Linear vs Switchmode supply

Thread Starter

Art

Joined Sep 10, 2007
806
Hi Guys,
I was debating with a friend about linear supply vs switched supply in products.
It is my position that the only reason to use a switchmode supply in any product is for cost effectiveness.
Weight of the supply could be included in cost effectiveness for shipping, etc.

It also occurs to me no audio equipment has switched supplies,
and for a linear supply you have a pi filter, and only have to filter the mains frequency.
I understand a switchmode supply can have the money spent on it, and have good filtering,
but that was all for noise the supply made anyway.

Is all this correct, and is there any benefit at all of a product having a switched supply?
 

Papabravo

Joined Feb 24, 2006
21,159
Amateur Radio transceivers are very commonly run from Switching Power supplies. The key design choice is the switching frequency, and they can very easily avoid the passband of the audio amplifier, and the IF amplifiers. This doesn't even include the magic of DSP which is just amazing in what it can do to pull in weak DX stations from the noise floor at -130 dBm.
 

AnalogKid

Joined Aug 1, 2013
10,989
Is all this correct, and is there any benefit at all of a product having a switched supply?
No, none of it is correct. Everything you said is completely or mostly incorrect, and significantly inexperienced.

Many audio products have switching supplies, including almost all high power automotive products. Because it is easy for switchers to operate above 100 kHz, it also is easy to keep heterodyne products and image frequencies out of the audio passband. Most people with normal hearing cannot distinguish the difference, even in a controlled listening environment.

Another reason you missed is size, independent of weight. A 1 kW 60 Hz power transformer not only weighs tens of pounds, but is almost twice the size of a softball. An equivalent switcher transformer is the size of a golf ball. Vicor makes a DC/DC converter that achieves 2,870 watts per cubic inch for the power train and magnetics combined. In round numbers it is the size of a 5-pack of chewing gum and weighs 1.5 ounces.

As blocco said, efficiency is a major factor in deciding to go with a switcher. A linear power supply that makes 400 W into the load also makes about 400 W into the air. An equivalent offline switcher can waste fewer than 40 W.

Another significant feature of switchers is universal AC input. From Japan to England, AC line voltage ranges fro 100 V to 240 V rms. To handle this without jumpers or switches, a linear supply needs either a multi-tap power transformer, a monitor and control circuit, and a high voltage high current AC switch, or a really really big fan. A switcher, particularly one with power factor correction, adjusts to anything from 85 to 250 VAC automatically, with very little change in efficiency.

And speaking of power factor and harmonic current correction, every year or two the minimum power level that requires PFC decreases. This has spelled doom for most high wattage linear supplies because passive PFC takes very large components because the frequency is so low. Anything with a bridge and filter cap on the input is horrible, but for switchers there's the old Zytec PFC patent (James Bucher, 1987), the basis for most PFC supply front ends today.

The transformer is a marvelous thing, and can be an efficient component when applied properly. But the vast majority of the time, low frequency linear electric power conversion, especially AC-to-DC, is inferior. Advantage switchers.

ak
 

MikeML

Joined Oct 2, 2009
5,444
... But the vast majority of the time, low frequency linear electric power conversion, especially AC-to-DC, is inferior. Advantage switchers...
With one important exception: RFI! The HF radio spectrum has been destroyed by the proliferation of switching power supplies.
 

ErnieM

Joined Apr 24, 2011
8,377
Once we get past the power supply to an audio amplifier we get to switched amplifiers such as the class D type where the audio signal is generated in a similar manner as a switching power supply.
 

#12

Joined Nov 30, 2010
18,224
Once upon a time, I was building an abomination of a variation of a Fender Champ amplifier and I had a 5 volt switchmode supply laying around...so I used it for the filament of the rectifier. Yeah, it's floating near 400 VDC, so I mounted it on the wood case to avoid high voltage danger. The vacuum tube nut-case of a customer never suspected it was in there.:D
 

Thread Starter

Art

Joined Sep 10, 2007
806
I claim to hear the difference in real instruments (piano, brass, wind, etc.) through a valve amp or valve domestic AM radio under heaps of static regardless of the source.

I did ask the question “is there any benefit at all” and in light of what has come up that is a bit close minded.
It is true though, the question is in the practical context of noise, you see valve amps and radios don’t have solid state PSUs either.
My mind should not be stuck in that context perhaps. There are appliances where noise at least couldn’t hurt the appliance.

This doesn't even include the magic of DSP which is just amazing in what it can do to pull in weak DX stations from the noise floor at -130 dBm.
What if there is noise in the air on the RF frequency you made, and you powered the radio or audio amp with that supply?

With one important exception: RFI! The HF radio spectrum has been destroyed by the proliferation of switching power supplies.
I don’t remember otherwise when HF was 64bit digital audio, but thousands of dirty little oscillators in plastic cases working together I wouldn’t be surprised what they could do.
 

Papabravo

Joined Feb 24, 2006
21,159
...
What if there is noise in the air on the RF frequency you made, and you powered the radio or audio amp with that supply?
...
I've been a ham for 11 years and have simply not encountered any trouble from the switching power supplies that I have used in my own shack, or at multi-transmitter events like Field Day. I chase very weak DX signals all the time and if there were noticeable interference from the power supplies this is when it would happen. There are other sources of noise and interference that happen all the time but you just have to live with those.
 

AnalogKid

Joined Aug 1, 2013
10,989
It's not like you can rebut people who can "hear" the difference between the same recording played back on a CD vs vinyl.
Easy with those quotes, sonny (I assume everyone is younger than me; it's a statistically reliable position). Just about anyone can hear that difference, even though much of the time it is because the CD has been remastered, noise cored, and usually compressed compared to the original vinyl. Granted there are the zealots who think oxygen-free copper is a gift from space aliens, but here in the real world the two media have significantly different native acoustic characteristics so it isn't a good example. As for hearing the difference in power supplies only, such as two phono preamps, one with a linear ps and one with a switcher (extreme example), the only reason you can measure and hear any sonic artifacts is because the ps designer did an incomplete job. I've designed power systems for TEMPEST rated equipment. It's not magic or secret, just a matter of effort and attention to detail. I contend that there is nothing inherent in switching power supplies that make them inescapably noticeable in audio systems.

ak
 

recklessrog

Joined May 23, 2013
985
Being an amateur radio ham, I just seem to have a natural dislike of switch mode power supplies, although I have a considerable amount of equipment that uses them. My Phillips PM3394 oscilloscopes and other other test instruments use them so I can not really rationalize my feelings towards them.
So much so, that back in the 1980's, I bought a Yeasu FT one HF transceiver, which I discovered had a SMP. This did cause some interference at odd frequencies which Muphy's law mostly happened to be the one i was operating on at the time. A year or two later, I bought an older version where the owner had blown up the o/p trying to get more power. This one had a linear power supply with a lovely heavy mains transformer. It was a simple task to swap it into my original set, and all the annoying internally generated emf was gone! I was very happy! (probably psychologicaly also)
The added bonus was that I repaired the second set, fitted the smp and sold it for a tidy sum!!!
I now have a Yeasu FT950 with DSP which is great, but unfortunately, with the way operators work stations nowadays, for me the thrill is gone.
Back to analogue vs smp. Logically, it makes sense also from an operating cost (secondary user cost) perspective to use smp's where possible, Think of powering modern T.V's with mains transformers. As pointed out in another post, not only would it make the set heavier increasing shipping costs, it would require tapped windings with voltage selectors, and be less efficient costing more to run. So even though something at the back of my mind still dislikes them, they seem to be the way to go for most things. (If only those darned nasty RF emmiters on sale from auction sites and market stalls where all crushed!!)
 

AnalogKid

Joined Aug 1, 2013
10,989
No. In the context of a thread bemoaning switching power supplies' noise pollution of HF radio bands, I think a switching radio transmitter output stage is worth noting, especially to those who are not familiar with this kind of crossover circuit.

ak
 

alfacliff

Joined Dec 13, 2013
2,458
which power supply, switching or linear depends on use, and whether it is a large production device or not. for a one off home project, it is pretty easy to put together a linear supply, much easier than to design a switching supply. a high power automotive aplication requiring higher voltage would have the advantage to the switcher, since a linear supply stating with low voltage dc and raising it to higher voltage would be rather silly. both supplies have their advantages and disadvantages.
 

ian field

Joined Oct 27, 2012
6,536
Hi Guys,
I was debating with a friend about linear supply vs switched supply in products.
It is my position that the only reason to use a switchmode supply in any product is for cost effectiveness.
Weight of the supply could be included in cost effectiveness for shipping, etc.

It also occurs to me no audio equipment has switched supplies,
I've seen several audio amplifiers with SMPSU, and Elektor magazine published a project to build one a few years back.

But as yet, most manufacturers don't want to mess with the screening and filtering to keep the switching transients out of the audio path.

There's also the difficulty of designing a SMPSU regulation control with good transient response - its just so much easier to use a dirty great toroidal mains transformer and design the amplifier to tolerate mains variations.
 

ramancini8

Joined Jul 18, 2012
473
Back in the sixties when switching power supplies first came on the scene the switching frequency was low (1 to 20 KHz) and RFI/EMI control was minimal. That is when switchers, deservedly, got their bad reputation. Now with switching frequencies increasing to 100 KHz and higher, coupled with excellent noise control, switchers don't deserve the bad reputation for being noise generators. But, like a wayward girl, reputations follow more than performance.
 
Top