I dont need to divide actual sine wave but I need to divide two AC values read by multimeter.

#### crutschow

Joined Mar 14, 2008
25,114
I think the way in which I had described the issue is not perfect. Sorry. Actually, I dont need to divide actual sine wave but I need to divide two AC values read by multimeter.
So how would you do that, other than with a calculator?

#### Romie77

Joined Jan 4, 2020
23
So how would you do that, other than with a calculator?
Exactly!! I was looking for a circuit to get output reading directly and bypass calculator uses.
I got the datasheet of AD532 which shows divider circuit. But I am unable to simulate it as I had discussed in post#6.

#### MrChips

Joined Oct 2, 2009
21,310
In that case, take the two AC signals, rectify and smooth to DC and then do the division.

#### Romie77

Joined Jan 4, 2020
23
I
In that case, take the two AC signals, rectify and smooth to DC and then do the division.
In this case, answer will be in DC itself. But required answer (division) should be in AC.

#### Papabravo

Joined Feb 24, 2006
13,924
I

In this case, answer will be in DC itself. But required answer (division) should be in AC.
You still have not answered the fundamental question of what you expect to happen when the divisor approaches zero from either direction. The answer is that you cannot make meaningful sense of the result. This fact makes your entire process suspect.

One more thing. Do you understand the meaning and implication of the phrase "two-quadrant"?

#### Romie77

Joined Jan 4, 2020
23
As I said before, I had describe the problem in wrong way. I dont want to divide actual sine wave. Hence no question of zero crossing raises in my mind. I just want to divide two AC readings read by multimeter without using calculator. Your suggestion to rectify signal is considerable.

I may dont have full knowledge about implication of two quadrant. But what we studied about quadrant is a graph divided into 4 part i.e. +x,-x,+y and -y. In electrical, single side rotation of motor can be called single quadrant while forward/reverse direction motor can be 2 quadrant.

I may misinterpret this division part of this IC. I am happy if you elaborate it for me. Thanks.

#### crutschow

Joined Mar 14, 2008
25,114
In this case, answer will be in DC itself. But required answer (division) should be in AC.
That statement makes little sense.
What do you mean the answer should be AC?
The multimeter displays a digital number that is a steady DC representation of the ACrms magnitude of the signal.
The answer from the divider will be a digital voltage representing the ratio of the two signal amplitudes, which can be converted to a digital number by an A/D converter if you like.
I may misinterpret this division part of this IC.
The IC will divide the two DC representations of the AC signal amplitudes.
If the divisor gets near zero than the output of the divider will saturate at its maximum value since the answer is becoming a very large value (approaching infinity as the divisor goes to zero.)
So what will be the minimum divisor signal level?

And what is that answer used for?

#### MrChips

Joined Oct 2, 2009
21,310
In this case, answer will be in DC itself. But required answer (division) should be in AC.
You just want a digital readout like what is shown on a DMM. Is that correct?
What is your definition of AC?
The readout can be scaled by any factor you choose.

#### Romie77

Joined Jan 4, 2020
23
Tha
That statement makes little sense.
What do you mean the answer should be AC?
The multimeter displays a digital number that is a steady DC representation of the ACrms magnitude of the signal.
The answer from the divider will be a digital voltage representing the ratio of the two signal amplitudes, which can be converted to a digital number by an A/D converter if you like.
The IC will divide the two DC representations of the AC signal amplitudes.
If the divisor gets near zero than the output of the divider will saturate at its maximum value since the answer is becoming a very large value (approaching infinity as the divisor goes to zero.)
So what will be the minimum divisor signal level?

And what is that answer used for?
Let me try to rectify both signal then give input to AD532. Answer in DC will be OK.

That answer is just to note down.

#### Romie77

Joined Jan 4, 2020
23
You just want a digital readout like what is shown on a DMM. Is that correct?
What is your definition of AC?
The readout can be scaled by any factor you choose.
I just want to divide two multimeter read outs (while knob on AC option on multimeter) without using any calculator.

#### Romie77

Joined Jan 4, 2020
23
Tha

Let me try to rectify both signal then give input to AD532. Answer in DC will be OK.

That answer is just to note down.
Minimum value will vary from 0.1V to almost 3V (while knob is on AC option on multimeter)

#### Papabravo

Joined Feb 24, 2006
13,924
The two quadrant restriction means both values should be positive and the divisor must be greater than zero, or both values should be negative and the divisor less than zero. To get what you want you need to square both signals, take the mean or average and then take the square root. This will give you the Root Mean Square value of the AC wave form and it will always be greater than zero. If you can scale and divide these two values you will have what you want. This procedure will work for non-sinusoidal waveforms. If you have have sinusoidal wave forms you can compute the RMS value by dividing the peak value by the square root of 2.

• atferrari