Protection triac gated mains, on 600W power amp problem

[N_Cook]

Last week I confirmed that the over-temperature sensing cct worked and the
triac cut the supply to the mains transformer. This is a Carlsbro S600M I
can find no info on . I was running on a variac and with about 90 percent
mains and with no load etc there was 0.1 amp current drawn from the mains
and a noticeable amount of transformer noise I took to be saturation noise
and current drain, more than I'd like but until I could confirm with the
owner about previous useage, I let be for then. This time I powered up and 1
amp being drawn with no load at only 70 percent mains. Disconnected the
secondaries from the amp and still the same.
Disconnected triac circuit and the transformer is happy to 100 percent
mains, fed directly, and only few mA of current drawn. What can be going on
? To get to the triac componentry proper requires a major strip down first.
Could it be half cycle firing and causing some sort of magnetic
mutual/resonant/self-induction effect that is causing excessive current? A
matter of isolation transformer and scoping some dropper in line with the
primary to see ? No obvious heating on the triac, only a small heatsink. I
was monitoring the DC last week and it was already a bit over the 63V rating
of each electro. This time direct to mains, no triac, it was much more
normal, forgot to note but about +/- 55V probably at 100 percent mains

 

[N_Cook]

Last week I confirmed that the over-temperature sensing cct worked and the
triac cut the supply to the mains transformer. This is a Carlsbro S600M I
can find no info on . I was running on a variac and with about 90 percent
mains and with no load etc there was 0.1 amp current drawn from the mains
and a noticeable amount of transformer noise I took to be saturation noise
and current drain, more than I'd like but until I could confirm with the
owner about previous useage, I let be for then. This time I powered up and 1
amp being drawn with no load at only 70 percent mains. Disconnected the
secondaries from the amp and still the same.
Disconnected triac circuit and the transformer is happy to 100 percent
mains, fed directly, and only few mA of current drawn. What can be going on
? To get to the triac componentry proper requires a major strip down first.
Could it be half cycle firing and causing some sort of magnetic
mutual/resonant/self-induction effect that is causing excessive current? A
matter of isolation transformer and scoping some dropper in line with the
primary to see ? No obvious heating on the triac, only a small heatsink. I
was monitoring the DC last week and it was already a bit over the 63V rating
of each electro. This time direct to mains, no triac, it was much more
normal, forgot to note but about +/- 55V probably at 100 percent mains

Scoped the secondaries and a cross-over distortion developes on increasing
the variac voltage, not particularly strongly, but definetely there, and
increases with voltage increase and "saturation" current/mechanical noise.
Placing an isolation transformer on the primary with a 25R/20W dropper to
"neutral" and scoping, then the problem was obvious ,now unfiltered by the
transformer. Could now run up much higher variac voltage because of damping
by the added R but even then there was 3mS duration spikes of peaks twice
that of the mains ac pk-pk of mains (UK) 20mS period. So some sort of bucky
inverter effect going on. What to look for around the triac when I get to it
?. To get to triac area, 8 XLR connectors with 3 stout wire standoff
conductors each to cut, probably by grinding through the most enclosed ones
(then desolder both cut ends and replace eventually ). Because audio input
and output and mains control all on the same board with no obvious isolation
band between the sections

 

[N_Cook]

What can go wrong with a triac control cct , visual and cold component test
of these comps shows nothing obviously amiss
TIC2360, BC214C,2x 1N4005,5V.1W zener, 2 thermal switches, 2x 100R w/w,
2K7,47R, 2.2uF 400V polyester and 2x 220uF,10V electros
I will have to draw it out as a schema

 

[Jamie]

What can go wrong with a triac control cct , visual and cold component test
of these comps shows nothing obviously amiss
TIC2360, BC214C,2x 1N4005,5V.1W zener, 2 thermal switches, 2x 100R w/w,
2K7,47R, 2.2uF 400V polyester and 2x 220uF,10V electros
I will have to draw it out as a schema

Triacs can develop latching issues and thus, depending on what you're
doing with them, they may take out a protection circuit (fuse), or
if used to select +/- outputs to generate a selective DC polarity, could
hang up on one side and this cause erratic behavior to the device
connected to it.

The transistor BC214, is a PNP type, know to open in the BASE/EMITTER
how ever, you've stated that you have done the cold test on that. The
other problem is, they lose gain and thus, what ever function they have,
they'll have no gain to perform it. This normally happens to a poorly
designed circuit or one that got over voltage, unexpectedly.

I would check the Zener to make sure it does have 5 volts across it
when power is on.

And also, Are you sure that is a zener? by the looks of the parts
list, I would say this may be a simple phase control device? Maybe you
have a failing DIAC/SIAC which can be past off as a diode? That 2.2 uf
Cap at 400V, is leading me towards that direction. Since that is a
common value for the trigger pulse on the gate which gets generated by
the DIAC.

 

[N_Cook]

Triacs can develop latching issues and thus, depending on what you're
doing with them, they may take out a protection circuit (fuse), or
if used to select +/- outputs to generate a selective DC polarity, could
hang up on one side and this cause erratic behavior to the device
connected to it.

The transistor BC214, is a PNP type, know to open in the BASE/EMITTER
how ever, you've stated that you have done the cold test on that. The
other problem is, they lose gain and thus, what ever function they have,
they'll have no gain to perform it. This normally happens to a poorly
designed circuit or one that got over voltage, unexpectedly.

I would check the Zener to make sure it does have 5 volts across it
when power is on.

And also, Are you sure that is a zener? by the looks of the parts
list, I would say this may be a simple phase control device? Maybe you
have a failing DIAC/SIAC which can be past off as a diode? That 2.2 uf
Cap at 400V, is leading me towards that direction. Since that is a
common value for the trigger pulse on the gate which gets generated by
the DIAC.

The overlay next to the zener says Z1 and I can read 5V1 on the body but not
anything else. Its a bit difficult to do "diode" checks because of the 47
and 100Rs around, but I suspect failed C-E of the transistor if anything.
But I will sketch out cct first tomorrow , rather than start replacing parts
on hunches.
Does not actually knock out fuses, mains transformer runs hot .

 

[N_Cook]

TIC236D = 400V 12A , misread it

 

[N_Cook]

This is probably the circuit, split primary is used (no connection at 47R) ,
2.2uF,100V polyester for high C unpolarised, presumably for a quasi DC
reference, BC214 emitter to top , view/message source, may have equi-spaced
font


240V

| mains primary
|C
|C------- 120V -----------------------------------------|
|C 47R |
| ___ |
| -|___|-| |
| /´ | |
| ,´ | |
| ,´ |< BC214 |
| / -| |
| / / |\ |
100R,25W| ,´ / ------- |
| ,´ / | | |
|----- / |-------| | V 2.2uF , 100V |
.-. _|_/ --- --10uF | - || |
| | V_A 10uF--- --- V ---||---| |
| | | | ___ | - | || .|. |
'-' |--------- -|___|- ZD | V | | |
|----| | - | |100R,10W |
| 2K7 | | '-' |
|TIC236D | | | |
| |-----|---|--- |
0V | |
| |
|------------------|
120V

 

[N_Cook]

I replaced the BC214 with a BC212 and as no TIC236 to hand , with a TIC226
for proving purposes, temp. There is now no sign of the broad spike trying
with the 25 ohm dropper, even to 100 percent mains. The 214 tests as 420 hFE
and I'll assume the triac is faulty. My triac test is little more than
go/no-go, as only substituting in a lamp dimmer set up, so may well work
with light duty and non inductive load. The 5.1V device tested as a 5V zener
, I wonder why 2x 1N4005 and not 2x 1N4001 in that area.

 

[Jamie]

I replaced the BC214 with a BC212 and as no TIC236 to hand , with a TIC226
for proving purposes, temp. There is now no sign of the broad spike trying
with the 25 ohm dropper, even to 100 percent mains. The 214 tests as 420 hFE
and I'll assume the triac is faulty. My triac test is little more than
go/no-go, as only substituting in a lamp dimmer set up, so may well work
with light duty and non inductive load. The 5.1V device tested as a 5V zener
, I wonder why 2x 1N4005 and not 2x 1N4001 in that area.

A Triac can short half of it's element. We have situations like this
with universal motors (series motor) for vacuum loaders. They use a
simple thyristor phase control with a small uC to operate them..
Once in a while, one will short in a why that allows only a DC to
appear to the motor. (Half wave /cycle), since these motors can operate
on DC/AC, hence universal motors, they normally operate at half speed
when the triac fails with no gate control signal. In other words, they
start up at half speed and just keep running, and when demand comes on,
they go up to full speed.
Diode test between B1 B2 terminals don't always reveal a shorted unit
because the meter is not placing the triac under the same conditions. A
short is obvious when performing a diode test between B1 and B2.

Also, Triac's are known to fire off the other side of the thyristor
at the transition, when you don't want it to. In other words, you may
want lets say the + side and then decide at the base of the + side, not
want the (-) side how ever, at times the (-) side will fire and vise versa.

Sensitive Gate triacs are tricky business.

They also have snubberless types.

 

[N_Cook]

A Triac can short half of it's element. We have situations like this
with universal motors (series motor) for vacuum loaders. They use a
simple thyristor phase control with a small uC to operate them..
Once in a while, one will short in a why that allows only a DC to
appear to the motor. (Half wave /cycle), since these motors can operate
on DC/AC, hence universal motors, they normally operate at half speed
when the triac fails with no gate control signal. In other words, they
start up at half speed and just keep running, and when demand comes on,
they go up to full speed.
Diode test between B1 B2 terminals don't always reveal a shorted unit
because the meter is not placing the triac under the same conditions. A
short is obvious when performing a diode test between B1 and B2.

Also, Triac's are known to fire off the other side of the thyristor
at the transition, when you don't want it to. In other words, you may
want lets say the + side and then decide at the base of the + side, not
want the (-) side how ever, at times the (-) side will fire and vise versa.

Sensitive Gate triacs are tricky business.

They also have snubberless types.

Thanks for that, I'll try exploring further. No snubber in this use.
Would you know of a term for the observed broad spike, of amplitude twice or
more the normal pk-pk amplitude, generated and passing through to the
secondaries , increasing the rectified voltage to exceed normal ratings ?
Pseudo bucky inverter fault ?

 

[N_Cook]

A Triac can short half of it's element. We have situations like this
with universal motors (series motor) for vacuum loaders. They use a
simple thyristor phase control with a small uC to operate them..
Once in a while, one will short in a why that allows only a DC to
appear to the motor. (Half wave /cycle), since these motors can operate
on DC/AC, hence universal motors, they normally operate at half speed
when the triac fails with no gate control signal. In other words, they
start up at half speed and just keep running, and when demand comes on,
they go up to full speed.
Diode test between B1 B2 terminals don't always reveal a shorted unit
because the meter is not placing the triac under the same conditions. A
short is obvious when performing a diode test between B1 and B2.

Also, Triac's are known to fire off the other side of the thyristor
at the transition, when you don't want it to. In other words, you may
want lets say the + side and then decide at the base of the + side, not
want the (-) side how ever, at times the (-) side will fire and vise versa.

Sensitive Gate triacs are tricky business.

They also have snubberless types.

This amp is sometimes used on a portable generator , I wonder if that is a
factor in causing the fault condition.
As an afterthought utility companies would not like a system that throws
high voltage spikes back into their lines

 

[N_Cook]

A Triac can short half of it's element. We have situations like this
with universal motors (series motor) for vacuum loaders. They use a
simple thyristor phase control with a small uC to operate them..
Once in a while, one will short in a why that allows only a DC to
appear to the motor. (Half wave /cycle), since these motors can operate
on DC/AC, hence universal motors, they normally operate at half speed
when the triac fails with no gate control signal. In other words, they
start up at half speed and just keep running, and when demand comes on,
they go up to full speed.
Diode test between B1 B2 terminals don't always reveal a shorted unit
because the meter is not placing the triac under the same conditions. A
short is obvious when performing a diode test between B1 and B2.

Also, Triac's are known to fire off the other side of the thyristor
at the transition, when you don't want it to. In other words, you may
want lets say the + side and then decide at the base of the + side, not
want the (-) side how ever, at times the (-) side will fire and vise versa.

Sensitive Gate triacs are tricky business.

They also have snubberless types.

no dioding either way at 15V unless < 2uamp

 

[Jamie]

Thanks for that, I'll try exploring further. No snubber in this use.
Would you know of a term for the observed broad spike, of amplitude twice or
more the normal pk-pk amplitude, generated and passing through to the
secondaries , increasing the rectified voltage to exceed normal ratings ?
Pseudo bucky inverter fault ?

A bad transformer can cause that. Do you have a megga meter you can use
on the transformer? Also, if I can remember that circuit, a cap failing
only at higher voltages can generate intermitting shorts and if in the
proper circuit, they'll just recover and go on. This kind of activity
can cause all kinds of strange spikes to appear.

Bridge rectifiers have been known to intermitly short, and then cause
a big spike to appear at the secondary. The filter caps (large one's)
can also do this..
So, you have a lot to look at here. If the transformer seems to be
running hot, I would disconnect the secondary of the transformer and
allow it to run unloaded, then perform some test via a scope and do
temperature checks.

P.S.
I've found transformers to become shorted that shows no sign of
electrical damage from heat but yet, exhibit very low Lx at the primary
side with secondary fully floating.. When doing a Lx check on the
primary the Lx value should not exceed a value that would generate over
the max expected load.. Some designs actually saturate or near saturate
the power transformer with no load which will cause it to heat up. This
is normal in design for a couple of reasons but should not measure
values of Sqr(Lx^2+R^2) that would exceed Max operating point. Normally,
this type of design is to reduce side effects from coupling circuits and
reduce cost in materials.

 

[N_Cook]

A bad transformer can cause that. Do you have a megga meter you can use
on the transformer? Also, if I can remember that circuit, a cap failing
only at higher voltages can generate intermitting shorts and if in the
proper circuit, they'll just recover and go on. This kind of activity
can cause all kinds of strange spikes to appear.

Bridge rectifiers have been known to intermitly short, and then cause
a big spike to appear at the secondary. The filter caps (large one's)
can also do this..
So, you have a lot to look at here. If the transformer seems to be
running hot, I would disconnect the secondary of the transformer and
allow it to run unloaded, then perform some test via a scope and do
temperature checks.

P.S.
I've found transformers to become shorted that shows no sign of
electrical damage from heat but yet, exhibit very low Lx at the primary
side with secondary fully floating.. When doing a Lx check on the
primary the Lx value should not exceed a value that would generate over
the max expected load.. Some designs actually saturate or near saturate
the power transformer with no load which will cause it to heat up. This
is normal in design for a couple of reasons but should not measure
values of Sqr(Lx^2+R^2) that would exceed Max operating point. Normally,
this type of design is to reduce side effects from coupling circuits and
reduce cost in materials.

This was a consistent spike (3mS or so broad ) , once per mains cycle. It
was possibly variable in degree as 0.1 amp or so of "saturation" current one
time and 1 amp or so another time. Repeated firing up with a TIC226 and no
return , will try with a TOP66 higher rating triac , perhaps today and see
if still no return to spikes, before a replacement TIC236 placed in there.
The transformer powered straight from the mains was perfect, no saturation
current / core or wiring rattle or whatever gave that nasty noise in a
toroidal tx

 

[N_Cook]

for the archives this is a corrected (but not necessarily correct) version
of the soft start circuit, now makes more sense

240V
| mains primary
|C
|C------- 120V ac---------------------------------------|
|C |
| |
| 47R |
| ___ |
| -|___|-| |
| /´ | |
| ,´ | |
| ,´ |< BC214 |
| / -| |
| / / |\ |
100R,25W| ,´ / ------- |
| ,´ / | | |
|----- / / | V 2.2uF , 250V |
.-. _|_/ / | - || ___ |
| | V_A / V ----||------|___|------|
| | | / - | || 94R,10W
'-' | | ZD | V
|----| | | -
| | | |
|TIC236D |----------|-----|
| |
| |
| |-------|
| --- --220uF
| 220uF --- ---
| | ___ |
|------------|___|-
| 2K7
0 V