PWM of AC?

Thread Starter

germeten

Joined Apr 18, 2014
26
Hi, I'd be interested to build a PWM adjustment/detector of the high points of an AC wave. I tend to think this would require a flip-flop and bi-directional peak detector of some kind that can be spread out from the highs to encompass greater or lesser amount of the wave. Can anyone point me in a direction where to start with such a project, or if something exists already? I did find a Chinese researcher who wrote a paper on such a generator but it's tied up in the legal copyright world of pay-for-me.

I found this from another thread of similar topic, but am having trouble making full sense of it or how to incorporate it.

http://www.realtechsupport.org/UB/MM/practical things/AC_powerlines.pdf
 
Last edited:

ian field

Joined Oct 27, 2012
6,543
Hi, I'd be interested to build a PWM adjustment/detector of the high points of an AC wave. I tend to think this would require a flip-flop and bi-directional peak detector of some kind that can be spread out from the highs to encompass greater or lesser amount of the wave. Can anyone point me in a direction where to start with such a project, or if something exists already? I did find a Chinese researcher who wrote a paper on such a generator but it's tied up in the legal copyright world of pay-for-me.

I found this from another thread of similar topic, but am having trouble making full sense of it or how to incorporate it.

http://www.realtechsupport.org/UB/MM/practical things/AC_powerlines.pdf
Its hard to make out what you're describing - but sounds a bit like triac phase control.

If you're trying to do peak detection, its fairly difficult. For a degree or two of phase angle, its almost DC at the tips of the peaks.

If the frequency is stable; you can detect the zero crossing point more precisely and predict the peak point with a fixed delay.
 

Thread Starter

germeten

Joined Apr 18, 2014
26
pwm waveform.jpeg pwm waveform.jpeg
If the frequency is stable; you can detect the zero crossing point more precisely and predict the peak point with a fixed delay.
OK, any idea how to go about that? I believe the link I posted describes that but it's still greek to me. I assume the detector/delay chip must power a mosfet and not sure how to incorporate both to run on and pass AC. Ideally I'd like to catch the high points with even distribution of pulses on both sides of peak (adjustable.) I don't think TRIACS can do that.
 
Last edited:

GopherT

Joined Nov 23, 2012
7,983
If you mean why bother responding, I couldn't agree more.
I think the "why bother" means, what advantage would your wish have over what is done with triacs? If you have a reasonable answer, then it may be worth the bother. Until then, why bother with the extra complexity other than the fact that it makes a more symmetrical wave and some people deem that to be prettier than a triac option.
 

Alec_t

Joined Sep 17, 2013
10,362
Why? What is the ultimate purpose? Is it to control power in a load?
 

Papabravo

Joined Feb 24, 2006
12,405
The "why bother" is because I don't think there is a single device, or a simple circuit that will do what you want it to do. Could some complex collection of devices and circuits do the job? Maybe, but nobody can see the purpose of developing such a circuit. If you can't articulate a reason then

 

ian field

Joined Oct 27, 2012
6,543
View attachment 93908 View attachment 93908

OK, any idea how to go about that? I believe the link I posted describes that but it's still greek to me. I assume the detector/delay chip must power a mosfet and not sure how to incorporate both to run on and pass AC. Ideally I'd like to catch the high points with even distribution of pulses on both sides of peak (adjustable.) I don't think TRIACS can do that.
It would be hard to do that other than threshold detection on both rising and falling slopes - and if the sinewave amplitude isn't absolutely stable, that would be futile.
 

crutschow

Joined Mar 14, 2008
23,349
I want to be able to PWM an AC wave.
Then for us to invest time and effort in helping you, we need a reason as to why the PMW approach is better then the standard SCR approach.
Or is this just an academic exercise?
 

ian field

Joined Oct 27, 2012
6,543
The "why bother" is because I don't think there is a single device, or a simple circuit that will do what you want it to do. Could some complex collection of devices and circuits do the job? Maybe, but nobody can see the purpose of developing such a circuit. If you can't articulate a reason then

As a theoretical exercise, its mildly entertaining - but only till the novelty wears off.......................

What I've suggested so far is; zero crossing detection and delays to generate a trigger at a phase angle past the zero crossing.

A more complex solution would be a PLL frequency multiplier.

Set the VCO to some binary multiple of the AC frequency and divide it down to the AC frequency to feed into the phase comparator alongside the AC.

The VCO running at some multiple of the AC frequency provides a chain of clock pulses that can effectively be used to divide up the AC wave into segments. From this point the complexity starts loose. For high resolution (lots of segments) I'd suggest an EPROM lookup table - feed a binary word to the address pins to select a particular pattern of bits to select switchin points on the AC wave.

For low resolution a diode matrix may be sufficient instead of an EPROM lookup table.
 

tcmtech

Joined Nov 4, 2013
2,808
I use a similar method for Grid Tie Inverter operation, to convert DC into AC and push power back into the AC lines, to set the zero cross dead band lower limits and prevent over current issues with the switching devices.

It's pretty easy to do with nothing more than basic op amp comparator circuits.
 

Thread Starter

germeten

Joined Apr 18, 2014
26
As a theoretical exercise, its mildly entertaining - but only till the novelty wears off.......................

What I've suggested so far is; zero crossing detection and delays to generate a trigger at a phase angle past the zero crossing.

A more complex solution would be a PLL frequency multiplier.

Set the VCO to some binary multiple of the AC frequency and divide it down to the AC frequency to feed into the phase comparator alongside the AC.

The VCO running at some multiple of the AC frequency provides a chain of clock pulses that can effectively be used to divide up the AC wave into segments. From this point the complexity starts loose. For high resolution (lots of segments) I'd suggest an EPROM lookup table - feed a binary word to the address pins to select a particular pattern of bits to select switchin points on the AC wave.

For low resolution a diode matrix may be sufficient instead of an EPROM lookup table.
Your first method is described in any earlier thread on AAC, similar topic, others here have already contributed to. I found some link references to delay circuits while theoretically simple, aren't easy to execute for beginners like me. Frequency doublers and comparators are even more complex, and cumbersome. Somehow I feel instinctively that detecting voltage level for on/off switching is more straight-forward. Simple dialing up or down the voltage detection would automatically capture more or less of the wave. The challenge is it working under either polarity. And lastly, I'd like to keep it as small as possible even while passing big loads.
 

GopherT

Joined Nov 23, 2012
7,983
Your first method is described in any earlier thread on AAC, similar topic, others here have already contributed to. I found some link references to delay circuits while theoretically simple, aren't easy to execute for beginners like me. Frequency doublers and comparators are even more complex, and cumbersome. Somehow I feel instinctively that detecting voltage level for on/off switching is more straight-forward. Simple dialing up or down the voltage detection would automatically capture more or less of the wave. The challenge is it working under either polarity. And lastly, I'd like to keep it as small as possible even while passing big loads.
Still no justification for trying to do this? Novelty has worn off for now. Thanks for stopping by.
 

Thread Starter

germeten

Joined Apr 18, 2014
26
I use a similar method for Grid Tie Inverter operation, to convert DC into AC and push power back into the AC lines, to set the zero cross dead band lower limits and prevent over current issues with the switching devices.

It's pretty easy to do with nothing more than basic op amp comparator circuits.
I'd be interested to see how you do this.
 

Thread Starter

germeten

Joined Apr 18, 2014
26
Still no justification for trying to do this? Novelty has worn off for now. Thanks for stopping by.
"Justifications" you're funny. Why not talk about prohibitions? Makes as much sense. If you don't know how to do something, or don't want to, why not leave it to someone who does? Please don't strain your brain on my account.
 

djsfantasi

Joined Apr 11, 2010
5,601
"Justifications" you're funny. Why not talk about prohibitions? Makes as much sense.
Because there is not always one answer to our questions. Without knowing the "justification", it may not be possible to identify which answers apply. Otherwise, simply reverse engineer a household dimmer. I too am interested in this question, so that I may control AC from a micro. So let's not let this get personal.
 
Top