Searching \ for '[EE] TRIAC related problem' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=triac+related+problem
Search entire site for: 'TRIAC related problem'.

Exact match. Not showing close matches.
PICList Thread
'[EE] TRIAC related problem'
2007\07\22@174227 by Trevor Page

flavicon
face
Hi list,

I have designed a PIC controlled device that is powered from the mains
supply. One of the functions of the PIC is to illuminate a mains light bulb
by means of a TRIAC.

When the mains supply is first switched on to the device, the bulb must not
illuminate.

The problem is that, depending on what part of the mains cycle we are at
when the mains is switched on, the TRIAC will sometimes do a false trigger
until the mains crosses zero again, and this causes the bulb to very briefly
flash once. This is very undesirable and I need to prevent it doing this.

I'm not experienced with the use of TRIACs but having done some reading, I
understand that my problem is dV/dt triggering. So I have tried a snubber
circuit across the TRIAC to try to prevent this, and while this works well
when an incandescent bulb is used, the small current that is always present
with the snubber circuit plays hell with energy saver type lamps (they will
illuminate dimly, or flash periodically).

I did order a number of different TRIACs from Farnell to see if any
particular type would exhibit the dV/dt problem but they all do. These are
just basic regular TRIACs that I've been using, though (including one or two
types that offered high dV/dt immunity).  

So is there any particular special type of TRIAC (like a phototriac, maybe)
I could use to solve this problem?

If this problem is going to be far too tricky to solve by simply using
another kind of TRIAC, I thought of another approach. This approach would be
to have two TRIACs in series. One TRIAC has the snubber circuit across it to
avoid dV/dt triggering and the other TRIAC sits after it to avoid residual
snubber current. Could this be a solution or is it a daft idea?

Many thanks,

Trevor.

2007\07\22@180735 by Peter P.

picon face
MOC3030, 3040 zero cross switches

Peter P.


2007\07\22@184928 by Richard Prosser

picon face
Peter,
I'm guessing, but the flashing of the CFLs is probably due to the
internal HV being pumped up by the leakage current. Placing a high
value resistor across the lamp _may_ be enough to stop it flashing.

RP

On 23/07/07, Peter P. <spam_OUTplpeter2006TakeThisOuTspamyahoo.com> wrote:
> MOC3030, 3040 zero cross switches
>
> Peter P.
>
>
> -

2007\07\22@191425 by Peter P.

picon face
> I'm guessing, but the flashing of the CFLs is probably due to the
> internal HV being pumped up by the leakage current. Placing a high
> value resistor across the lamp _may_ be enough to stop it flashing.

You may find that the energy saver lamps have diac starters and will start from
'nothing' charging the filter cap (if they have one) to the firing voltage. I
know of cases where the wiring ac leakage is enough (with lamp off) to make the
lamp flash once every 30 seconds or so in the dark. I Installed a FL capacitor
to return in the circuit to make that go away.

Ime the only way to prevent them from working in leaky circuit is the addition
of a leak across them (series RC shunting them). This should cure the op's
problem. But having two snubbers with X1 or X2 caps will increase cost and size.
So I think that the real problem is slowing down the input. This is normally
done with a large choke in series with mains and with a relatively small snubber
across the mains after it.

Basically I think that there is no snake oil solution. It is not reasonable to
control a load that can vary from 3W to 3000W with 'zero' leakage (which may
mean 0.03W for the 3W load - 140uA of leakage will do that - that's within range
for *insulated* chassis suitable for human contact). Basically accomodating
3000W and 3W at leakage as shown means an isolation ratio of >100,000 or 80dB.
Not very reasonable for ANY single part active switch as you can see, especially
not for $0.8 class parts.

Although a DPDT relay would do just exactly that. So either it's a low power
only circuit and it uses an optomos (which is usually not prone to dv/dt - IF
the problem is dV/dt at all - could be checked by adding the choke upstream and
seeing what helps) directly, or it's a high power one and then there *will* be
leakage from snubbers etc, or it's an any load type and a good old relay and a
varistor and/or small snubber fix the problem for good. But I don't think that
there is a $0.82 solution to the problem as formulated. imho. In fact, the op
did not ask for a $0.8 solution, I just inferred it.

Peter P.


2007\07\22@195551 by Timothy J. Weber

face picon face
Trevor Page wrote:
> So is there any particular special type of TRIAC (like a phototriac, maybe)
> I could use to solve this problem?

Once you add a snubber, opto interface, etc... you might just use a
canned solid-state relay.  E.g., the Omron G3MB series is ~$4 in
singles, depending on your desired flavor.
--
Timothy J. Weber
http://timothyweber.org

2007\07\23@024914 by Trevor Page

flavicon
face
Gents,

Many thanks for your ideas and discussion so far. This is greatly
appreciated.

I shall be exploring some of the other ideas suggested (opto interface,
solid state relay, etc). I'll place an order for some components to play
with today. For starters, Farnell stock those solid state relays and I like
the way that they have a relatively high 'impedance' on the 'coil' side,
since there is only about 15mA to 20mA available to switch a device. In this
unit I'm getting my low voltage from a resistor/zener and X2 capacitor
divider.

In the meantime, what do you guys think to the two TRIAC method I suggested,
whereby one TRIAC contains a snubber to eliminate dV/dt triggering, and then
a second TRIAC is placed after it as a barrier to the small snubber current?
Their gate terminals would be connected in common. This is something I'll
have a go at today, though I just get a feeling that this method has a flaw
that I haven't thought of! It'd be lovely if it worked, though: two very
cheap and compact DPAK TRIACs and just one snubber circuit.

Regards

Trev

{Original Message removed}

2007\07\23@133758 by Harold Hallikainen

face
flavicon
face
I designed a series of light dimmers (the "Shoebox" series at
http://www.dovesystems.com). I originally had an RC snubber across the
triac, but recently we went to triacs with high dV/dT ratings (I think
it's the Q4015L5) and were able to remove the snubbers. One thing that
helps a lot is to have a trigger pull-down resistor. I believe we used 100
ohms. The opto pulls the trigger towards MT2, and the pull-down pulls the
trigger towards MT1, as I recall (I don't have the schematics in front of
me). This also helps prevent false triggering at high temperatures.

We were having problems with snubber leakage current when driving mains
dimmable fluorescent fixtures. The ballast power supply would slowly
charge with snubber leakage current, then the lamp would flash. The cycle
would repeat every minute or so.

Harold


--
FCC Rules Updated Daily at http://www.hallikainen.com - Advertising
opportunities available!

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...