50W electronic halogen transformer

[nukie]

During lunch, I went to my local electrical distributor store today to see if they have any new+silly+expensive LED lighting, and I stumble upon a kit set of 12VAC 50W electronic transformer halogen set for approx AU$8. It can't be real... the kit consist of an 240v>12v electronic transformer that complies with electrical safety AS/NZ standards, a white two piece diecast aluminium fixture and a 50W halogen bulb. Well it was no brainer, the power supply was very small I wanted to see what's in it so I bought it and waited impatiently to knock off and get home.

I just renovated my 20 yr old home, and I have standard coil transformer for my halogen downlights. I want to see how these cheap power supply works and to my surprise, quite well. I wired it up and hooked up a kill-a-watt, with the standard 50W bulb, the kill-a-watt reports 53W. On the true rms Fluke is telling me output 12.0xxVAC which is very nice voltage regulation. Then I tried a bi-pin 20W bulb, the kill-a-watt says 23W.

Two high power TO-220 NPN for switching and a TO-92 NPN all I can see, and some caps and resistors, 7 diodes, lots of magnetics. 12V output is fully isolated via a toroid winding. I'll post some pictures of the internal tomorrow.

[sacherjj]

Wow, I'd like to see that.  I'm trying to figure out if it is feasible to use LED lighting for the new closed in porch that I'm finishing.  It would be cool to be able to vary the voltage in the switching circuitry for using a dimmer without losing tons of efficiency.

[mikeselectricstuff]

the kit consist of an 240v>12v electronic transformer that complies with electrical safety AS/NZ standards,

.. if you believe the stickers...

[Zero999]

They're soo efficient because there so simple.

EMC filter -> rectifier -> oscillator - > ferrite transformer

The power factor will be very good because the current waveform is fairly sinusoidal due to the absence of a massive reservoir capacitor after the rectifier.

I doubt the regulation will be that great: the output voltage will vary in proportion to the primary voltage.

The output waveform will be a 20kHz to 200kHz AC square wave modulated with the mains frequency after full wave rectification. Have a look at the waveform with an oscilloscope and try varying the input voltage with a variac.

[PetrosA]

You don't want regulated output in these transformers or else they wouldn't work with dimmer switches...

I just installed six 50W MR16 fixtures each with its own power supply (in the wall). They work very nicely:

[Uncle Vernon]

the kit consist of an 240v>12v electronic transformer that complies with electrical safety AS/NZ standards,

.. if you believe the stickers...

That's the beauty of our C-Tick system[sarcasm]. If it has a valid C-Tick number it complies all the onus is pushed onto the distributors. Who needs engineers testing everything when that can be replaced by a file of assumed tests diligently checked by someone skilled in import/export?

It's win/win for governments, no need to pay compliance testers and all blame can be pointed elsewhere whenever some member of the public gets fried.

So very likely it is compliant and deemed fit for purpose by the seller. Does it meet or exceed Australian Standards? That is anybodies guess.

[Neilm]

the kit consist of an 240v>12v electronic transformer that complies with electrical safety AS/NZ standards,

.. if you believe the stickers...

That's the beauty of our C-Tick system[sarcasm]. If it has a valid C-Tick number it complies all the onus is pushed onto the distributors. Who needs engineers testing everything when that can be replaced by a file of assumed tests diligently checked by someone skilled in import/export?

It's win/win for governments, no need to pay compliance testers and all blame can be pointed elsewhere whenever some member of the public gets fried.

So very likely it is compliant and deemed fit for purpose by the seller. Does it meet or exceed Australian Standards? That is anybodies guess.

All too true. 

The C-tick is similar to the CE mark required for Europe, but there are many cases of companies that do not do the required "Due diligence". For example, multimeters fall under the requirement of IEC61010. This requires a full report to be generated. The old version even gave a tick sheet and test report templates to be filled in.

The company I work for wanted to put a factored CAT III multi-meter on the market. The company who made the meter sent us an example meter and a certificate of conformity from a test house. All looked fine. However, it is company policy to run these past the product qualification people who assess the in house designed products. I was working in that department at the time and it landed on my desk.

First thing I do (as Dave says) "Take it Apart". I opened the back of the instrument and could not believe my eyes - glass fuses (which I believe is banned in IEC61010) screw heads between the legs of a transient suppressor and between the holders of the fuse. Internal wires not restrained. Cheap test leads that showed signs of wear from the travel from China. If the test house was real then they didn't see this instrument. The other samples were just as bad. Needless to say, the list of faults was sent to the salesmen who wanted to put this on the market together with a note that any other samples would require a FULL test (both safety and EMC).

The one good thing that did come out was that test house. It was added to the list of those not believed.

Neil

[nukie]

Hello all,

I think hero999 nailed it.
Here's the pics.. hopefully they are in right order the filename describes the scope readings.

50W is full load. Sorry I was using my older rigol with silly bad STN screen.

[Zero999]

The no load waveform looks odd but it's no surprise since these little transformers aren't designed to work unloaded.

The frequency is roughly inversely proportional to the load 45KHz at 20W and 26kHz at full load which makes sense as it's probably a self oscillating circuit so the frequency will vary widely depending on the load.

Try setting the the time base to 20ms/div to look at the waveform envelope.

I've often wondered if it's possible to use one of those transformers to power an electronic project.

I wonder how well the transformer would cope with a bridge rectifier and filter capacitor connected to the secondary? Probably not very well since the capacitor charging may trip the over current protection and it might not allow the current to be drawn in pulses so you'd have to add active power factor correction unit. Another option may be to power it from rectified and filtered mains so you could use a small filter capacitor but it might not like that either.

Of course you could omit the filter capacitor and use the transformer to power a load which doesn't need a smooth DC supply such as a motor or LED (an efficient capacitor ballast or inductor could be used as current limiting rather than a resistor) and have a separate low current smooth DC supply to power some control logic.

[Zero999]

I just found this application note and remembered this thread.

Going from the photograph, the components seem to match the one listed in the paper linked below - I bet it's nearly identical.
http://www.rlocman.ru/i/File/2007/05/07/EL_transf_12V.pdf