开云体育

I used to work fixing stage-n-lighting light dimmers. We used
oscilloscopes triggered by the incoming line AC. The output sine-wave
never shifted left and right as dimming was varied. Just the sharp
vertical line from "zero" where the triac triggered ... went "left and
right".

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

I think we could be getting into a semantic war with this thing. So ....
instead of my talking anymore ... please ... everyone open up the *.bmp
attachment of this email. THIS .... is what I saw on the scope output of
the triac dimmer. The arrows indicate what moves as I dim the lamps up and
down. The overall wave doesn't shift ... but that vertical line (indicated
by the arrows) ... shifts left and right.

As the area under the curve varies ... so does the intensity of the lamps.
I thought a picture might settle this.

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

Got it.

I'll try modifying future mail so you can see it.

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

Question: Why is the current waveform sinusodal despite the voltage output
being non-sinusodal?? How are you measuring it?? I would think it'd be
nothing-snap-full-comingdownsinusodal-zero-nothing-snap-full-comingbackupsin
usodal-zero. Just like the voltage.

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

The quadrant thing refers to the gate sensitivity differences between the
"forward SCR with a positive trigger", the "forward SCR with a negative
trigger", the "reverse SCR with a positive trigger" and lastly the "reverse
SCR with a negative trigger". Some modes of operation are more sensitive
than others.

The RC network is to prevent, say a fast line spike ... from triggering the
triac "on" inadvertently. If I remember correctly, the triac has a small
internal capacitance between the gate and main terminal. And a fast rising
spike can be conducted through it ... to the gate

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

I'm considering using triacs to switch 110VAC instead of relays.
Never used triacs before though and I want to double check my
understanding.

First thing is gate signal. This is AC, not DC correct? The AC
signal opens one SCR on the positive side of the sine wave and the
opposite SCR on the negative side?

A triac output optocoupler such as the MOC3022 would be used to turn
the triac on and off. Its triac side gets power from the AC line
directly. Is a simple triac opto good for a spindle motor or do I
need to get a zero-crossing opto like the MOC3040?

Last part - Do triacs generally fail open or closed? If they fail
closed, that could be a problem if used to switch motors on...

Merci!

On Tue, 13 Apr 2004 14:11:30 -0000, ghidera2000 <ghidera2000@...> wrote:

I'm considering using triacs to switch 110VAC instead of relays.
Never used triacs before though and I want to double check my
understanding.

First thing is gate signal. This is AC, not DC correct? The AC
signal opens one SCR on the positive side of the sine wave and the
opposite SCR on the negative side?

it only needs to be a pulse, it is sayd the optimum is a pulse chain for
best triggering but it can be a continous signal too.

I'm not entirely sure if the pulse can be both polarities or only the
one in which the current should flow, the datasheet can tell i hope.
(please answer if someone knows).
Tomi Engdahl has told me it can be both polarities, no matter which polarity
the main current has but i am not sure about this.

A triac output optocoupler such as the MOC3022 would be used to turn
the triac on and off. Its triac side gets power from the AC line
directly. Is a simple triac opto good for a spindle motor or do I
need to get a zero-crossing opto like the MOC3040?

this is a good and widely used solution.

the zero crossing has the advantage you need no (or not much) choke inductance.
if you have no zero crossing you need enough inductance to prevent triac destruction
(max dcurrent/st from datasheet).
on the other hand you have no possibility to make phase angle control with zero crossing detector.
you also need to keep HF signals low, because of EMC, especially if you switch fast this can be a problem.
In this case the inductance of the motor might very well be enough and you will have no such problems at all.

(looking in the average electronically controlled power drill the inductance seems to be enough).

Last part - Do triacs generally fail open or closed? If they fail
closed, that could be a problem if used to switch motors on...

i think they fail closed more often but i'm not too sure...


sam's repair faq on microwave triacs:


7.20) Testing and replacing the triac

A triac may fail in a variety of ways:

* A shorted triac would result in the oven coming on as soon as the door is

closed or the power being stuck on high no matter what the touchpad setting.

* An open triac or one that didn't respond to the gate would result in no heat

and possibly other things like the fan and turntable not working as well.

* A triac that didn't turn off would result in the parts of the oven continuing

to run even after the timer counted to zero.

* A triac where one half was shorted would result in a blown fuse due to it

acting as a rectifier pumping DC through the HV transformer.

* A triac where one half doesn't properly turn off would result in the main

fuse blowing when the cook cycle completed.

Nearly all triac failures will be shorts. Thus, measuring across the
MT1 and MT2 terminals of the triac (the power connections) should read
as a high resistance with a multimeter. A few ohms means a bad triac.

As noted above, triacs can fail in other - possibly peculiar ways - so
substitution or bypassing may be necessary to rule out all possibilities.

Replacement is very straightforward - just don't get the wires mixed up.

i knew there was some text about it in there..

ST

Merci!

On Tuesday 13 April 2004 10:58 am, Stefan Trethan wrote:

I'm not entirely sure if the pulse can be both polarities or only the
one in which the current should flow, the datasheet can tell i hope.
(please answer if someone knows).
Tomi Engdahl has told me it can be both polarities, no matter which
polarity the main current has but i am not sure about this.

This is correct. However a triac _will_ have different sensitivity in
different "quadrants" (the term that the datasheet will used to refer to the
different combinations of gate polarity and other polarity).

On Tue, 13 Apr 2004 11:05:53 -0400, Roy J. Tellason <rtellason@...> wrote:

On Tuesday 13 April 2004 10:58 am, Stefan Trethan wrote:

I'm not entirely sure if the pulse can be both polarities or only the
one in which the current should flow, the datasheet can tell i hope.
(please answer if someone knows).
Tomi Engdahl has told me it can be both polarities, no matter which
polarity the main current has but i am not sure about this.

This is correct. However a triac _will_ have different sensitivity in
different "quadrants" (the term that the datasheet will used to refer to the
different combinations of gate polarity and other polarity).

that means if you apply a dc current to the gate (higher than trigger current in all quadrants)
the triac will be on all the time, throughout the full sine wave?


in which magnitude should one expect the difference between the quadrants?



I did forget to add you will need a RC network across the triac to protect it from possible high voltages.
you can find circuit examples and i think even a guide how to calculate the component values
at epanorama.

ST

I dunno. If I recall correctly, a DC signal should work just fine. You
just won't be using one of the quadrants that "every other Joe" usually uses
(the negative MT/negative gate). You'd be using the negative MT/POSITIVE
gate quadrant.

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

--- In Electronics_101@..., Stefan Trethan
<stefan_trethan@g...> wrote:

Nearly all triac failures will be shorts.

Doh, oh well. Back to relays for this application. I'll check
epanorama though, good to know these things!

Thanks again.

On Tuesday 13 April 2004 11:30 am, Stefan Trethan wrote:

On Tue, 13 Apr 2004 11:05:53 -0400, Roy J. Tellason

<rtellason@...> wrote:

On Tuesday 13 April 2004 10:58 am, Stefan Trethan wrote:

I'm not entirely sure if the pulse can be both polarities or only the
one in which the current should flow, the datasheet can tell i hope.
(please answer if someone knows).
Tomi Engdahl has told me it can be both polarities, no matter which
polarity the main current has but i am not sure about this.

This is correct. However a triac _will_ have different sensitivity in
different "quadrants" (the term that the datasheet will used to refer to
the different combinations of gate polarity and other polarity).

that means if you apply a dc current to the gate (higher than trigger
current in all quadrants)
the triac will be on all the time, throughout the full sine wave?

I would think so, yeah.

in which magnitude should one expect the difference between the quadrants?

I guess the answer to that one would be on some data sheets. I don't remember
offhand, just that there were differences.

I did forget to add you will need a RC network across the triac to protect
it from possible high voltages.

That network is to prevent the device from triggering when the voltage across
it rises too rapidly, if I'm remembering right.

--- In Electronics_101@..., "Roy J. Tellason"
<rtellason@b...> wrote:

On Tuesday 13 April 2004 11:30 am, Stefan Trethan wrote:

On Tue, 13 Apr 2004 11:05:53 -0400, Roy J. Tellason

<rtellason@b...> wrote:

On Tuesday 13 April 2004 10:58 am, Stefan Trethan wrote:

I'm not entirely sure if the pulse can be both polarities or

only the

one in which the current should flow, the datasheet can tell i

hope.

(please answer if someone knows).
Tomi Engdahl has told me it can be both polarities, no matter which
polarity the main current has but i am not sure about this.

This is correct. However a triac _will_ have different

sensitivity in

different "quadrants" (the term that the datasheet will used to

refer to

the different combinations of gate polarity and other polarity).

that means if you apply a dc current to the gate (higher than trigger
current in all quadrants)
the triac will be on all the time, throughout the full sine wave?

I would think so, yeah.

in which magnitude should one expect the difference between the

quadrants?

I guess the answer to that one would be on some data sheets. I

don't remember

offhand, just that there were differences.

I did forget to add you will need a RC network across the triac to

protect

it from possible high voltages.

That network is to prevent the device from triggering when the

voltage across

it rises too rapidly, if I'm remembering right.

Yes, a fast dv dt across the anode and cathode terminals will cause a
SRC or TRIAC to trigger without any gate signal applied. The voltage
does not need to be particularly high so the RC network is not there
to protect against high voltages, it is there to slow things down a bit.

No, A DC gate signal will not 'turn on the TRIAC' it will only rectify
the current to the load if the power source is from the AC mains. A
TRIAC is the same as two reverse parallel SCRs. If a DC gate signal is
applied, only one of the 'SCR halves' will ever turn on.

A DC signal should work fine for a photo-coupled TRIAC.I have driven
small motors and lights directly from an optocoupled triac using a DC
signal; it worked fine.

Yes, TRIACs almost always fail closed. Most light dimmers that go bad
stop dimming but the light will still go full on as long as the switch
works. Don't forget that relays can also weld their contacts together
and fail short. Look for a more positive way to shut off the circuit
if there is a failure like remove all power, short the motor and blow
the fuse, etc.

Good luck!

On Tuesday 13 April 2004 08:45 pm, manifold wrote:

--- In Electronics_101@..., "Roy J. Tellason"

<...>

I did forget to add you will need a RC network across the triac to
protect it from possible high voltages.

That network is to prevent the device from triggering when the
voltage across it rises too rapidly, if I'm remembering right.

Yes, a fast dv dt across the anode and cathode terminals will cause a
SRC or TRIAC to trigger without any gate signal applied.

That's the expression I was reaching for and couldn't quite recall, dv/dt!

The voltage does not need to be particularly high so the RC network is not
there to protect against high voltages, it is there to slow things down a
bit.

This meshes with my understanding as well.

No, A DC gate signal will not 'turn on the TRIAC' it will only rectify
the current to the load if the power source is from the AC mains. A
TRIAC is the same as two reverse parallel SCRs. If a DC gate signal is
applied, only one of the 'SCR halves' will ever turn on.

Oh?

A DC signal should work fine for a photo-coupled TRIAC.I have driven
small motors and lights directly from an optocoupled triac using a DC
signal; it worked fine.

Yes, TRIACs almost always fail closed. Most light dimmers that go bad
stop dimming but the light will still go full on as long as the switch
works. Don't forget that relays can also weld their contacts together
and fail short. Look for a more positive way to shut off the circuit
if there is a failure like remove all power, short the motor and blow
the fuse, etc.

Yes, I was thinking that fuses were good to have in there. As well as way
over-spec'ing the triacs, as I sure haven't seen many fail.

You're right, A DC voltage applied to the gate will allow the TRIAC to
conduct in both directions as long as the applied voltage gets over
the threshold voltage. I guess I have just never used one like that
and did not think through the analogous SCR circuit well enough (duh).

That would mean that the output would always have a notch out of the
of the applied voltage waveform and that notch height would be equal
to the threshold voltage. The notch location would be determined by
the value of the DC gate voltage up the limit of 10V.

Is that right?

--- In Electronics_101@..., "Curtis Sakima" <csakima@m...>
wrote:

I dunno. If I recall correctly, a DC signal should work just fine.

You

just won't be using one of the quadrants that "every other Joe"

usually uses

(the negative MT/negative gate). You'd be using the negative

MT/POSITIVE

gate quadrant.

Curtis

Dazzle Mom this coming Mother's Day season with flowers!




----- Original Message -----
From: manifold <manifold_1@y...>

No, A DC gate signal will not 'turn on the TRIAC' it will only

rectify the

current to the load if the power source is from the AC mains. A

TRIAC is the

same as two reverse parallel SCRs. If a DC gate signal is applied,

only one

of the 'SCR halves' will ever turn on.

On Wednesday 14 April 2004 06:59 pm, manifold wrote:

That would mean that the output would always have a notch out of the
of the applied voltage waveform and that notch height would be equal
to the threshold voltage. The notch location would be determined by
the value of the DC gate voltage up the limit of 10V.

Seems to me that with any kind of a triac circuit you're *always* going to
have a bit of a notch, in that the applied voltage has to rise to a certain
point before triggering can take place...

This is why there are times when I'll get interference from a lamp dimmer
here, no matter where it's set -- even "full on".

I think THAT part is because even at "full brightness" ... the phase shift
mechanism of most dimmers *still* does not have enough shift to shift the
trigger point all the way to "zero" (IE: the lamp is a 95% ... but the pot
is banged all the way to "full". Can't increase the pot any more.).

Curtis

Dazzle Mom this coming Mother's Day season with flowers!

Not sure if a triac "creates" a phase offset between voltage and current.
AFAIK, a triac only PWM's the input voltage/current ... and that's what
"controls" the brightness of the lamp (heat from the heating element).

I thought only capacitive/inductive elements creates angle shifts .......??


Curtis

Dazzle Mom this coming Mother's Day season with flowers!

On Wed, 14 Apr 2004 22:59:36 -0000, manifold <manifold_1@...> wrote:

You're right, A DC voltage applied to the gate will allow the TRIAC to
conduct in both directions as long as the applied voltage gets over
the threshold voltage. I guess I have just never used one like that
and did not think through the analogous SCR circuit well enough (duh).

That would mean that the output would always have a notch out of the
of the applied voltage waveform and that notch height would be equal
to the threshold voltage. The notch location would be determined by
the value of the DC gate voltage up the limit of 10V.

Is that right?

dunno...

will have to actually try it when i need it.

i thought a scr is current triggered, not voltage triggered, is that right?

If you think it through with two seperate SCRs it looks like it will not work
but a triac is not exactly only 2 scrs, like a bipolar transistor is not only two diodes.

It might just be that with the opposite polarity of the trigger pulse the other PN junction
might be used, which would mean with DC trigger always the same junction is used to trigger,
no matter which direction the main current is...


not sure myself...

ST

On Wed, 14 Apr 2004 19:18:18 -0400, Roy J. Tellason <rtellason@...> wrote:

On Wednesday 14 April 2004 06:59 pm, manifold wrote:

That would mean that the output would always have a notch out of the
of the applied voltage waveform and that notch height would be equal
to the threshold voltage. The notch location would be determined by
the value of the DC gate voltage up the limit of 10V.

Seems to me that with any kind of a triac circuit you're *always* going to
have a bit of a notch, in that the applied voltage has to rise to a certain
point before triggering can take place...

This is why there are times when I'll get interference from a lamp dimmer
here, no matter where it's set -- even "full on".

yes, true, the effect is not that much compared to 240Veff line voltage...
also you need sufficient inductance and maybe filtering for EMC.

One thing to keep in mind is that a system using phase angle control (like a triac)
can produce a phase offset between voltage and current and you draw reactive power.
the interesting thing is you have nothing to store this energy in so it seems to be converted
to heat/light which is efective power. this seems a bit odd to me because it should not be possible
tu "use" reactive power to gain energy. another thing is that the kWh meter which measures how much
you will pay does not seem to measure reactive power.

This is one point i need to think about some day... It's in the back of the mind but i never
really thoght it through if it is correct.

ST