Analog zero-point synced AC PWM: is there a better way?

[Archie Medes]

Hi,

Thinking of a circuit to control the AC power to a heating element. I wanted something fairly linear, and analog. I don't have anything against using a programmable controller, I just wanted to think of an analogue solution, seeing as I don't need to learn how to write a program.

The attached image shows the flow diagram and circuit sub-section schematics. It works in my simulator.

Just wondering if there is a better way to go about this? Switching on/off whole/half waves in a duty cycle?

Cheers,

Archie

*edit - that's a 555 IC in the pwm section

 

[rbelli1]

You can get opto isolated drivers that are zero cross detecting (the first one I found on DigiKey: https://www.digikey.com/product-detail/en/lite-on-inc/MOC3063S-TA1/160-1723-1-ND/670018). You want a driver in any case to shield you from high voltages.

Your circuit will be much simplified and you can set your PWM frequency to seconds or minutes depending on the thermal mass of the heater. You get less EMI too. Don't use the zero cross detectors on inductive devices (Motors, Transformers, etc.).

BoB

 

[Archie Medes]

Thanks for your reply BoB, I will check out the drivers.

I can set PWM period with the pot and cap on the trigger/threshold input of the 555. 1 second with half-waves gives me 100 points of resolution at 50 Hz. Unless you mean that I can vary the period without altering the circuit, which would be great. That was the next part I was going to try to do.

I'm unsure of what you mean by a few things:

You want a driver in any case to shield you from high voltages.
Which part needs shielding from the high voltages, and what is the source of them?

You get less EMI too.
By lengthening the period? What kind of periods are troublesome for switching AC waves on/off at zero point?

Don't use the zero cross detectors on inductive devices (Motors, Transformers, etc.)
So, if I am taking 240VAC, inputting it to a magnetic transformer, which outputs a low voltage AC wave to be rectified into DC for the controller, then I shouldn't use a zero point crossing on the 240V signal? Or do you mean that I shouldn't zero-point detect on the stepped down AC signal, because it will be out of phase?

*edit, actually, you mean because the current will be out of step with the supply voltage on the load, right?

I understand AC motors need a variable frequency drive, something else I need to learn. This is purely for driving resistive heating elements

 

[rbelli1]

I can vary the period without altering the circuit

The period is set by the RC circuit on the 555.

Which part needs shielding from the high voltages, and what is the source of them?

The high voltage comes from the AC voltage you want to control. If the voltage you are controlling is referenced to logic ground you may be able to get away with no isolation. This puts your logic ground at neural voltage which may present a shock hazard in some instances. Triac optos will work. There are also purpose built triac drivers with various advanced features. In any case you need more than the NPN output of the 555 to properly drive the triac.

The EMI comes from switching at non-zero potentials. If you put proper filtering on the output it can be made an acceptable level. That is what standard triac dimmers will do. However eliminating it is better if that is an option.

The zero-cross in regards to the inductive loads is for the load. That does not apply to a transformer driving a resistive load. With inductive loads you get the lowest inrush if you switch at peak. For small loads it is not so important. If you are designing for that you would use a non zero-cross driver (random trigger it is sometimes called) or use a peak trigger for large loads. I have seen mechanically switched transformers in the several hundred kilowatt size. The transformers would make quite a racket when the breaker closed at close to zero. My comment was more of an aside. Disregard it for your heater project.

For AC motors you can use a triac dimer type speed controller for certain kinds. I know that universal motors and shaded pole motors will work. Other types I don't know.

When designing PWM controllers with 555 timers you will often get some frequency variation as well as percentage modulation. This is completely OK as long as the frequency stays low enough to maintain the desired resolution and high enough to not cause overheating during the on time. Be sure to keep the charge and discharge currents with acceptable datasheet ranges.

BoB

Edit: If you are full wave rectifying the AC into your load then you can use the output of the 555. I forgot they are totem pole outputs.

 

[Archie Medes]

Cheers for the detailed explanation BoB

The period is set by the RC circuit on the 555

Yeah, that is what I meant by setting the period with the pot and cap on the 555 comparator inputs. I would prefer a controller that gave me both period and duty control, and was wondering if your suggested device offers that (I haven't got around to looking at it yet)

In any case you need more than the NPN output of the 555 to properly drive the triac

The 555 output is driving the D input on the flip flop, the Q output is driving the triac gate, but yes, it is just a lazy connection at the moment, is should drive something else to power the gate. I'm not across the foibles of triacs yet.

This puts your logic ground at neural voltage which may present a shock hazard in some instances.

This is the kind of thing I really need to understand (I don't). My education had no focus on designing power supplies or motor controllers - AC or DC, the only design focus was on logic. I really need to find myself some good textbooks/online courses on the subjects.

My comment was more of an aside. Disregard it for your heater project.

A very informative aside, thank you.

Be sure to keep the charge and discharge currents with acceptable datasheet ranges.

Thanks again, hadn't calculated the discharge through the 555.

 

[dlgoff]

I've use this awesome UAA2016 IC for a temperature controller. Here's the circuit.

edit: BTW works well at 120 vac.

 

[Archie Medes]

Cheers dlgoff, that looks pretty handy