How to trick a 12V switch mode PS??

[Sam]

Hi all,

Trying to trick a 12V, 20-60VA output, switch mode power supply (AKA
electronic transformer)designed for halogen lamps, (ie. off the shelf from
your local electrical supplier) to drive this regulator etc. Is there an
L-R-C trick? Suggestions and direction is what I am after, specifics not
that important but would be a help as well. Circuit controls lamp with rest
of circuit consisting of sensor, driver and relay. Of course it would be
easy just to use a "wall wart" but why use 2 transformers when one should
suffice.

Front end - view in courier font
____
| |
|---------------->|--o----o---o---|7812|----o----o---- +12V,(load
| |----->|--| | | |____| | | 120mA)
| | | | | | |
o-- 12V~ --o 1000u = =0.1u | 0.1u = = 470u
| | | | | | |
| |-----|<--| | | | | |
o---------(------|<--o----o---o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)


Thanks for any advice.

 

[Mark Harriss]

You would need a load on the switchmode output
to make it work typically.

 

[Matt2 - Amstereo]

try a 5watt or bigger wire wound resister
might be enough of a resistive load... somthing like a 100ohm so that
you dont present a dead short

 

[The Real Andy]

On Tue, 09 Aug 2005 09:46:23 GMT, "Sam" <no@spam.com> wrote:

Hi all,

Trying to trick a 12V, 20-60VA output, switch mode power supply (AKA
electronic transformer)designed for halogen lamps, (ie. off the shelf from
your local electrical supplier) to drive this regulator etc. Is there an
L-R-C trick? Suggestions and direction is what I am after, specifics not
that important but would be a help as well. Circuit controls lamp with rest
of circuit consisting of sensor, driver and relay. Of course it would be
easy just to use a "wall wart" but why use 2 transformers when one should
suffice.

Front end - view in courier font
____
| |
|---------------->|--o----o---o---|7812|----o----o---- +12V,(load
| |----->|--| | | |____| | | 120mA)
| | | | | | |
o-- 12V~ --o 1000u = =0.1u | 0.1u = = 470u
| | | | | | |
| |-----|<--| | | | | |
o---------(------|<--o----o---o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)


Thanks for any advice.

I am looking at your circuit an wondering what the output from the PSU
looks like. Do you have a cro? Can you give us some idea of the
frequency output?

Assuming a fairly typical SMPS design, if the output is not filtered,
you will need a fairly quick diode and an inductor.

Something like:
10uH

|---------->|---o----^^^^^----o---|7812|----o----o---- +12V,(load
| | | |____| | | 120mA)
| | | | | |
o-- 12V~ --o =100uF =47uF | 0.1u = = 10u
| | | | | | |
| | | | | | |
o |----o-------------o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)

Whilst i am not sure of any common schottky diodes, a quick look at a
jaycar cat shows a 1N5819. If that fails, they also have a UF4003.

 

[Sam]

"The Real Andy" <will_get_back_to_you_on_This> wrote in message
news:dqjjf1124omrdegotd2j6bhufngm8nit2v@4ax.com...

On Tue, 09 Aug 2005 09:46:23 GMT, "Sam" <no@spam.com> wrote:


Hi all,

Trying to trick a 12V, 20-60VA output, switch mode power supply (AKA
electronic transformer)designed for halogen lamps, (ie. off the shelf
from
your local electrical supplier) to drive this regulator etc. Is there an
L-R-C trick? Suggestions and direction is what I am after, specifics not
that important but would be a help as well. Circuit controls lamp with
rest
of circuit consisting of sensor, driver and relay. Of course it would be
easy just to use a "wall wart" but why use 2 transformers when one should
suffice.

Front end - view in courier font
____
| |
|---------------->|--o----o---o---|7812|----o----o---- +12V,(load
| |----->|--| | | |____| | | 120mA)
| | | | | | |
o-- 12V~ --o 1000u = =0.1u | 0.1u = = 470u
| | | | | | |
| |-----|<--| | | | | |
o---------(------|<--o----o---o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)


Thanks for any advice.


I am looking at your circuit an wondering what the output from the PSU
looks like. Do you have a cro? Can you give us some idea of the
frequency output?

Assuming a fairly typical SMPS design, if the output is not filtered,
you will need a fairly quick diode and an inductor.

Something like:
10uH
|---------->|---o----^^^^^----o---|7812|----o----o---- +12V,(load
| | | |____| | | 120mA)
| | | | | |
o-- 12V~ --o =100uF =47uF | 0.1u = = 10u
| | | | | | |
| | | | | | |
o |----o-------------o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)

Whilst i am not sure of any common schottky diodes, a quick look at a
jaycar cat shows a 1N5819. If that fails, they also have a UF4003.

The output on a CRO looks like multiple sine waves - very hard to lock on. I
imagine it must be fairly high freq as they have to be EMC compliant. These
transformers are becoming very common being sold with just about every
recessed downlight you buy, however they are designed to drop off under a
certain value of VA - not sure why.

It's interesting what you suggest - only one schottky diode required? Are
you suggesting that this will trick the SMPS into firing or are you saying
this is what I should have regardless.

Cheers.

 

[Phil Allison]

"Sam"

The output on a CRO looks like multiple sine waves - very hard to lock on.

** Set the CRO's timebase to "line" or "AC" trigger.

The waveform is amplitude modulated at double mains frequency - so this
will lock it .

( Hope you are not using one of those awful hand held LCD things. )

I
imagine it must be fairly high freq as they have to be EMC compliant.

** Total non sequitur.

The oscillation frequency is about 50 to 100 kHz - keeps interference
below the broadcast band.




.............. Phil

 

[Terry Given]

Sam wrote:

"The Real Andy" <will_get_back_to_you_on_This> wrote in message
news:dqjjf1124omrdegotd2j6bhufngm8nit2v@4ax.com...

On Tue, 09 Aug 2005 09:46:23 GMT, "Sam" <no@spam.com> wrote:


Hi all,

Trying to trick a 12V, 20-60VA output, switch mode power supply (AKA
electronic transformer)designed for halogen lamps, (ie. off the shelf

from

your local electrical supplier) to drive this regulator etc. Is there an
L-R-C trick? Suggestions and direction is what I am after, specifics not
that important but would be a help as well. Circuit controls lamp with

rest

of circuit consisting of sensor, driver and relay. Of course it would be
easy just to use a "wall wart" but why use 2 transformers when one should
suffice.

Front end - view in courier font
____
| |
|---------------->|--o----o---o---|7812|----o----o---- +12V,(load
| |----->|--| | | |____| | | 120mA)
| | | | | | |
o-- 12V~ --o 1000u = =0.1u | 0.1u = = 470u
| | | | | | |
| |-----|<--| | | | | |
o---------(------|<--o----o---o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)


Thanks for any advice.


I am looking at your circuit an wondering what the output from the PSU
looks like. Do you have a cro? Can you give us some idea of the
frequency output?

Assuming a fairly typical SMPS design, if the output is not filtered,
you will need a fairly quick diode and an inductor.

Something like:
10uH

|---------->|---o----^^^^^----o---|7812|----o----o---- +12V,(load
| | | |____| | | 120mA)
| | | | | |
o-- 12V~ --o =100uF =47uF | 0.1u = = 10u
| | | | | | |
| | | | | | |
o |----o-------------o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)

Whilst i am not sure of any common schottky diodes, a quick look at a
jaycar cat shows a 1N5819. If that fails, they also have a UF4003.




The output on a CRO looks like multiple sine waves - very hard to lock on. I
imagine it must be fairly high freq as they have to be EMC compliant. These
transformers are becoming very common being sold with just about every
recessed downlight you buy, however they are designed to drop off under a
certain value of VA - not sure why.

If you use a digital scope and cant trigger on a signal, suspect
aliasing - crank the timescale right down and try again. The first time
I ever used a digital scope I did exactly that, couldnt trigger and my
12MHz oscillator was a few kHz.... very red-faced when a tech twiddled
the knob for me. doh.

It's interesting what you suggest - only one schottky diode required? Are
you suggesting that this will trick the SMPS into firing or are you saying
this is what I should have regardless.

Cheers.

Cheers
Terry

 

[Terry Given]

Phil Allison wrote:

"Sam"


The output on a CRO looks like multiple sine waves - very hard to lock on.



** Set the CRO's timebase to "line" or "AC" trigger.

The waveform is amplitude modulated at double mains frequency - so this
will lock it .

( Hope you are not using one of those awful hand held LCD things. )



I
imagine it must be fairly high freq as they have to be EMC compliant.



** Total non sequitur.

Indeed. LOL

The oscillation frequency is about 50 to 100 kHz - keeps interference
below the broadcast band.

............. Phil

Most conducted EMI standards start at 150kHz (MIL-STD-461 or whatever it
morphed into goes so low - 9kHz IIRC - AC line harmonics show up). The
tradeoff is > 20kHz (inaudible), < 150kHz. Just under 30kHz is a good
number, the 5th harmonic still doesnt matter.

Above a few hundred kHz things get rather tricky, and minimum duty-cycle
often becomes a problem - eg at 500kHz T = 2us so a 50% duty cycle is
1us wide. Finite switching times, current sense delays etc. typically
limit minimum on-time to 300ns - 500ns, which makes the minimum duty
cycle 15-25% or so.

Cheers
Terry

 

[Chris Jones]

The Real Andy <will_get_back_to_you_on_This> wrote:

On Tue, 09 Aug 2005 09:46:23 GMT, "Sam" <no@spam.com> wrote:


Hi all,

Trying to trick a 12V, 20-60VA output, switch mode power supply (AKA
electronic transformer)designed for halogen lamps, (ie. off the shelf from
your local electrical supplier) to drive this regulator etc. Is there an
L-R-C trick? Suggestions and direction is what I am after, specifics not
that important but would be a help as well. Circuit controls lamp with
rest of circuit consisting of sensor, driver and relay. Of course it would
be easy just to use a "wall wart" but why use 2 transformers when one
should suffice.

Front end - view in courier font
____
| |
|---------------->|--o----o---o---|7812|----o----o---- +12V,(load
| |----->|--| | | |____| | | 120mA)
| | | | | | |
o-- 12V~ --o 1000u = =0.1u | 0.1u = = 470u
| | | | | | |
| |-----|<--| | | | | |
o---------(------|<--o----o---o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)


Thanks for any advice.


I am looking at your circuit an wondering what the output from the PSU
looks like. Do you have a cro? Can you give us some idea of the
frequency output?

Assuming a fairly typical SMPS design, if the output is not filtered,
you will need a fairly quick diode and an inductor.

Something like:
10uH
|---------->|---o----^^^^^----o---|7812|----o----o---- +12V,(load
| | | |____| | | 120mA)
| | | | | |
o-- 12V~ --o =100uF =47uF | 0.1u = = 10u
| | | | | | |
| | | | | | |
o |----o-------------o------o------o----o----- 0
| |
| |
| |
| |
--/ --O---
relay lamp (50W)

Whilst i am not sure of any common schottky diodes, a quick look at a
jaycar cat shows a 1N5819. If that fails, they also have a UF4003.

You have to be slightly careful if you use a single diode to rectify the
output of a transformer: The average DC current in the secondary winding of
the transformer is equal to the load current and so it is sometimes
possible to saturate the transformer core which can be destructive. If you
use a bridge rectifier or two diodes wired as a voltage doubler then you
won't have that problem. You need fast diodes though!

Also you may or may not be aware of a difference between an ordinary SMPS
and one of these halogen inverter things: They sometimes (often?) don't
fully smooth the rectified mains, they only use a small capacitor. This
means that the high frequency signal out of the transformer comes and goes
100 times a second, and this means that the filter cap at the output of
your new rectifier will have to be much bigger than in a normal SMPS.

I guess that the small filter cap in the halogen inverter thing is probably
to make the power factor nice or to reduce the harmonic content of the
mains current or maybe just to save the cost of a capacitor.

Chris

 

[The Real Andy]

On Sun, 14 Aug 2005 23:41:15 +0100, Chris Jones
<lugnut808@nospam.yahoo.com> wrote:

<snip>

You have to be slightly careful if you use a single diode to rectify the
output of a transformer: The average DC current in the secondary winding of
the transformer is equal to the load current and so it is sometimes
possible to saturate the transformer core which can be destructive. If you
use a bridge rectifier or two diodes wired as a voltage doubler then you
won't have that problem. You need fast diodes though!

Also you may or may not be aware of a difference between an ordinary SMPS
and one of these halogen inverter things: They sometimes (often?) don't
fully smooth the rectified mains, they only use a small capacitor. This
means that the high frequency signal out of the transformer comes and goes
100 times a second, and this means that the filter cap at the output of
your new rectifier will have to be much bigger than in a normal SMPS.

I guess that the small filter cap in the halogen inverter thing is probably
to make the power factor nice or to reduce the harmonic content of the
mains current or maybe just to save the cost of a capacitor.

Can I take a wild stab in the dark and guess that the small cap maybe
there to reduce EMI? (self confessed no nothing about halogen
transformers )

Chris

 

[Terry Given]

The Real Andy wrote:

On Sun, 14 Aug 2005 23:41:15 +0100, Chris Jones
lugnut808@nospam.yahoo.com> wrote:

snip

You have to be slightly careful if you use a single diode to rectify the
output of a transformer: The average DC current in the secondary winding of
the transformer is equal to the load current and so it is sometimes
possible to saturate the transformer core which can be destructive. If you
use a bridge rectifier or two diodes wired as a voltage doubler then you
won't have that problem. You need fast diodes though!

Also you may or may not be aware of a difference between an ordinary SMPS
and one of these halogen inverter things: They sometimes (often?) don't
fully smooth the rectified mains, they only use a small capacitor. This
means that the high frequency signal out of the transformer comes and goes
100 times a second, and this means that the filter cap at the output of
your new rectifier will have to be much bigger than in a normal SMPS.

I guess that the small filter cap in the halogen inverter thing is probably
to make the power factor nice or to reduce the harmonic content of the
mains current or maybe just to save the cost of a capacitor.


Can I take a wild stab in the dark and guess that the small cap maybe
there to reduce EMI? (self confessed no nothing about halogen
transformers )

Chris

I'd go for option D - all of the above.

Cheers
Terry

 

[Terry Given]

Chris Jones wrote:

The Real Andy <will_get_back_to_you_on_This> wrote:


On Sun, 14 Aug 2005 23:41:15 +0100, Chris Jones
lugnut808@nospam.yahoo.com> wrote:

snip

You have to be slightly careful if you use a single diode to rectify the
output of a transformer: The average DC current in the secondary winding
of the transformer is equal to the load current and so it is sometimes
possible to saturate the transformer core which can be destructive. If
you use a bridge rectifier or two diodes wired as a voltage doubler then
you won't have that problem. You need fast diodes though!

Also you may or may not be aware of a difference between an ordinary SMPS
and one of these halogen inverter things: They sometimes (often?) don't
fully smooth the rectified mains, they only use a small capacitor. This
means that the high frequency signal out of the transformer comes and goes
100 times a second, and this means that the filter cap at the output of
your new rectifier will have to be much bigger than in a normal SMPS.

I guess that the small filter cap in the halogen inverter thing is
probably to make the power factor nice or to reduce the harmonic content
of the mains current or maybe just to save the cost of a capacitor.

Can I take a wild stab in the dark and guess that the small cap maybe
there to reduce EMI? (self confessed no nothing about halogen
transformers )


Chris


Ok, sorry, I could have said that more clearly. Of course the presence of
*some* capacitor is necessary for EMI reduction, efficiency amongst other
reasons. I wsa commenting that the cap in a normal SMPS is much bigger
because the capacitor after the mains rectifier in your laptop's power
supply has to store enough energy to power the load for about 10ms (half a
mains cycle), whereas the halogen "transformer" doesn't bother with such a
big cap and so the high frequency waveform powering the light goes off
every time the mains voltage goes through zero.

I was speculating that by not using a big capacitor, there are not the usual
large current peaks which occur during the brief period when the bridge
rectifier is conducting, when a very large filter cap is used. I think
there are rules about the harmonic content of the mains current and the
power factor, at least in some countries.

Chris

and you are right.

Cheers
Terry

 

[Chris Jones]

The Real Andy <will_get_back_to_you_on_This> wrote:

On Sun, 14 Aug 2005 23:41:15 +0100, Chris Jones
lugnut808@nospam.yahoo.com> wrote:

snip

You have to be slightly careful if you use a single diode to rectify the
output of a transformer: The average DC current in the secondary winding
of the transformer is equal to the load current and so it is sometimes
possible to saturate the transformer core which can be destructive. If
you use a bridge rectifier or two diodes wired as a voltage doubler then
you won't have that problem. You need fast diodes though!

Also you may or may not be aware of a difference between an ordinary SMPS
and one of these halogen inverter things: They sometimes (often?) don't
fully smooth the rectified mains, they only use a small capacitor. This
means that the high frequency signal out of the transformer comes and goes
100 times a second, and this means that the filter cap at the output of
your new rectifier will have to be much bigger than in a normal SMPS.

I guess that the small filter cap in the halogen inverter thing is
probably to make the power factor nice or to reduce the harmonic content
of the mains current or maybe just to save the cost of a capacitor.

Can I take a wild stab in the dark and guess that the small cap maybe
there to reduce EMI? (self confessed no nothing about halogen
transformers )


Chris

Ok, sorry, I could have said that more clearly. Of course the presence of
*some* capacitor is necessary for EMI reduction, efficiency amongst other
reasons. I wsa commenting that the cap in a normal SMPS is much bigger
because the capacitor after the mains rectifier in your laptop's power
supply has to store enough energy to power the load for about 10ms (half a
mains cycle), whereas the halogen "transformer" doesn't bother with such a
big cap and so the high frequency waveform powering the light goes off
every time the mains voltage goes through zero.

I was speculating that by not using a big capacitor, there are not the usual
large current peaks which occur during the brief period when the bridge
rectifier is conducting, when a very large filter cap is used. I think
there are rules about the harmonic content of the mains current and the
power factor, at least in some countries.

Chris