controlling AC load

Thread Starter

bug13

Joined Feb 13, 2012
2,002
Hi guys

I am looking for information about how to control an AC(240V) load from a lower DC voltage, says about 5V.

The AC load I want to control is/are heating element/s, the goal is to control the element through a MCU with thermal-couple feedback, PID loop in software.

  • control the load at a constant temperature
  • ramp up the temperature at a constant rate, say about 2-3 degree C per second
  • ramp down the temperature 3-4 degree C per second

I am thinking using solid state relay and controller it with PWM, but I am not sure if I can PWM AC.

so what is the usual approach to control an AC load like that?
 

GopherT

Joined Nov 23, 2012
8,009
Usually a triac or "silicon controlled rectifier" SCR is used to control AC. You are best off turning heaters completely on then off once set point is reached. The two devices I mentioned can also be used to Pwm the 50hz pulse of the AC. Read the Datasheets for info. An example would be Q601sr (from mu junk box - it is not an industry standard. Also a nice write up in Wikipedia for each device.

Sensitive gate type units can control from 5v DC.
Units are available for 15 amps under $1
 

#12

Joined Nov 30, 2010
18,224
ramp down the temperature 3-4 degree C per second
I think you mis-spoke. You can't "ramp down the temperature" without one of those cooler chips (Peltier). You usually just allow the heat to dissipate into the air with the power turned off.

Edit: I did one of these last week and found that the 3032 is cheaper than the 3031 or the 3033. (The lowest and highest required diode current are more expensive than the one in the middle. The difference between the 303x series and the 304x series is that the former has a voltage limit of 250V and the latter has a voltage limit of 400V. The MOC's are excellent chips because they have a zero crossing circuit in them.
 
Last edited:

Potato Pudding

Joined Jun 11, 2010
688
If you need to work with power plug AC, then keep in mind that a background working with electronics might make you more likely to get hurt because of the way the mind works and the habits that develop around low power electronics, that can get you killed when you forget what you are doing and try them with high voltage electricity.

For reasons of safety and keeping your insurance agency from disqualifying potential claims, develop a relationship with a certified electrician. Have an electrician drop by for a coffee, look at what you are doing, make the final connections, and sign off on it.

Knowing that you have questions is good. Make sure that all of the questions that need to be asked have been answered .

For example what kind of heating element are you talking about?
Resistive elements are common and you might be able to run one with PWM by conversion of your AC to High Voltage DC.

Other types of elements like an Inductive heater might need very special circuitry.

What type of element do you have or plan to use is the first question.
What wattage and max temperature is the next question.
 

#12

Joined Nov 30, 2010
18,224
You can PWM the power line, but you need to lower your expectations to about 50 Hz. Way cheaper than turning it into DC and trying to PWM at kilohertz frequencies.
 

crutschow

Joined Mar 14, 2008
34,459
Typically AC heaters are controlled by a low-frequency ON/OFF duty cycle of less than a cycle per second using zero crossing SCR circuits to minimize power line interference. An AC solid-state relay is the easiest way to control the AC power.

In effect you have a low-frequency PWM where the on period is at least one complete AC cycle.
 

Thread Starter

bug13

Joined Feb 13, 2012
2,002
I think you mis-spoke. You can't "ramp down the temperature" without one of those cooler chips (Peltier). You usually just allow the heat to dissipate into the air with the power turned off.
maybe it's my English problem (again), what I was thinking is, if I just turn the power off, the temperature might drop faster than I want, so I was thinking controlling the heating element/s, input some heat in the system so it doesn't cool down too fast.

or maybe I am just paranoid, I don't need that at all. :)
 

Thread Starter

bug13

Joined Feb 13, 2012
2,002
What type of element do you have or plan to use is the first question.
What wattage and max temperature is the next question.
Thanks for pointing this out, the heating element I am going to use is properly ceramic or quartz, I hope they are not inductive.

my guess the maximum temperature is going to be about 350 degree C, but I think most of the time the maximum temp. is about 300 degree C, I need to google a bit more to find out the exact temp. but those numbers are pretty close.

500W is more than I need, but I will properly get something between 1000W to 2000W (I don't have much control on this one)
 

Thread Starter

bug13

Joined Feb 13, 2012
2,002
Typically AC heaters are controlled by a low-frequency ON/OFF duty cycle of less than a cycle per second using zero crossing SCR circuits to minimize power line interference. An AC solid-state relay is the easiest way to control the AC power. .
I am a bit confused, so can I get a solid-state relay with zero crossing SCR in it? it looks like something I want.

when you mention zero crossing, are you referring the property of SCR, which cut off the electrical connection when the voltage is lower than the threshold voltage of the SCR?
 

#12

Joined Nov 30, 2010
18,224
Yes, you can buy a solid state relay with zero crossing included. Zero crossing means the triac refuses to start until the power voltage is zero. An AC wave is always changing amplitude and polarity, so when it hits zero volts, it is on the way to somewhere else. That is why it's called, "zero crossing". It's only crossing zero voltage on its way to more voltage (either more positive or more negative). Starting the triac at zero volts avoids the sudden change in current that would happen if the triac turned on at any other voltage.

and please, an SCR only works in one polarity, it's the triac that is used for AC.
 

tubeguy

Joined Nov 3, 2012
1,157
Light dimmers work by syncing to the AC sine wave and triggering at different points along the AC cycle to accomplish dimming/voltage control.
Google 'How dimmers work"

It would be helpful to have an input into your MCU which reads the AC wave and then sync to it. Then you could trigger at the same place every cycle.

You need to use an isolation device - either a transformer or an AC input optoisolator to accomplish this. If you are using a transformer for your control circuit power supply you can get the sync signal with some extra diodes placed before the main rectifier/filter capacitors.
 

Thread Starter

bug13

Joined Feb 13, 2012
2,002
Light dimmers work by syncing to the AC sine wave and triggering at different points along the AC cycle to accomplish dimming/voltage control.
Google 'How dimmers work"
what's the advantage of syncing to AC sine wave and triggering at different points, compares to normal PWM.

from what I can understand it, is this method can achieve a better control over the power output?
 

tubeguy

Joined Nov 3, 2012
1,157
When 'normal' PWM is used it's typically used with a steady DC voltage, which means that every pulse applied at a given duty cycle will produce the same average voltage output.

Now, imagine applying, for example a 50% duty cycle pulse to a sine wave.

If it is not synced exactly to the AC signal it will be impossible to have the pulse occur during the same portion of every cycle. Edit: So the 'pulse height' will be inconsistent, so the average voltage will be inconsistent,

A Triac, shuts off at/near the zero crossing point every cycle, and must be re-triggered on the 2nd half of the cycle.

This is an advantage, because for a give pulse width you only need to trigger the sine wave once at the same point each 1/2 cycle and the pulse will 'finish' automatically.

Edit: Clarify 'Same point' - For a 50HZ sine wave the period of 1/2 the wave is 10ms. So for a 50% duty cycle you would trigger at 5ms, then again at 15ms. For a 10% duty cycle you would trigger at 9ms and again at 19 ms.
 
Last edited:

tubeguy

Joined Nov 3, 2012
1,157
The AC load I want to control is/are heating element/s, the goal is to control the element through a MCU with thermal-couple feedback, PID loop in software.

  • control the load at a constant temperature
  • ramp up the temperature at a constant rate, say about 2-3 degree C per second
  • ramp down the temperature 3-4 degree C per second
You know, after thinking more about what you are doing, My opinion is that the PWM requiremnet is making this project more complicated than it needs to be.

Why?
Because heating elements are relatively slow to respond and so are temperature changes. I bet that simply turning a solid state relay on/off would work. You can still do this many times a second, but probably don"t need on/off 50m times a second. Just need one that can be controlled with 5VDC logic level input.

Just my opinion...
 
Last edited:

Thread Starter

bug13

Joined Feb 13, 2012
2,002
You know, after thinking more about what you are doing, My opinion is that the PWM requiremnet is making this project more complicated than it needs to be.

Why?
Because heating elements are relatively slow to respond and so are temperature changes. I bet that simply turning a solid state relay on/off would work. You can still do this many times a second, but probably don"t need on/off 50m times a second. Just need one that can be controlled with 5VDC logic level input.

Just my opinion...
That makes sense, 50Hz is too much, I haven't really thought about how many times to turn on/off yet, but I don't think it will be 50Hz.

What would you recommend?

At this stage, I am educating myself about SCR, SSR and triac.

edit:
SSR(with triac and zero crossing) are quite expensive that I expected, about $50+, am I looking at the wrong component?
edit2:
I am looking for 10Arms, 240V, 5V logic
 
Last edited:
Top