advertisement


Power consumption of transformers: standby vs. idle

N

nomad101

Member
Hi

I wondering about how transformers use power (for example in an Amplifier or DAC).

There can be the following states of a device:
  1. Off
  2. Standby
  3. Idle (no output)
  4. In Use (playing output)
For (3), the transformer can remain warm. How significant a difference is this in terms of power consumption compared to (2)? Is it a mere few watts or potentially much more? And is leaving it in (3) a bad idea concerning 'wearing out' the transformer (as eddy currents are circulating)?

Thanks
 
Good question....I was interested in this. I always wondered if it was better to leave my old LK280s in scenario 3 or 4 to warm up before use and it's always surprised me how hot they get even in 3...I assumed they were running in 'class A' mode at low/no input and the heat was coming from the power transistors, didn't occur to me it could be the toroid.
 
Long story short:

1. Vinny is entirely correct SMPS are different. Efficiencies under load and especially when unloaded, are high by design focussing on absolutely minimising losses (by running a special tiny transformer at 200-500Khz, not a big one at 50-60Hz, and clever control strategies to do this)

2. Transformers of a size used in audio equipment are efficient, and get better the bigger they are; at the 1KVA level, 1-1.5% 'loss' at full power would be about right. Still only 10-15W, and (waves hands a bit) that will stay roughly proportional down to the 100VA scale used in small amplifiers; might be 5-8% there, so - still, small wasted power.

3. The losses are a little higher proportionally at 'idle/ standby' levels - there is a necessary magnetising current, even without output power being drawn and that means current flowing in the copper; and a corresponding 'iron loss' in the core. Again these small-numbers in well designed transformers of any core type - toroidal, the most efficient (least copper resistance, excellent magnetic coupling) and lower in E-I, split-bobbin types (slightly worse on those two accounts) - and can even be deliberately made poor for other reasons like some kinds of specialised power-filtering transformers not used in audio.

Example for 2, 3 above: I've a very nice H&F 100VA toroidal power transformer rescued from a Cyrus shoebox amp, now used to power my headphone amp. The transformer, unloaded, on its own, draws under 2w connected to the mains. Idling at 18watts load - it still only runs 3-4degC above room temp, max (measured with IR thermometer.) The summary - total 'losses' are small, in modern transformers.

3. These standing losses are proportionally-worse, the smaller the transformer gets. This is why heavy transformer based wall-warts went away, replaced by SMPS types (no use having a 5v phone charger / a chunky wall wart for a small halogen desk light, where the thing wastes 4+w idling all the time it is plugged in... you'll remember how warm those things used to get )

4. There really is NO wear-out mechanism for transformers, at all! Which is nice. The only real way to damage one is a gross & sustained overload. In fact larger domestic-scale power transformers are incredibly robust, running 20% overload for say 20% or more of the time is no issue at the 300VA+ scale for toroidals of good quality - because of the inherent efficiency. But under gross overload the failure mode is terminal - the insulation will fail (stinky, give long, early warning) , a winding 'shorts' and as a result that will blow the mains fuse.

5 Note small transformers, the wall-wart scale, actually often have fusible-links in inside the windings as a protective measure - because they always run hot, as a result of inefficiency.

6. Leaving equipment powered-on though will eventually shorten the life of the reservoirs capacitor(s) - which may or may not be a problem. This is why Naim recommended replacement at 7-8yrs old for the amplifiers they recommended be left powered-up 24/7 (BTW the maths on that works perfectly, I can dig up a former post on such). Note this bit does still apply to SMPS - they have a reservoir cap, powered by a bridge rectifier directly off the mains AC input and as such that cap is somewhat exposed to any Mains AC voltage transients. Good SMPS have mitigation for that, but still - the DC link cap dying is the usual reason for an SMPS failing. Since that does not directly affect the quality of the DC output owing to the voltage-regulation on the output side, really good quality SMPS can run for decades.

Hope that wasn't too boring/patronising/wordy/hand-wavey...


eta: 28.06.24/ 6pm - typos fixed for clarity.
 
Last edited:
Vinny said:
This is especially useful -

engineering.stackexchange.com

Does a transformer use power when output isn't under load?

I was reading about how AC to DC converters work with a step-down transformer and then a diode bridge to convert the lower, stepped down AC voltage into DC. What I don't understand is since the inp...
engineering.stackexchange.com engineering.stackexchange.com
Click to expand...
That article is a poor choice. Nearly all AC to DC converters rectify the mains and chop it at high frequencies through a small transformer. They often can be shut down, so the stand-bypower consumption is tiny.
 
davidsrsb said:
That article is a poor choice. Nearly all AC to DC converters rectify the mains and chop it at high frequencies through a small transformer. They often can be shut down, so the stand-bypower consumption is tiny.
Click to expand...

The OP asked about transfrmers, not SMPS, which I also mention as being different.
In the great majority of audio kit, the gas-guzzlers at any stage of operation, will be transformers.
 
Vinny said:
The OP asked about transfrmers, not SMPS, which I also mention as being different.
In the great majority of audio kit, the gas-guzzlers at any stage of operation, will be transformers.
Click to expand...
The transformers in a conventional power supply don't use a lot of "real" power when idling or they would get very hot, unless they are not rated for the actual supply voltage.
 
Look up "non watty" component and Faraday.
The best SMPSU are bridge resonant type and you can get them to 98% eff. Like the ones on the space station.
Flyback is the most common and they are not so good noise and bad on no load .
Toridal TX produce and the wire exit magnet fields a pain on pre amps.
I normally add 20% extra copper wire .
Sorry vision still crap
 
Hi Colin,

Would you explain the purpose of using additional copper / more windings, over bare-minimum for function please ?
It might be of interest to others in future.

Best wishes
 
Thanks for the responses, esp @martin clark for the insightful post!

My question was prompted by noticing that my integrated amp doesn't have a standby mode, and when left idling appears to run cool - I can't feel any warmth from the transformer location which is beefy toroidal. In comparison, my dac, which also has a relatively large toroidal for the size of the dac runs warm when idling; though its box is very compact and no vents at the top, unlike the amp which has a larger space inside and vented.
 
Your integrated is likely running class AB - and will be designed to deal with the heat developed at max output - it should run cool!
Even then: almost all that dissipation , at idle - is a result of the amplifier circuit: not the transformer (back of envelope guess - by about 8-10:1)

Modern dacs, using a modern /' low -profile transformer, have to trade two or three or even more things, off:
  • Compact size in a compact case
  • Use of convenient , UL-rated flatpack split-bobbin types, like the excellent 'Block' transformers (- which have good advantage in terms of being split-bobbin / lower susceptibility to Mains HF noise)
  • yet that means also, are not quite as efficient as can be. but this is hair -splitting.
  • internal voltage regulation after the raw dc supplies are established. These, stable & reliable across the various markets [the product] is sold-into, likley-dominate. With a ' linear supply' the designer merely to cover the EU might need to ensure it will work from, say 210- 260vAc input, reliably: design for that low end, means .. quite some waste heat, if the mains is higher : in a market like the UK; and so on.

I suggest - don't worry about it: in a small unventilated case, powering a fair amount of current-draw at low voltage, where perhaps much of the heat will be from the necessary 'drop' to run many little voltage regulators... well, to feel such a small package run warm is not remotely unusual.

tl;dr:
its not the transformer alone, but many other aspects. Plug -in , enjoy!
 
martin clark said:
Long story short:

1. Vinny is entirely correct SMPS are different. Efficiencies under load and especially when unloaded, are high by design focussing on absolutely minimising losses (by running a special tiny transformer at 200-500Khz, not a big one at 50-60Hz, and clever control strategies to do this)

2. Transformers of a size used in audio equipment are efficient, and get better the bigger they are; at the 1KVA level, 1-1.5% 'loss' at full power would be about right. Still only 10-15W, and (waves hands a bit) that will stay roughly proportional down to the 100VA scale used in small amplifiers; might be 5-8% there, so - still, small wasted power.

3. The losses are a little higher proportionally at 'idle/ standby' levels - there is a necessary magnetising current, even without output power being drawn and that means current flowing in the copper; and a corresponding 'iron loss' in the core. Again these small-numbers in well designed transformers of any core type - toroidal, the most efficient (least copper resistance, excellent magnetic coupling) and lower in E-I, split-bobbin types (slightly worse on those two accounts) - and can even be deliberately made poor for other reasons like some kinds of specialised power-filtering transformers not used in audio.

Example for 2, 3 above: I've a very nice H&F 100VA toroidal power transformer rescued from a Cyrus shoebox amp, now used to power my headphone amp. The transformer, unloaded, on its own, draws under 2w connected to the mains. Idling at 18watts load - it still only runs 3-4degC above room temp, max (measured with IR thermometer.) The summary - total 'losses' are small, in modern transformers.

3. These standing losses are proportionally-worse, the smaller the transformer gets. This is why heavy transformer based wall-warts went away, replaced by SMPS types (no use having a 5v phone charger / a chunky wall wart for a small halogen desk light, where the thing wastes 4+w idling all the time it is plugged in... you'll remember how warm those things used to get )

4. There really is NO wear-out mechanism for transformers, at all! Which is nice. The only real way to damage one is a gross & sustained overload. In fact larger domestic-scale power transformers are incredibly robust, running 20% overload for say 20% or more of the time is no issue at the 300VA+ scale for toroidals of good quality - because of the inherent efficiency. But under gross overload the failure mode is terminal - the insulation will fail (stinky, give long, early warning) , a winding 'shorts' and as a result that will blow the mains fuse.

5 Note small transformers, the wall-wart scale, actually often have fusible-links in inside the windings as a protective measure - because they always run hot, as a result of inefficiency.

6. Leaving equipment powered-on though will eventually shorten the life of the reservoirs capacitor(s) - which may or may not be a problem. This is why Naim recommended replacement at 7-8yrs old for the amplifiers they recommended be left powered-up 24/7 (BTW the maths on that works perfectly, I can dig up a former post on such). Note this bit does still apply to SMPS - they have a reservoir cap, powered by a bridge rectifier directly off the mains AC input and as such that cap is somewhat exposed to any Mains AC voltage transients. Good SMPS have mitigation for that, but still - the DC link cap dying is the usual reason for an SMPS failing. Since that does not directly affect the quality of the DC output owing to the voltage-regulation on the output side, really good quality SMPS can run for decades.

Hope that wasn't too boring/patronising/wordy/hand-wavey...


eta: 28.06.24/ 6pm - typos fixed for clarity.
Click to expand...
It’s not necessarily the transformer itself that is inefficient (in terms of waste heat) in a linear power supply but the regulation. Linear regulators give off a lot of waste heat, switching supplies use pulse width modulation to regulate the voltage supplied from the secondary windings of the transformer, which is technically a lot more efficient, the devices doing the regulation are on or off rather like linear regulators.
 
Hi Martin.
Adding 20% Cu.
We did this at Inca Tech, the TX were made then by new company run by a dear friend Peter Harris in Bishopstort Ford tha company is Airlink.. oh part no for Claymore TX is 918.
We had problem in parts of UK with audio buzzing from the mechanical parts of the transformer. Due to lam and wire movement.
Just a note if wire is moving it is a waste of energy.
The buzzing is often due to magnetic saturation due to high primary volts but most often DC or flat topping sin wave.

The big chunky test Gig I made for Naim listen and check for harmonics.
Mains for 50Hz has 20mS cycle so if the input is overvolted and ever 10mS peak of We see flat topping thus core saturation we hear a 100Hz hum. DC one sided saturation we see just clipping on one peak we hear 50Hz
Saturation simply is too few turns and losing magnetic property of core.
Transformers at very high frequency and magnetic core but air or vacuum do not saturate. But a 50Hz version and at a few 100VA would be the size of a house.
For study look up B/H curve .
Tesla and Maxwell and not forget my hero Faraday. Forget Bell and Marconi money greedy humans, and there are still to many alive.

Got to stop now eyes gone bonkers look up "Charles Bonnet Syndrome " oh what fun.
Seen Dino's in living room and Red Busses in garden. But word jumping is odd.

Best Col
 
@CJ14 Great explanation - thankyou. I thought that was the reason, but then transformer action is in some ways counter-intuitive.
Add more turns over the minimum at first glance would seem to add more induction and demand more flux in the core- yet , the way transformers actually work, the better coupling actually reduces the flux in the core. Faraday and Lenz' laws at work.

ATB with your vision; what an utterly-disruptive condition to live with.

Thanks -
Martin
 
Amps per turn.
Vision with new specs work well can see better thru the 1inch slot like a fish eye lens wierd and spooky. But made a amp for Dave of Audiochews today and it worked yippee.

Best Col
 
You must log in or register to reply here.


advertisement


Back
Top