Questions about triacs
GiveMeL
voltage is needed at the gate to trigger a triac ? I'm looking at
the datasheet for bt137...
IGT Gate trigger current. Conditions : VD = 12 V; IT = 0.1 A
T2+ G+ typ. 5mA
T2+ G- typ. 8mA
T2- G- typ. 11mA
T2- G+ typ. 30mA
Say I want to control mains voltage and trigger the triac at different
points of the AC cycle. If, say, T2 is at ~50V, does the gate need to
be above that voltage so that at least 5mA of current flow into the
gate ?
I've seen schematics showing a microcontroller's output connected
directly to a 'logic level' triac. Do logic level triacs have some
sort of isolation between the main terminals and the gate ?
I'm wondering if mains current can flow from the main terminals
through the gate ? If it does, how do I limit it so that the triac
doesn't blow up ?
Your answers are much appreciated ! =]
P.
Phil Allison
"GiveMeL"
>
> I don't understand the details of triac triggering.
** Baffles lots of newcomers, for some strange reason.
> What kind of
> voltage is needed at the gate to trigger a triac ?
** About 1.5 volts or so.
> I'm looking at the datasheet for bt137...
>
> IGT Gate trigger current. Conditions : VD = 12 V; IT = 0.1 A
>
> T2+ G+ typ. 5mA
> T2+ G- typ. 8mA
> T2- G- typ. 11mA
> T2- G+ typ. 30mA
>
> Say I want to control mains voltage and trigger the triac at different
> points of the AC cycle. If, say, T2 is at ~50V, does the gate need to
> be above that voltage so that at least 5mA of current flow into the
> gate ?
** All triac triggering is done by applying a current between the gate and
MT1.
As you can see from the table above, the amount of current needed depends
somewhat on what voltage is on MT2 and the polarity of the trigger voltage.
When MT2 is positive relative to MT1 and the gate voltage is also positive,
triggering is best.
> I've seen schematics showing a microcontroller's output connected
> directly to a 'logic level' triac. Do logic level triacs have some
> sort of isolation between the main terminals and the gate ?
** No.
The common rail of such a circuit is connected to MT1.
> I'm wondering if mains current can flow from the main terminals
> through the gate ?
** Yep.
> If it does, how do I limit it so that the triac doesn't blow up ?
** Meaningless question without a schem of what you are considering.
Keep well away from triacs until you understand what you are doing.
.... Phil
Charles
GiveMeL
understand
what's going on =]
>Baffles lots of newcomers, for some strange reason.
Heh. Well, for DC circuits I can be certain that ground == 0 volts
no ?
But that doesn't seem to be true for AC circuits ?
(Where I live electrical outlets usually have only 2 holes. Which
means, I guess,
that people here don't much care about which one is hot and which one
is neutral
since you can connect any appliance one way or the other by simply
turning the plug around...)
Anyway, I got this schematics from http://www.epanorama.net/circuits/semiconductor_relays.html#triacuse
R1
100 N1 Load
+----/\/\/---------------+----------o o-------> 120V
| |
Live
| | MT2
o +-+
SWITCH \ G | | TRIAC
o / | |
| / +-+
| | | MT1
+-----------------+ |
|
+-----------------------> 120V
Neutral
When the switch is open, N1 follows the mains voltage ?
Now, say the switch is closed when Vmains = 60V (first quarter of the
cycle) ? What's the amount of current flowing through R1 into G ?
I guess it's determined by the voltage G is at ? But what voltage is
that ?
GiveMeL
Phil Allison
"GiveMeL"
>
> Phil, thanks for replying. Yeah, I won't try anything before I really
> understand what's going on =]
** That may be a while yet ......
>>Baffles lots of newcomers, for some strange reason.
>
> Heh. Well, for DC circuits I can be certain that ground == 0 volts
> no ?
> But that doesn't seem to be true for AC circuits ?
** With AC mains power, you must always assume that either or both current
carrying conductors are live.
> Anyway, I got this schematics from
> http://www.epanorama.net/circuits/semiconductor_relays.html#triacuse
>
> R1
> 100 N1 Load
> +----/\/\/---------------+----------o o-------> 120V
> | |
> Live
> | | MT2
> o +-+
> SWITCH \ G | | TRIAC
> o / | |
> | / +-+
> | | | MT1
> +-----------------+ |
> |
> +-----------------------> 120V
>
> Neutral
>
>
>
> When the switch is open, N1 follows the mains voltage ?
** Correct.
> Now, say the switch is closed when Vmains = 60V (first quarter of the
> cycle) ? What's the amount of current flowing through R1 into G ?
** About 600 mA - for a couple of microseconds. Then the triac is
triggered "on" and there is a low voltage ( say 1 volt) between MT2 and MT1.
The triac will briefly go off when the supply voltage drops to under 1 volt
and be re-triggered almost immediately when it rises again in the other
polarity - long as the switch stays closed.
> I guess it's determined by the voltage G is at ? But what voltage is
> that ?
** G is at about 1.5 to 2 volts above MT1 ( assuming MT2 is positive
compared to MT2 ) at the moment of triggering ( ie for about 2 uS) then
drops to the same voltage as MT2.
.... Phil
GiveMeL
R1
100 N1 Load
+----/\/\/---------------+----------o o-------> 120V
| |
| |
Live
| | MT2
o +-+
SWITCH \ G | | TRIAC
o / | |
| / +-+
| | | MT1
+-----------------+ |
|
+----------------------->
120V neutral
GiveMeL
So, in that schem the current for the gate comes directly from the
mains and is limited by R1. If the triac was a bt137 which has a peak
current of 2A, and Vmains was 220v, R1=150 (say) would limit gate
current to a safe value.
And it is also possible to use a separate source of current for the
gate and that's what the circuits that interface a uController to a
triac do, I guess ? In that case the current comes from the
uController's PS - So 5V DC can be used to trigger the triac, just
like 110/220 AC.
Phil Allison
"GiveMeL"
** There are three main categories of triggering circuit used with triacs:
1. Pulse triggering ( aka pulse firing )
2. Continuous triggering (aka hard firing )
3. DC triggering
The schem with the switch and the 100 ohm resistor is, despite appearances,
a form of pulse triggering - cos a brief current pulse flows at the moment
of switching and at the beginning of each AC cycle.
Continuous triggering uses a stream of 2 to 5 uS long pulses, spaced about
50uS apart, to trigger the gate - usually via a pulse transformer.
DC triggering simply applies a continuous current to the gate, sufficient to
fire the particular triac in all quadrants.
BTW:
Once you have figured out how to make a triac conduct - you can start
worrying about how to make sure it goes off when you want it to.
Cos that is whole nuther can of worms.
.... Phil
GiveMeL
>worrying about how to make sure it goes off when you want it to.
Hehe. Well, I'll start with a resistive load and see what happens...
Thanks a lot for your help Phil!
P.
Robert Baer
Trigger voltage is a Vbe (near 0.6V), but beware using voltage as you
can zap it with too much current.
To trigger at different parts of the AC waveform, one controls the
timing of the trigger.
There is no such thing as a "logic" triac.
Phil Allison
"Robert Baer"
>
> Trigger voltage is a Vbe (near 0.6V), but beware using voltage as you can
> zap it with too much current.
> To trigger at different parts of the AC waveform, one controls the timing
> of the trigger.
> There is no such thing as a "logic" triac.
** Try reading a triac data sheet sometime.
http://www.st.com/stonline/products/literature/ds/7471.pdf
... Phil