Driving Triac Directly with 555 Output?

[Jamie]

"Jamie"




** You are a fucking LIAR !!




** What you said was CRAP.






** Triacs will work fine with + DC drive on the gate.

Never come across one that did not.

**** you - asshole.


..... Phil

Ha, you make me laugh Phil.
I've done so much work with SCR's and Triacs in various
applications, I don't even know where to start..

To coin one of your phrase;

"Piss off! you autistic piss of shit"

Jamie

 

[Phil Allison]

"John Fields"

Notice the reference to a four quadrant triac?

** Something almost ANY triac is.

A search of the Australian version of the Element 14 ( aka Farnell ) printed
catalogue indicates that the vast majority of triacs listed are speced in
all 4 quadrants - and the rest are still perfectly usable in all 4.

http://www.littelfuse.com/data/en/Product_Catalogs/PowerThyristorDataSheets.pdf

** Posting a link to a 466 page PDF file is a tad un-cool, with no warning.

BTW:

Larkin is beating a smelly dead donkey with a bent stick for his own
perverted amusement.

As per fucking usual.



..... Phil

 

[Spehro Pefhany]

On Fri, 11 Feb 2011 19:54:52 -0600, John Fields

Obviously beating a dead horse since the OP admitted his quest was
bogus but, just for for fun, here's how to do it with a four quadrant
triac, no opto, and no full-wave bridge:

.AC>--+-------------------------------------+
. |A |
. [1N4002] |
. | |
. | 78L05 |
. +-I O-+-------------+---------+ [LOAD]
. |+ G | | | |
. [C1] | [R1] |8 | |
. | | | +----+----+ | |
. +---+ | | Vcc _| | |
. | +-----7-O|D R|O-4-+ |
. | | | | MT2
. | [R2] +--6-|TH OUT|-3-[R3]---G
. | | | |__ | MT1
. | +---+-2-O|TR 555| |
. | | | GND | |
. | [C2] +----+----+ |
. | | |1 |
.AC>--+-------+-------------+---------------+

Most triacs are not spec'd to fire in quadrant 4.

John

---
You said that earlier, and it's true enough, but irrelevant.

Did you not read the text preceding the schematic?

Here, I'll post it again:

"Obviously beating a dead horse since the OP admitted his quest was
bogus but, just for for fun, here's how to do it with a four quadrant
triac, no opto, and no full-wave bridge:"

Notice the reference to a four quadrant triac?

Something like one of Teccor's "L" or "O" series triacs at:

http://www.littelfuse.com/data/en/Product_Catalogs/PowerThyristorDataSheets.pdf

would work nicely.

Using the "L" series would allow the use of a 7555 and greatly reduce
the size of C1.

Careful with too close connection of chips and a triac gate.. the gate
pin isn't completely benign.


Best regards,
Spehro Pefhany

 

[Jamie]

John Fields wrote:



Yes, that circuit is bogus, I am sure its just a mistake...

---
Can you fix it?
---

One point that I must make here is this, it was asked about
driving the Gate directly from the output of the timer. And
unless I am over looking something here, it looks like a
Q3 operation.. When the timer is on, you'll get gate current
in both directions how ever, when off the 555 goes into a
sink operation, I see a problem here.


Jamie

---
Using this circuit: (view with a fixed-pitch font:

.AC>--+-------------------------------------+
. |A |
. [1N4002] |
. | |
. | 78L05 |
. +-I O-+-------------+---------+ [LOAD]
. |+ G | | | |
. [C1] | [R1] |8 | |
. | | | +----+----+ | |
. +---+ | | Vcc _| | |
. | +-----7-O|D R|O-4-+ |
. | | | | MT2
. | [R2] +--6-|TH OUT|-3-[R3]---G
. | | | |__ | MT1
. | +---+-2-O|TR 555| |
. | | | GND | |
. | [C2] +----+----+ |
. | | |1 |
.AC>--+-------+-------------+---------------+

will result in the triac being driven in quadrants 1 and 4, as shown
on page 17 of:

http://www.littelfuse.com/data/en/Product_Catalogs/PowerThyristorDataSheets.pdf


Since MT1 is referenced to what the circuit sees as ground, when the
555's output is low it'll merely follow "ground" and there'll be
nothing to drive the gate with.

When its output goes high, however, it'll fire the triac in QI if MT2
is positive with respect to MT1, and in QIV if MT2 is negative with
respect to MT1.

Not exactly sure what you are trying to point out here but a Q4 will
fire either way as long as there is current in the gate through A1(MT1).


My comment was based on the fact of induction kicking back through
the sink mode of the timer.. Seen this kind of circuit and it causes
miss fires a long with damage if the voltages are too high. Hence the
reasoning for using Opto's. Using an Emitter follower is still better
than using a direct link like this, at least it removes some of the
return path to the gate, when off.

Years ago (~25), I redesigned a Siltronics EDM drive and sensor
controls. The supply had a CT source so I simply used a bridge rectifier
on that source and 2 Q4 Triacs being driven via 2 555's using a phase
angle trigger from the supply to control speed. I made several trips
there replacing the 555's over a period of 6 months before I understood
what was going on. I ended up using the 4N24 family type to fix the
problem by isolating the 555 from the triac gate. I was able to
duplicate the problem by jerking the attached DC motor up/down rapidly
there by, causing the DC motor to back generate some voltage.. This
motor had snubbers and large chokes on it but still didn't fix this
problem..

When It was finished, the new drive, PID and sensor electronics
occupied only 1/4 of the real-estate over the original drive. As far as
know, that machine still runs today! Back in those days, we didn't have
a lot of Triacs you could depend on not latching up. Today, they are
getting much better.. But I still have doubts about using them in a
quad type output bridge.

Then there was the Cincinnati EDM, that is another story, 3-400Z's

Jamie

 

[Jamie]

Please pay attention and don't start clucking half-cocked.

<[email protected]>

My point was to be careful about quadrants. Some triacs trigger in all
four quads, but a lot won't.

John

I know how that is, we have a load of 50 amp Q4 studded types that has
issues where only 1 out of 4 will fire properly in all quadrants at
their prescribed 100 ma trigger. The failed one's have at least 1Q that
requires like 250ma or more to fire. Must of been some fire sale I got
them from!

Jamie

 

[Jamie]

Obviously beating a dead horse since the OP admitted his quest was
bogus but, just for for fun, here's how to do it with a four quadrant
triac, no opto, and no full-wave bridge:

.AC>--+-------------------------------------+
. |A |
. [1N4002] |
. | |
. | 78L05 |
. +-I O-+-------------+---------+ [LOAD]
. |+ G | | | |
. [C1] | [R1] |8 | |
. | | | +----+----+ | |
. +---+ | | Vcc _| | |
. | +-----7-O|D R|O-4-+ |
. | | | | MT2
. | [R2] +--6-|TH OUT|-3-[R3]---G
. | | | |__ | MT1
. | +---+-2-O|TR 555| |
. | | | GND | |
. | [C2] +----+----+ |
. | | |1 |
.AC>--+-------+-------------+---------------+

---
JF

Most triacs are not spec'd to fire in quadrant 4.

John

---
You said that earlier, and it's true enough, but irrelevant.

Did you not read the text preceding the schematic?

Here, I'll post it again:

"Obviously beating a dead horse since the OP admitted his quest was
bogus but, just for for fun, here's how to do it with a four quadrant
triac, no opto, and no full-wave bridge:"

Notice the reference to a four quadrant triac?

Something like one of Teccor's "L" or "O" series triacs at:

http://www.littelfuse.com/data/en/Product_Catalogs/PowerThyristorDataSheets.pdf

would work nicely.

Using the "L" series would allow the use of a 7555 and greatly reduce
the size of C1.

Careful with too close connection of chips and a triac gate.. the gate
pin isn't completely benign.


Best regards,
Spehro Pefhany

They don't seem to be too worried about that. It's evident.

Jamie

 

[Jamie]

John Fields wrote:

On Fri, 11 Feb 2011 21:14:44 -0500, Jamie



John Fields wrote:



Several circuits I've seen use an opto-isolator or a transistor between
the 555 output and the triac gate, i.e.
<http://www.circuitstoday.com/wp-content/uploads/2008/03/lamp-flasher12.jpg>
or
<http://www.electro-tech-online.com/images/projects/triac_light_flasher.gif>.

The triac I'm using has a 50mA gate current and the 555 can sink and
source 200mA. The triac will be controlling a non-inductive AC load
where the AC voltage will not exceed 18V (coming from a bicycle dynamo),
so there is no shock hazard here.

Can anyone think of a reason I need that extra component between pin 3
of the 555 and the triac gate? I was just going to put in a current
limiting resistor of 10 ohms or so.


---
The circuit at:

http://www.circuitstoday.com/wp-content/uploads/2008/03/lamp-flasher12.jpg>
can't work because there's no supply for the 555, unless I'm missing
something.

The other circuit needs the optoisolator because there's no common
ground between the dynamo, the 555, and the TRIAC's MT1.

It also has some extra stuff that's not needed for your application,
and I'll post a schematic for you some time today.

---
JF

Yes, that circuit is bogus, I am sure its just a mistake...


---
Can you fix it?
---



One point that I must make here is this, it was asked about
driving the Gate directly from the output of the timer. And
unless I am over looking something here, it looks like a
Q3 operation.. When the timer is on, you'll get gate current
in both directions how ever, when off the 555 goes into a
sink operation, I see a problem here.


Jamie


---
Using this circuit: (view with a fixed-pitch font:

.AC>--+-------------------------------------+
. |A |
. [1N4002] |
. | |
. | 78L05 |
. +-I O-+-------------+---------+ [LOAD]
. |+ G | | | |
. [C1] | [R1] |8 | |
. | | | +----+----+ | |
. +---+ | | Vcc _| | |
. | +-----7-O|D R|O-4-+ |
. | | | | MT2
. | [R2] +--6-|TH OUT|-3-[R3]---G
. | | | |__ | MT1
. | +---+-2-O|TR 555| |
. | | | GND | |
. | [C2] +----+----+ |
. | | |1 |
.AC>--+-------+-------------+---------------+

will result in the triac being driven in quadrants 1 and 4, as shown
on page 17 of:

http://www.littelfuse.com/data/en/Product_Catalogs/PowerThyristorDataSheets.pdf


Since MT1 is referenced to what the circuit sees as ground, when the
555's output is low it'll merely follow "ground" and there'll be
nothing to drive the gate with.

When its output goes high, however, it'll fire the triac in QI if MT2
is positive with respect to MT1, and in QIV if MT2 is negative with
respect to MT1.

Not exactly sure what you are trying to point out here but a Q4 will
fire either way as long as there is current in the gate through A1(MT1).


My comment was based on the fact of induction kicking back through
the sink mode of the timer.. Seen this kind of circuit and it causes
miss fires a long with damage if the voltages are too high. Hence the
reasoning for using Opto's. Using an Emitter follower is still better
than using a direct link like this, at least it removes some of the
return path to the gate, when off.

Years ago (~25), I redesigned a Siltronics EDM drive and sensor
controls. The supply had a CT source so I simply used a bridge rectifier
on that source and 2 Q4 Triacs being driven via 2 555's using a phase
angle trigger from the supply to control speed. I made several trips
there replacing the 555's over a period of 6 months before I understood
what was going on. I ended up using the 4N24 family type to fix the
problem by isolating the 555 from the triac gate. I was able to
duplicate the problem by jerking the attached DC motor up/down rapidly
there by, causing the DC motor to back generate some voltage.. This
motor had snubbers and large chokes on it but still didn't fix this
problem..

When It was finished, the new drive, PID and sensor electronics
occupied only 1/4 of the real-estate over the original drive. As far as
know, that machine still runs today! Back in those days, we didn't have
a lot of Triacs you could depend on not latching up. Today, they are
getting much better.. But I still have doubts about using them in a
quad type output bridge.

Then there was the Cincinnati EDM, that is another story, 3-400Z's

Jamie

The MOC3010/3020 series of opto triac drivers are cool. They solve a
bunch of problems.

John

YEs, I've used those types a lot in pure AC circuits, they do work
nicely. The old 808 Eurotherm PID controllers use those for the
AC output option card to drive the Triac.. Our plant still uses lots of
those units so we keep them on hand. What's nice about those
controllers are, we can make our own little output interface and the
processor does not know the difference


Jamie

 

[Jasen Betts]

Most triacs are not spec'd to fire in quadrant 4.

just flip everything, drive the gate negative (needs one more part)

AC>--+---+---------------------+---------+-----+
|+ | | | |
[C1] | 1N4000 |8 | |
| | +-->|--+ +----+----+ | |
| | | | | Vcc _| | |
| | +-[R2]-+-3|OUT R|4---+ |
| | | | | MT1
| | [R2] +----2|TH DIS|7--[R3]---G
| | | | |__ | MT2
| | +---+----6|TR 555| |
| | | | GND | |
| | [C1] +----+----+ |
| | | |1 |
+-[7905]---+--------------+ load
| i o |
V |
T 1N4000 |
AC>--+-----------------------------------------+

 

[SMS]

---
The circuit at:

http://www.circuitstoday.com/wp-content/uploads/2008/03/lamp-flasher12.jpg>
can't work because there's no supply for the 555, unless I'm missing
something.

The other circuit needs the optoisolator because there's no common
ground between the dynamo, the 555, and the TRIAC's MT1.

It also has some extra stuff that's not needed for your application,
and I'll post a schematic for you some time today.

Here is how I'm doing it now:

<http://i53.tinypic.com/iee5ck.jpg>

The MR16 LED lamp runs off AC or DC (since it is used for a replacement
lamp in track lighting which uses just a transformer, as well as in DC
applications, the internal circuit (which I don't have access to) to the
lamp allows AC or DC operation).

So now, when the light needs to be steady on, I run it directly off the
dynamo, and when I want the strobe I drive it with DC, through Q1.

I'd rather run it off AC all the time and somehow turn it off and on
with a triac (or other device). I would have less losses. Normally the
VCE and diode drop losses are not worth worrying about, but with a
dynamo that produced 6-9V at 500mA, the losses are significant enough to
worry about.

 

[SMS]

---
The circuit at:

http://www.circuitstoday.com/wp-content/uploads/2008/03/lamp-flasher12.jpg>
can't work because there's no supply for the 555, unless I'm missing
something.

Yeah, there's a mistake there. Should be
<http://www.circuitstoday.com/wp-content/uploads/2008/03/Lamp-Flasher-Circuit-Using-NE-555.bmp>

 

[SMS]

On Wed, 16 Feb 2011 10:05:03 -0600, John Fields


---
Oops...

R1 and R2 were 680k and 300k respectively, changed to 100k and 470k
below:

.AC>--+--------------------------------------------+
. |A |
. [1N4002] |
. |D1 |
. | LED1 |+
. +-------+-------------+---------+ [MR16]
. |+ |R1 | | |
. [470µ] [100k] |8 | |
. |C1 | +----+----+ | |
. | | | Vcc _| | |
. | +-----7-O|D R|O-4-+ SSR |
. | |R2 | | R3 +----+ |
. | [470k] +--6-|TH OUT|-3--[1k]--|+ L|-+
. | S1 | | |__ | | |
. | |<-o-+---+-2-O|TR 7555| +-|- L|-+
. | | |C2 | GND | | +----+ |
. | O [0.47µ] +----+----+ | K1 |
. | | | U1 |1 | |
.AC>--+-+-----+-------------+-------------+--------+

With S1 in the open position the MR16 will flash at about 3Hz with
a 557% duty cycle, while with the switch closed
it'll be on continuously.

Also, I suggest you use a CMOS 555 and a solid-state relay, as shown
above.

The 7555 will eliminate the crossover transient the 555 suffers from
and the SSR will eliminate any load nasties getting into the 555
through the triac's gate. (Thanks for the heads-up, Speff

A nice one would be a Clare CPC1018N since it only needs 1mA to drive
the output ON.

Mouser has them in stock at:

Thanks, that's very close to what I want to do. I think I'd use a 3PDT
(center off) switch to a) disconnect the 555 from the circuit, b) short
the relay outputs for non-strobe mode (as there is no need to have the
555 and the relay operating and using power during non-strobe mode),and
c) need an on/off switch anyway.

<http://i51.tinypic.com/idr42o.jpg>

I had looked into SSRs earlier, but obviously not enough, as the ones I
find were essentially triacs with a snubber circuit.

I presume you meant 55.7%

The other issue is that the SSR increases the cost significantly versus
just using a PNP transistor and operating the MR16 on DC during strobe
mode. For one or two it's not an issue, but for mass production it's an
issue.

I always stick a cap on pin 5, per recommendations for the 555. Do you
find that not necessary?

 

[SMS]

Do you have any spec's for the lamp?

Probably inside the MR16 module the circuit is similar to Figure 15

However, if you use a bipolar 555 it's _mandatory_ to put a 100nF cap
across pins 1 and 8, as close as possible to the package; preferably
right across the pins.

If you use a CMOS 555 it isn't necessary.

Right, I'd use a CMOS part since I don't need high output current.

 

[SMS]

Have you considered a schottky bridge to kep the DC operating losses
down?

Yes, I'm building a bridge out of Schottky diodes for that reason. I was
trying to do through-hole, not surface mount, and I could not find any
through-hole Schottky bridge rectifiers.

 

[SMS]

Mouser has them in stock at:

http://www.mouser.com/Search/Refine.aspx?Keyword=CPC1018N

and a data sheet is available at that link.

I think I need a higher power relay. The power dissipation could be up
to 700mW or so (600mA *11V). Thinking of the Clare LCA710.

<https://www.jameco.com/Jameco/Products/ProdDS/1583041.pdf1>

I have all the parts on order from AllElectronics other than the relay.
Jameco I can do will-call from, so I'd prefer them over shipping.

 

[SMS]

---
That's how much your load will be dissipating.

If your load draws 600mA and the maximum ON resistance of the relay is
0.5 ohm, then it'll dissipate:


P = I²R = 0.6A² * 0.5R = 0.18 watts,


so the CPC1018N should work OK.

Having said that, THE LCA710 has a lower ON resistance than the
CPC1018N, and you've got the 10mA drive available from a CMOS 555, so
I'd go with the LCA710 just for the safety of that extra 400mA the
output can handle.

But, I forget how you said you were going to wire the thing up.

Do you have a schematic you can post?

It's complex because the final design also includes what's called a
"stand light" being able to power the light for few minutes while the
bicycle is not in motion, and being able to charge a NiMH battery. So
there's a lot of switching necessary to keep the battery from
discharging at all when the bike is not in use, as well as to
automatically switch from dynamo (AC) power to battery power when you
stop for a short time.

I've posted the schematics here, with current paths.

<http://nordicgroup.us/s78/mr16currentpaths/currentpaths.htm>

I'm still not sure if I'll go down this route but I'll try this circuit.
It looks cheaper to run the light of DC when in flash mode rather than
pulsing the AC. That loses the expensive relay and replaces it with a
PNP power transistor.

Thanks.