Alternating current 0-crossing

[danielsmusic]

I want to make a very crude (cheap 'n' nasty) 0-crossing detecter for a small AC current (5V - -5V max). It should send a pulse everytime it crosses 0v. I've not tested it yet and want to know your thoughts.

 

[Nigel Goodwin]

Well your bridge is wired completely wrongly, but even if wired correctly it's going to switch well before zero crossing and switch back well after, due to the switching points of the gate.

Better to use a single NPN transistor, with a resistor feeding the base from the bridge, and a pullup on the collector - this will give a narrower pulse than the gates.

 

[danielsmusic]

Something like this? How could I determine the value of the resistor?

 

[ericgibbs]

Something like this? How could I determine the value of the resistor?

Why do you think this version is any different from the first one.?

 

[Nigel Goodwin]

The bridge is still wrong, and it needs to be an NPN transistor, with emitter to chassis, and a pullup resistor on the collector.

 

[flat5]

"The bridge is still wrong"

Your input needs to work for both voltage directions.
Both input legs need to see the bridge both ways.
-l<- and ->l-
clear?

 

[Roff]

This is pretty cheap, but not too nasty. You'll get narrower pulses than with the diode bridge circuit, which is a good thing.

 

[flat5]

Please Roff, can you walk me through that circuit or provide a link?

 

[Roff]

Please Roff, can you walk me through that circuit or provide a link?

Do you know how to use LTspice?

 

[Roff]

On positive half cycles, the active circuitry consists of a saturating common-emitter stage consisting of R3, Q3, and R5. The other transistors are off. D1 keeps the Vbe of Q1 from breaking down in the case where Vcc is higher than the breakdown voltage (generally around 6V).
On negative half cycles, Q3 is off, and Q2 is a saturating common base stage. It has no current gain, and its collector voltage will be about -0.6V wwhen it is saturated. Q1 is an emitter follower, providing the current gain necessary to drive R5 and also the level shifting reqiured to keep the output slightly above 0V while Q2 is saturated.
Therefore, during both positive and negative half cycles of the input, the output is near 0V. When -0.6V<Vin<+0.6V, All transistors are off, and the output rises to Vcc, resulting in a short positive pulse during each zero crossing.
EDIT: I am not the **broken link removed**, although I had to do a little Googling to find it. I had reconstructed it from memory, and wound up with a slightly different circuit. The original is simpler, in that it doesn't require the diode. It also has more symmetrical rise and fall times, at least in simulation.

 

[MrAl]

Hi there,


Here are a couple of simpler zero crossing detectors.

One is very basic for use when the exact zero cross point is not that
important.

The second circuit can be easily adjusted to make the rise of the output
pulse exactly equal to the zero crossing points of the input ac wave.

Both circuits only require a single 2N2222 (or similar) transistor.


**broken link removed**

I should add a note however, that a comparator based circuit is not that complex yet
works even better than a transistor based circuit.

 

[eblc1388]

The original is simpler, in that it doesn't require the diode. It also has more symmetrical rise and fall times, at least in simulation.

For anyone interested in the circuit, here is the asc file.

 

[flat5]

Thank you Roff for your time and expertise.
I will print the circuit and read your text until I can understand it...or most of it as the case may be.

LTSpice I have downloaded but have not learned to use it yet.
I find it to be a difficult program. I suppose I need more background concerning spice, in general.

 

[SYE]

Hi,
I've always used this circuit whenever I need to detect zero crossings.

 

[flat5]

"It's All Clear Now" (Humphrey Bogart)

Thanks again, Roff.

 

[MrAl]

Hi,
I've always used this circuit whenever I need to detect zero crossings.

That's an interesting little circuit too there, although it may trigger
a little high on the sine wave it's still usable im sure.

In the past i have always used a comparator that can detect a
low level change, and a pretty high input sine wave so that the
zero detect is only off by 1 or 2 degrees. A little hysteresis and
it's good to go. I've used this in 100 amp line switches too.

The two circuits i posted where mainly to keep parts count low, and
i thought the OP was wanting to use only transistors and no IC's,
but i could be wrong. I'll wait however to see if he ever even
notices them he he

 

[danielsmusic]

Sorry I'm replying a bit late, I have some 3904's around but I seem to have misplaced them in a big pile of random parts hehe. I've not had a good chance to experiment yet and I will probably try all the circuits when I get a good chance. Thanks for posting them