I hope this is a simple question with a simple answer!
I am building an Arduino based heater control circuit, the final load of which will be a 6kW Heatrae Sadia electric boiler. At the moment the control is based on using a relay to interrupt the call from the room thermostat; a simple on-off action.
The next iteration of this project will be to measure the room temperature directly into the Arduino, and control the heater output via a triac. I have read an artical in the Playground -
This article tells me that the triac has a minimum gate pulse width to ensure it turns on, and I understand that part. The implication is that the gate pulse must EXCEED the minimum time (as given by the triac spec, I suppose).
My question is this (you may wish to refer to the 'Theory of operation' in the link above), for very low outputs I must fire the triac very late in the mains half-cycle, so is there a danger that the gate pulse will overlap into the next mains half cycle and turn on the load fully during the second half cycle...and the inferrence is that demands for heat below 2% or so will result in 100% heat output because of this overlap?
My question is this (you may wish to refer to the 'Theory of operation' in the link above), for very low outputs I must fire the triac very late in the mains half-cycle, so is there a danger that the gate pulse will overlap into the next mains half cycle and turn on the load fully during the second half cycle...and the inference is that demands for heat below 2% or so will result in 100% heat output because of this overlap?
For a heater you don't use phase control, you use burst firing. Heaters take ages and ages (well, seconds or 10s of seconds, maybe minutes) to heat up and cool down, so you don't need to control 100 (or 120 depending where you are) half cycles every second. Instead you have whole cycles but control how many, so for half power you would turn the triac on for 1 cycle then off (not turn on) for 1 cycle, for quarter power on for 1 cycle, off for 3. Note I said 1 cycle, not 1 half cycle. You need to do this in whole cycles otherwise you are rectifying the mains to some extent. You must switch at zero crossing or very close.
Using phase control generates electrical noise due to the rapid switching part way through a cycle, the noise should be suppressed with inductors, big inductors for 6kW. Switching whole cycles means almost no noise as you are switching at zero crossing.
Thank you both, I consider that to be case closed and it's burst fire for me for my heater project...
...but suppose...
...well, suppose I had a project that used phase angle control of an output (this is purely for my own understanding at present). If I triggered the gate so late in the half cycle that the trigger pulse overlapped into the next half cycle then would it be true that very low output commands could give rise to full power?
Glorymill:
Thank you both, I consider that to be case closed and it's burst fire for me for my heater project...
...but suppose...
...well, suppose I had a project that used phase angle control of an output (this is purely for my own understanding at present). If I triggered the gate so late in the half cycle that the trigger pulse overlapped into the next half cycle then would it be true that very low output commands could give rise to full power?
Yes.... You have grasped the concept of triac phase control.
Tom...
Using phase control generates electrical noise due to the rapid switching part way through a cycle, the noise should be suppressed with inductors, big inductors for 6kW. Switching whole cycles means almost no noise as you are switching at zero crossing.
Not true when the load is purely resistive. You might create some voltage distortion with a high impedance source, but that's about the worst that would happen.
OTH, if you wanted to create problems... you could add an inductor...
Not true when the load is purely resistive. You might create some voltage distortion with a high impedance source, but that's about the worst that would happen.
OTH, if you wanted to create problems... you could add an inductor...
I respectfully disagree, and I'd like to know why you think that, as I might learn something.
At 230V (UK and several other countries) a 6kW heater draws 26A RMS, 36.5A peak. If the triac switches at the peak then the current goes from 0A to 36.5A almost instantly. Rapid transitions show up as high frequencies so rapid switching generates high frequency noise, which an inductor will limit.
At the moment the control is based on using a relay to interrupt the call from the room thermostat; a simple on-off action.
Virtually every heating/cooling system works that way. You don't need to "dim" a heater because temperature doesn't change instantly.
In rare cases you might find low-medium-high settings , usually with multiple heating elements (or multiple gas jets) switched on & off. But this can get tricky in a feedback-controlled system (i.e. a thermostat). It's easier to make a stable system with on/off control (with some hysteresis).
I assume you're using optical isolation? If you're using phase (or cycle) control, of course the zero-crossing/phase detector also has to be isolated.
There are zero-crossing opto-isolators or zero-crossing solid-state relays so if you use "traditional" on/off controls you can get quiet(er) switching without phase-detection.
Ideally you'd tailor a snubber circuit to the load being switched, just an inductor will slow down current rise but generate high dV/dt edges at switch off instead, which also cause a lot of interference. Zero-crossing switching a resistive load is so much less problematical, but needs the complication of a zero-crossing detector circuit.
SSR's can do the job for you as well as giving opto-isolation, just pick the right one...
Remember there is ALWAYS an inductor involved with AC power distribution. At a minimum there is the transformer supplying the reduced voltage from the power distribution system. The chopped energy is not directional.
Glorymill:
f I triggered the gate so late in the half cycle that the trigger pulse overlapped into the next half cycle then would it be true that very low output commands could give rise to full power?
True.
But then the gate signal is in the tune of 10 us.
A half wave at 50 Hz is 10 ms.
So this becomes only an issue if you try to reduce the power to <1%, so that's easily prevented to set a minimum power level in your code. I've used <3% is off; >97% is on. Those very high and very low percentages have very little practical use anyway, if you routinely have to supply power at <1% you have to get a lower powered appliance.
