why triac blows by non-dimmable CFL even when its off?

R

ravi

Guest
I am trying to use non-dimmable CFL with triac dimmer but just to turn
on and off the lamp without dimming. 13W CFL lamp flickers even when
its off. This must have been caused by continuous leakage current
through snubber and triac. But when I was removing the CFL lamp from
socket, it popped triac even when it was off. I tried to repeat
problem i am seeing upward of 30A current spike for a brief time. I
don't understand where it does it get's this energy(current) to pop
triac? If I understand correctly, leakage current should be in mA.
 
"ravi"
I am trying to use non-dimmable CFL with triac dimmer but just to turn
on and off the lamp without dimming.
Click to expand...
** Bad idea - unless the triac is a "snubberless" type and is continuously
triggered rather than pulse fired.


13W CFL lamp flickers even when its off.
Click to expand...
** Very dimly - right ?


This must have been caused by continuous leakage current
through snubber and triac.
Click to expand...
** The current flowing through a snubber cap will faintly light most CFLs or
tern them into slow strobe lights.


But when I was removing the CFL lamp from
socket, it popped triac even when it was off. I tried to repeat
problem i am seeing upward of 30A current spike for a brief time. I
don't understand where it does it get's this energy(current) to pop
triac?
Click to expand...
** The CFL has a filter electro inside that comes right after the bridge
rectifier - when AC is first applied the cap charges very rapidly to the
peak value - hence the current spike.

The triac must be rated to take this spike current and with multiple CFLs on
the same circuit it can build up to hundreds of amps.


If I understand correctly, leakage current should be in mA.
Click to expand...
** See above.

Best keep CFLs away from all triacs.


..... Phil
 
"ravi"
But when I was removing the CFL lamp from
socket, it popped triac even when it was off. I tried to repeat
problem i am seeing upward of 30A current spike for a brief time.
Click to expand...

** BTW:

How did you measure that current spike ?

That ain't so trivial to do safely and accurately.

I use a specially made interface with a Hall Effect current transducer for
such jobs - mine has 100kHz bandwidth Hall transducer followed by a DC to
true RMS converter IC and a 3.5 digit LED display.

The most used feature though is a BNC output of the current wave that goes
straight to my scope.



..... Phil
 
"ravi"
"Phil Allison"
But when I was removing the CFL lamp from
socket, it popped triac even when it was off. I tried to repeat
problem i am seeing upward of 30A current spike for a brief time.

** BTW:

How did you measure that current spike ?

That ain't so trivial to do safely and accurately.

I use a specially made interface with a Hall Effect current transducer for
such jobs - mine has 100kHz bandwidth Hall transducer followed by a DC to
true RMS converter IC and a 3.5 digit LED .

The most used feature though is a BNC output of the current wave that goes
straight to my scope.
Click to expand...
I used simple 1ohm resistor and connected in in line to CFL to measure
with my scope.

** That ain't so safe - unless your scope is a plastic case portable.


While CFL lamp was in socket properly i saw hardly any
current flowing, but as i was unscrewing the lamp from socket, i
measured upward of 20A current spike and if I repeat it few times it
blows BTA16 (16A) snuberless triac easily.

** The BTA16 is a standard triac with a short pulse current rating in the
hundreds of amps.

The BTA16-600BW or BTA16-600CW are the only snubberless versions

Must be something fishy about your triggering arrangement ( or the lack of a
suitable snubber) for the triac to be damaged by a 20 amp current pulse that
lasts only 50 uS or so.

Possibly at the instant of connection, the dv/dt rating of the triac is
being exceeded and so it tries to commutate "on" under the extreme load
presented by the CFLs rectifier and internal electro cap.

That could kill it.


..... Phil
 
On Jan 22, 5:26 pm, "Phil Allison" <phi...@tpg.com.au> wrote:
"ravi"



But when I was removing the CFL lamp from
socket, it popped triac even when it was off. I tried to repeat
problem i am seeing upward of 30A current spike for a brief time.

** BTW:

How did you measure that current spike ?

That ain't so trivial to do safely and accurately.

I use a specially made interface with a Hall Effect current transducer for
such jobs -  mine has 100kHz bandwidth Hall transducer followed by a DC to
true RMS converter IC and a 3.5 digit LED .

The most used feature though is a BNC output of the current wave that goes
straight to my scope.

....  Phil
I used simple 1ohm resistor and connected in in line to CFL to measure
Click to expand...
with my scope. While CFL lamp was in socket properly i saw hardly any
current flowing, but as i was unscrewing the lamp from socket, i
measured upward of 20A current spike and if I repeat it few times it
blows BTA16 (16A) snuberless triac easily.
 
"ravi"

Yes, I am using CW version. It could be lack of suitable snubber as
you pointed out. This dimmer was designed years ago when there were no
CFLs. Current pulse lasts for over 200usec. Next I will try to
measure dv/dt and try to compare with maximum rating of datasheet.


** And when you have done all that experimenting - you might just wake up
to how much simpler it would be to use a relay instead.

BTW

Have you considered the issue of the triac's holding current yet ?

A CFL draws zero current up to 80% of the time it is on.



..... Phil
 
On Jan 22, 6:18 pm, "Phil Allison" <phi...@tpg.com.au> wrote:
"ravi"
"Phil Allison"





But when I was removing the CFL lamp from
socket, it popped triac even when it was off. I tried to repeat
problem i am seeing upward of 30A current spike for a brief time.

** BTW:

How did you measure that current spike ?

That ain't so trivial to do safely and accurately.

I use a specially made interface with a Hall Effect current transducer for
such jobs - mine has 100kHz bandwidth Hall transducer followed by a DC to
true RMS converter IC and a 3.5 digit LED .

The most used feature though is a BNC output of the current wave that goes
straight to my scope.

I used simple 1ohm resistor and connected in in line to CFL to measure
with my scope.

**  That ain't so safe  -  unless your scope is a plastic case portable.

While CFL lamp was in socket properly i saw hardly any
current flowing, but as i was unscrewing the lamp from socket, i
measured upward of 20A current spike and if I repeat it few  times it
blows BTA16 (16A) snuberless triac easily.

** The BTA16 is a standard triac with a short pulse current rating in the
hundreds of amps.

The BTA16-600BW  or  BTA16-600CW are the only snubberless versions

Must be something fishy about your triggering arrangement ( or the lack of a
suitable snubber) for the triac to be damaged by a 20 amp current pulse that
lasts only 50 uS or so.

Possibly at the instant of connection, the dv/dt rating of the triac is
being exceeded and so it tries to commutate "on" under the extreme load
presented by the CFLs rectifier and internal electro cap.

That could kill it.

....  Phil
Click to expand...
Yes, I am using CW version. It could be lack of suitable snubber as
you pointed out. This dimmer was designed years ago when there were no
CFLs. Current pulse lasts for over 200usec. Next I will try to
measure dv/dt and try to compare with maximum rating of datasheet.
Thanks
 
On Jan 22, 7:27 pm, "Phil Allison" <phi...@tpg.com.au> wrote:
"ravi"

Yes, I am using CW version. It could be lack of suitable snubber as
you pointed out. This dimmer was designed years ago when there were no
CFLs.  Current pulse lasts for over 200usec. Next I will try to
measure dv/dt and try to compare with maximum rating of datasheet.

** And when you have done all that experimenting  -  you might just wake up
to how much simpler it would be to use a relay instead.

BTW

Have you considered the issue of the triac's holding current yet ?

A CFL draws zero current up to 80% of the time it is on.

....  Phil
Click to expand...
I know relay would be nice, but let's just say its out of control. At
this point everything is on the table. I can probably add 35mA worth
of resistive load parallel to CFL and measure the effect if it solve
the problem..
Thanks
 
"ravi"
"Phil Allison"
"ravi"
Yes, I am using CW version. It could be lack of suitable snubber as
you pointed out. This dimmer was designed years ago when there were no
CFLs. Current pulse lasts for over 200usec. Next I will try to
measure dv/dt and try to compare with maximum rating of datasheet.

** And when you have done all that experimenting - you might just wake up
to how much simpler it would be to use a relay instead.

BTW

Have you considered the issue of the triac's holding current yet ?

A CFL draws zero current up to 80% of the time it is on.
Click to expand...
I know relay would be nice, but let's just say its out of control. At
this point everything is on the table. I can probably add 35mA worth
of resistive load parallel to CFL and measure the effect if it solve
the problem..

** You need to post your schematic somewhere for me or anyne to give any
real help.

I fear you have a drive circuit that draws its power via the lamp load -
and that is a disaster with CFLs.


..... Phil
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top