ADC concept for 500VAC to 500mV for sampling with harmonics, transient, flicker

[eebhoi01]

ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Hello Guys,

I am planning to start a new project, and right now I am doing concept hunting. I wish to gather information before I would dive in some experiments to minimize cost.

I am going to create a 500VAC to 500mV Analog to Digital (ADC) converter which I will be interfacing with an MCU. I do have an approach in mind which can attain this function, such as, using OpAmps or a combination of transformer and OpAmps. However the part I am having doubts is, will it capture Harmonics (THD and TDD)?

I understand that Harmonics occur starting as twice the fundamental frequency and above. My question really is will there be such ADC that its resolution is high enough to capture the harmonics.

Any idea, advice or shedding light on my troubled mind would really help me a lot. I am looking forward of hearing from you guys. Thank you

Regards,

EEbhoi

 

[barry]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

One word: Nyquist. If you don't know what that means then you probably shouldn't be messing with sampled data systems. When you ask "will there be an ADC that its resolution is high enough..." Resolution has nothing to do with it, sampling frequency does. But what harmonics are you expecting, 10th? 100th? You can add an anti-aliasing filter. You need to determine what frequencies you want to sample, then select your ADC accordingly.

 

[KlausST]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Hi,

Converting high voltage AC to a digital value is a frequently asked question here. There are many discussions. Read through them.
There are some with simplified hardware and there are some with high hardware effort. Each of them has benefits and drawbacks and they are discussed.

Developing an electronics circuit starts with specifications: Values with units, tolerance, precision...
"500V AC" tells almost nothing.
-->
It could be 50Hz, it could be 50MHz. It could be almost pure sine, it could be a mixture of a whole range of frequencies.
It cold be the expected fundamental RMS voltage, it could be the expected measurement range Vpp of the transients.

"Will it capture harmonics"?
--> it depends on frequency (harmonics frequency and sample frequency)
It depends on harmonics voltage and ADC resolution
It depends on analog filters (anti aliasing filter)
It depends on calculation method
...additionally you should ask: does the used circuit suppress existing harmonics or generate additional harmonics....and does this start a new problem?

As so often: Our answers can only be as good as your question (given informations).

Klaus

 

[eebhoi01]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Hi,

Converting high voltage AC to a digital value is a frequently asked question here. There are many discussions. Read through them.
There are some with simplified hardware and there are some with high hardware effort. Each of them has benefits and drawbacks and they are discussed.

Developing an electronics circuit starts with specifications: Values with units, tolerance, precision...
"500V AC" tells almost nothing.
-->
It could be 50Hz, it could be 50MHz. It could be almost pure sine, it could be a mixture of a whole range of frequencies.
It cold be the expected fundamental RMS voltage, it could be the expected measurement range Vpp of the transients.

"Will it capture harmonics"?
--> it depends on frequency (harmonics frequency and sample frequency)
It depends on harmonics voltage and ADC resolution
It depends on analog filters (anti aliasing filter)
It depends on calculation method
...additionally you should ask: does the used circuit suppress existing harmonics or generate additional harmonics....and does this start a new problem?

As so often: Our answers can only be as good as your question (given informations).

Klaus

Thank you, you got my point. I am not good really in expressing my ideas, specially on english. But yes, I would like to know if the circuit I am thinking about would suppress harmonics.

I have read some of the threads that talks about converting AC to digital value but none of them talk about harmonics.

Yes, I am sorry for being vague, I intend to gather harmonic data out of the AC power mains which runs at 50/60Hz. Having to think of converting the waveform to a digital one I believe I might suppress the Harmonic signal that I wish to measure. To simplify everything, its somewhat similar to a power quality analyzer.

To Sir Barry:
I also know the Nyquist theorem and anti-aliasing, it is because i understand these concepts, it makes me wonder how can I design an ADC in such a way that it will not suppress harmonics since waveform that has harmonics is distorted and I am not confident that by setting the right sampling frequency would be sufficient enough to prevent anti-aliasing. That is why I think more of the resolution and not the sampling frequency alone.

Anyhow, thank you for answers. Please do correct my wrong assumptions. ^_^

 

[FvM]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Front ends for mains voltage measurement are a serial topic at Edaboard. Besides 500 VAC range and sufficient bandwidth for harmonics measurement, do you need true isolation or can you afford some Mohm common mode "leakage" resistance. If so, a differential voltage divider would an economic and accurate method.

For me, 500 VAC refers to 50/60 Hz fundamental (unless stated otherwise). Useful range for grid analysis is up to 40th or 50th harmonic, respectively you'll end up with >3000 Hz bandwidth. ADC sampling frequency > 7 kHz, depending on the anti-alias filter steepness.

A standard µC ADC can acquire the signal with well considered software design and adequate anti-aliasing filter. For a good dynamic range, you'll want >= 12 Bit ADC resolution, an oversampling sigma-delta ADC can ease analog filter design considerably.

 

[eebhoi01]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Front ends for mains voltage measurement are a serial topic at Edaboard. Besides 500 VAC range and sufficient bandwidth for harmonics measurement, do you need true isolation or can you afford some Mohm common mode "leakage" resistance. If so, a differential voltage divider would an economic and accurate method.

For me, 500 VAC refers to 50/60 Hz fundamental (unless stated otherwise). Useful range for grid analysis is up to 40th or 50th harmonic, respectively you'll end up with >3000 Hz bandwidth. ADC sampling frequency > 7 kHz, depending on the anti-alias filter steepness.

A standard µC ADC can acquire the signal with well considered software design and adequate anti-aliasing filter. For a good dynamic range, you'll want >= 12 Bit ADC resolution, an oversampling sigma-delta ADC can ease analog filter design considerably.

Thank you. I am starting to get a bit more confident now. One more clarification sir, I will be using a transformer to bring down the 500V into a safe level to be inputted into an OpAmp circuit. Will harmonics, transient and flicker wont be suppressed by the transformer?

Or are there any other approach that will proportionally reduce the voltage level to 500mV while preserving the quality of the waveform and its harmonics for accurate reading?

My main goal for now is to come up a reliable means to convert 500V into 500mV level while as much as possible preserving the original waveform so I can accurately read the data.

Thank you so much for your patience.

 

[FvM]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

A transformer can easily generate new harmonics, thus I would avoid it. Attenuation of higher harmonics might also happen, depending on the transformer load. It's also quite difficult to get hold of a high quality measurement transformer.

Flicker measurement is mainly a matter of the fundamental frequency. no problem in this regard. Don't which kind of "transient" measurements you are referring here.

Why did you specify 500 mV level? Have you a specific ADC in mind?

 

[KlausST]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Hi,

A transformer can only work if your input (primary) signal is free of DC.

There are many different types of transformers. Power transformaers may not be suitable for signal analysis, because they may suppress/introduce harmonics. There may be audio signal transformers with better behaviour.

Without a circuit, without components, without specifications...we are not able to validate your idea.

***
Sampling frequency:
Nyquist requirement is relatively simple: your sample rate needs to be more than twice the max frequency of interest.
Anti-aliasing filter needs to suppress all signal components above half sampling frequency to a level that does not harm your measurement precision/accuracy.

Example: you want to measure 60Hz mains and it's 21st harmonics. So you are interested in 60Hz x 21 = 1260Hz max signal frequency.
Your sample rate needs to be higher than 2 x 1260 Hz.
Anti aliasing filter: It needs to pass frequencies up to 1260Hz with tolerable (you specify) attenuation.
And it needs to suppress all frequencies above f_sample/2 to a tolerable value (you specify)
Because there is no filter with step response (pass / suppress) ... you need to consider a (much) higher sampling frequency than 2520Hz.

Klaus

 

[eebhoi01]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Actually, I was interested with the AD9000 device. And as I read its manual I wanted to create some sort of development board out of it.

Based on its manual, I would need to feed its input with +-.707mVrms Max of voltage signal. I automatically think of using ADC but as what I am having right now, is the problem of harmonics might be suppressed.

 

[KlausST]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Hi,

AD9000:
* obsolete
* 6 bit resolution --> too low resolution for harmonics analysis
* 77MSPS sampling rate --> I'd say not suitable for mains harmonics frequency range
* ECL level outputs --> not a usual standard to play around with
* I assume extremely expensive

Klaus

 

[FvM]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

AD9000 is an obsolete high speed 6-bit ADC. You surely don't mean this device.

 

[eebhoi01]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

This is the reason why I wanted to ask first before purchasing stuffs to mess with. Anyhow thank you for your inputs.

But to wrap things up, assuming I will use this device. I will still have to reduce the voltage to a level this device can read. What is the best approach to convert such high voltage without suppressing any harmonic value? Thank you so much

Also, could you suggest a device that I can look into that is more recent. Thank you again ^^,

 

[FvM]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Differential voltage divider (e.g. dual 5 Mohm:5kohm) + unity gain differential amplifier.

Forget about using a 6-bit ADC.

 

[eebhoi01]

Re: ADC concept for 500VAC to 500mV for sampling with harmonics,transient,flicker

Differential voltage divider (e.g. dual 5 Mohm:5kohm) + unity gain differential amplifier.

Forget about using a 6-bit ADC.

I understand, you are suggesting to better design a ADC circuit myself rather than using the device. Thank you so much. I will update you guys with progress ^_^, thank you thank you.