H11aa1 as ZCD

[Kirigozo]

Hi folks,

I've been looking at the application of H11AA1 optocouplers as zero crossing detectors and I can't see how they can possibly work. The data sheets are no help here at all.

Looking at the sequence of operation the mains input cycle goes positive, turn the led on and drives the transistor on quickly. As the input goes negative the first led extinguishes and the transistor powers down slowly. In the mean time the other led begins emitting and turns the transistor on again quickly.

From the data that is available it looks like any change in the output state of the transistor is more likely a glitch and not a reliable state indicator.

Has anyone used this device for ZCD or is this application just an internet myth?

[David_AVD]

The LEDs will be off during the zero crossing as the voltage dips under the Vf value.  Yes the pulse is relatively narrow, but that's what you want.

[Kirigozo]

The LEDs are not the problem. The iffy area is the transistor switching times. Turn on is only 1.5us but turn off is 20+us.

My math is not good enough to compute the time for the 1.5v Vfd transition.

[Circuitous]

I have used the H11AA1 to detect AC line voltage, and it should be fine for tracking zero crossings. 

In my case, I wanted to just detect the presence of power so I had to smooth the output by connecting a 100nF between the base and emitter.  That gave me a nice flat output for a micro to read.  Without the capacitor I was picking up the zero crossings, which I didn't want.

[djacobow]

I've built a bunch of dimmer type circuits that use an H11AA1 for ZCD. It works fine. With a pullup on the output, it generates a pulse at each crossing. Because both diodes have a minimum forward voltage and current before they emit enough light to turn on the transistor, the pulse width depends on the supply voltage and resistors you are using to control the current. The pulse well be centered on the crossing, but of course, if you re triggering a uC interrupt or something, you'll get a positive edge slightly before the ZC and a negative edge slightly after. It doesn't seem to matter for dimming applications, but if you need to, you can use the uC timers to make an adjustment for the width.

I've also used this to build a line frequency monitor for fun. By timing/counting pulses, you can see the slight fluctuations in frequency that the power system generates. You need a good stable oscillator, as these fluctuations are usually small (10's of mHz) unless you live on an island or Something Major is happening to the grid.
 

Hi folks,

I've been looking at the application of H11AA1 optocouplers as zero crossing detectors and I can't see how they can possibly work. The data sheets are no help here at all.

...

Has anyone used this device for ZCD or is this application just an internet myth?

[mij59]

The LEDs are not the problem. The iffy area is the transistor switching times. Turn on is only 1.5us but turn off is 20+us.

My math is not good enough to compute the time for the 1.5v Vfd transition.

Hi,

Optocouplers don't do math, take a look at page 5 of the data sheet.
The turn on and turn off time is determent by the load resistor.