I have been experimenting with LED lighting to compare to CFL and found that although my 9w fluorescent draws 18.5w at the plug, a 10w LED module uses 23w with all the adjustments I can make.

I'm using a 120/24v transformer into a FW bridge and 470uf cap giving 34-35v unloaded. I then use an LM317 in constant current wiring to provide approx 28.4 @ 332ma. this loading drops the rectified supply to about 32.3v (the first one I got required 38v to drive at 330ma! Definitely a factory second.) At this combination the LM317 doesn't get warm and the transformer gets only marginally above ambient, but when I disconnect the load the draw is still around 12.3-13w. .

If I get a different transformer(the existing has a 3a sec.), maybe a 500ma rated one, would it be more efficient?

 

If you're concerned with efficiency, be sure you are checking watts and not volt amps.

 

Unfortunately very small transformers naturally have poor efficiency. There is no way around it.

At 100VA they are around 90 - 95% efficient but at 10 VA you are lucky to see 50 - 60% efficiency.
Below that the efficiency drops of even faster which is why micro SMPS are used in most small power packs now.
Even at 80% average efficiency in the 5 - 10 watt range SMPS are still two - five times more energy efficient than their transformer equivalent.

 

If I get a different transformer(the existing has a 3a sec.), maybe a 500ma rated one, would it be more efficient?

I think that it would. But be careful about how you measure efficiency. As suggested above, some of the current you measure is reactive, and so the phase must be considered in the power calculation.

 

Unfortunately very small transformers naturally have poor efficiency. There is no way around it.

At 100VA they are around 90 - 95% efficient but at 10 VA you are lucky to see 50 - 60% efficiency.
Below that the efficiency drops of even faster which is why micro SMPS are used in most small power packs now.
Even at 80% average efficiency in the 5 - 10 watt range SMPS are still two - five times more energy efficient than their transformer equivalent.

I didn't know small transformers were that low.
Thought maybe it was just that they were poor quality.

So I checked a good quality industrial 50va transformer.
It's only 85% at a little over half load.

Which backs up your figures.


Op, are you sure it's watts?

 

I didn't know small transformers were that low.
Thought maybe it was just that they were poor quality.

Depending on how they are made and quality of materials used once you go below the ~100 VA range working efficiency drops off pretty bad.

The tiny 2 - 3 VA units used for early LED alarm clock were lucky to see 25 - 35% at their working power levels.

What's interesting is that once you get large enough efficiency gets very close to 100%! Massive 100's of MVA units will see numbers in the mid 99% range!

Not so relevant here on the large stuff but when dealing with small sub to few watt loads line frequency transformer based power supplies tend to get pretty inefficient. Especially if it's a low voltage power supply.

 

Tiny transformers are difficult to make, too. Here are some primary winding numbers for 115V, 60 Hz. on grain oriented silicon steel EI cores.

300 watt core: 107 turns of #13 wire
30 watt core: 350 turns of #20 wire
5 watt core: 875 turns of #32 wire

Don't ask me to prove these numbers. I didn't design them, I just copied the data from work.

 

How are you measuring the mains power?

 

I'm using a Kill'a'Watt to measure the apparent power. I can't attest to the accuracy, but the manual states .02%. It does have several units of measure including PF and an accumulator.

 

the transformer gets only marginally above ambient, but when I disconnect the load the draw is still around 12.3-13w. .

I'm using a Kill'a'Watt to measure the apparent power.

Is the "around 12.3-13w" apparent power or true power?

 

Re inefficiency of small transformers: Note that the little ones are usually designed not to start a fire if the secondary is shorted, called Class 2 transformers. Here is a description:

Class II transformers have a maximum VA (Volt-Ampere) rating of less than 100 and a maximum secondary output of 30 VAC. The maximum VA generally offered is 75 and the most common secondary voltage is 24 VAC. All Class II transformers are either inherently or non-inherently limited. This means that the maximum output current of the transformer is limited, either by the intrinsic coil impedance or by a fuse or circuit breaker.

Likely, this contributes to their apperent inefficiency.

 

I'm using apparent power because both the ballasted CFL and the transformer input to the power supply are reactive and I don't know how well the wattmeter corrects for PF. The transformer, FWIW is a Triad F-61U dry rectifier type and I thought it would be high quality.

 

Re inefficiency of small transformers...

Thanks. I searched half the night for clues.

 

I'm using apparent power because both the ballasted CFL and the transformer input to the power supply are reactive and I don't know how well the wattmeter corrects for PF. The transformer, FWIW is a Triad F-61U dry rectifier type and I thought it would be high quality.

I've found the Kill'a'Watt to be very accurate.

When comparing efficiency I can't think of a reason not to compare watts.
The utility charges for watts.

The watts in to watts out will already take into account the I squared R losses.

The VA reading is important for transformer capacity.

 

I'm using apparent power

Apparent power has almost nothing to do with efficiency. Apparent power is not what the power company meter measures on a residence. You have to use real power to get any useful numbers.

 

OK, so I am using A Triad-Utrad F-61U which is rated for 100va(?) 24-26-30-33-36v@3a.

Is there any other option so I lose less in the supply? My experience with the inexpensive products from overseas are that they are Minimally adequate/not dependable. I saw a preamp board that had caps rated at 1/2 the actual working voltage and the vendor didn't care when it was pointed out.

 

You might look into torroid transformers. They are somewhat more efficient. Or you might try one of the small, inexpensive switchers in ebay. You thread has me thinking about trying to come up with something to use, if I can find the time to work on it.

 

OK, so I am using A Triad-Utrad F-61U which is rated for 100va(?) 24-26-30-33-36v@3a.

Is there any other option so I lose less in the supply? My experience with the inexpensive products from overseas are that they are Minimally adequate/not dependable. I saw a preamp board that had caps rated at 1/2 the actual working voltage and the vendor didn't care when it was pointed out.

What is the power use with no load?

Not apparent power, but the power actually being consumed.

 

According to the Kill'a'Watt it draws 13w without load attached and 23w with load of 9.43w from LED.

 

<snip>

Downside it line driven LED circuits are not discussed here.

Moderator Comment:

Please do not violate the ToS rules here. It is obvious from the above comment you knew this was a violation. This set was created for clear cut safety reasons, which we at AAC take very seriously.

A moderated copy of the comment has been saved around post #25 for moderator review.