Author Topic: When are SMPS not used these days?  (Read 3144 times)

0 Members and 1 Guest are viewing this topic.

Offline lordvader88Topic starter

  • Frequent Contributor
  • **
  • !
  • Posts: 935
  • Country: ca
When are SMPS not used these days?
« on: June 23, 2019, 05:27:23 pm »
I see they are used in oscilloscopes, even my 1980s analog scope. How is it actually cleaned up enough, to not interfere with something as sensitive as a scope ? Or do they use circuits that can cancel it out or ignore it ?

And digital stuff like computers use them all the time. What types of stuff, besides maybe tube amp's, still use main's transformers ?
 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: When are SMPS not used these days?
« Reply #1 on: June 23, 2019, 05:38:18 pm »
High resolution bench DMM.
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 5170
  • Country: ro
  • .
Re: When are SMPS not used these days?
« Reply #2 on: June 23, 2019, 05:44:07 pm »
My nightstand lamp uses a 20VA transformer (has one 12v halogen bulb).
The transformer actually has just the right amount of weight to prevent the lamp from falling on its sides or shaking.

You may use them in devices where noise can affect it (radios, transmitters etc) , or maybe in devices that use negative and positive voltages (small audio amplifiers, small boomboxes etc)
« Last Edit: June 23, 2019, 05:46:05 pm by mariush »
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: When are SMPS not used these days?
« Reply #3 on: June 23, 2019, 08:23:53 pm »
I see they are used in oscilloscopes, even my 1980s analog scope. How is it actually cleaned up enough, to not interfere with something as sensitive as a scope ? Or do they use circuits that can cancel it out or ignore it ?

You have two slightly wrong assumptions here:
1) Oscillosscopes being extra sensitive. Most are not; think about the fact that most digital scopes digitize at only 8 bits! The timebase accuracy is more important, most scopes more or less suck at measuring low-level noise, or getting very accurate voltage readings. There are exceptions on the high-end market, of course.

2) Switch-mode power supplies being noisy. They can be designed to be fairly clean. A properly designed one can be surprisingly clean even without filtering, e.g., by using as high frequency as possible, reducing edge rate a bit at the cost of efficiency, using RC snubbers, keeping layout good and switching loop area minimal, with ground plane shielding all around it, and good low-ESR, low-ESL capacitors at the input and output (so that the ripple currents actually can go through these local capacitors and won't take a longer loop over the "load" side capacitors!)

Adding LC filtering to kill the remaining ripple current (fairly low frequency differential mode noise), or ferrites for really high frequency common mode noise, can reduce the noise further.

If you see a linear supply in a modern device, it's most likely for legacy reasons, such as a design reuse. It's also possible that they really need some very low-noise operation and designing such low noise SMPS with enough post-filtration would end up being too complex or expensive.

They may also have design culture around simplicity and robustness, and while an SMPS can be made very robust, long-life and low noise, it requires more understanding to do so, and more trust on your SMPS engineers that they are capable of pulling that off.
« Last Edit: June 23, 2019, 08:25:51 pm by Siwastaja »
 
The following users thanked this post: lordvader88

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: When are SMPS not used these days?
« Reply #4 on: June 23, 2019, 08:37:25 pm »
At some point they realized that for high precision work, you need to have local regulation anyway, just where the load is. These typically do not handle large currents, because each site uses their own regulators, the burden is shared. They want to use local regulation for DC accuracy, as well.

At the same time, something else happened as well: now there are linear regulator ICs with really excellent Power Supply Rejection Ratios, even at high frequencies, and they aren't even expensive. Compare to older regulators, if you try to post-filter your noisy SMPS with a LM317 or 7805, it won't do shit after a few dozen kHz, just passing it through. But some modern regulators regulate at MHz or tens of MHz ranges just fine, which is the territory where you'd otherwise need some real inductors and capacitors, which then again could have their own resonances where they are ineffective or even make the situation worse, unless you dampen them which again makes them less effective over a wider range.

But now if the linear regulator is capable of taking care up to some MHz, you are only left with really high-frequency noise which is dealt with ferrites, very small SMD caps, proper layout, shielding etc. Suddenly, an SMPS isn't so bad after all.

They need to deal with noise, anyway: it reflects back from all the other SMPS's, and this may include some very crappy ones, for example, old ones that passed less stringent regulations back then, or some illegally performing imports. If you just use a mains transformer (which has some coupling capacitance from input to output) and a 78xx style linear stage, assume it's guaranteed to be very low noise, and call it a day, you may fail due to external noise getting in through power line.

So, say, your 10x better than average SMPS design inside your low-noise product may not produce more noise than what's already getting in because of a 10x worse than average PSU in the neighbor's power outlet! Which brings us to filtration, and local regulation and local shielding.
« Last Edit: June 23, 2019, 08:40:07 pm by Siwastaja »
 
The following users thanked this post: tooki

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8176
  • Country: us
Re: When are SMPS not used these days?
« Reply #5 on: June 24, 2019, 04:22:08 am »
I see they are used in oscilloscopes, even my 1980s analog scope. How is it actually cleaned up enough, to not interfere with something as sensitive as a scope ? Or do they use circuits that can cancel it out or ignore it ?

And digital stuff like computers use them all the time. What types of stuff, besides maybe tube amp's, still use main's transformers ?

A lot of consumer audio equipment still uses mains transformers and linear power supplies.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 4034
  • Country: us
Re: When are SMPS not used these days?
« Reply #6 on: June 24, 2019, 02:25:28 pm »
Linear supplies are still common in bench supplies.  It is just a lot easier to design a linear power supply that is stable over a huge range if voltage, current and output load while providing good transient response and low noise.  Bench supplies also usually want to do all that with low output capacitance to have good current limiting.  You could do all that in a switching supply (and people have) but a linear regulator is way easier. 
 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 13157
  • Country: ch
Re: When are SMPS not used these days?
« Reply #7 on: June 24, 2019, 03:42:46 pm »
I see they are used in oscilloscopes, even my 1980s analog scope. How is it actually cleaned up enough, to not interfere with something as sensitive as a scope ? Or do they use circuits that can cancel it out or ignore it ?

And digital stuff like computers use them all the time. What types of stuff, besides maybe tube amp's, still use main's transformers ?

A lot of consumer audio equipment still uses mains transformers and linear power supplies.
Yep. Not including sound bars and home-theater-in-a-box style devices, most home cinema receivers and stereo amplifiers still use linear power supplies.
 

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: When are SMPS not used these days?
« Reply #8 on: June 24, 2019, 04:26:00 pm »
That is largely because a linear supply can have a peak rating way higher then its average one (Low frequency transformers have LOTS of thermal mass) and audio has a high crest factor. This lets you size the linear supply for an audio application way below the peak power demand.
Unfortunately the time constants that matter in a switcher are very much shorter so you wind up having to size the expensive parts for something much closer to worst case.

One nice trick for doing precision things with switched mode power is to lock the switching frequency to some harmonic of the audio clock rate, that way the switching products all alias down to DC and are trivial to filter out. You also see this trick in precision volt meters where you sometimes see different crystals for 50 Vs 60Hz markets to put the null in the right place.

Regards, Dan.
 

Offline Dave

  • Super Contributor
  • ***
  • Posts: 1356
  • Country: si
  • I like to measure things.
Re: When are SMPS not used these days?
« Reply #9 on: June 24, 2019, 06:15:37 pm »
maybe in devices that use negative and positive voltages (small audio amplifiers, small boomboxes etc)
Bipolar supplies are trivial to produce with isolated switching topologies.
<fellbuendel> it's arduino, you're not supposed to know anything about what you're doing
<fellbuendel> if you knew, you wouldn't be using it
 

Offline rstofer

  • Super Contributor
  • ***
  • Posts: 9964
  • Country: us
Re: When are SMPS not used these days?
« Reply #10 on: June 24, 2019, 07:49:57 pm »
maybe in devices that use negative and positive voltages (small audio amplifiers, small boomboxes etc)
Bipolar supplies are trivial to produce with isolated switching topologies.

They're pretty easy with 78xx and 79xx devices also.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: When are SMPS not used these days?
« Reply #11 on: June 24, 2019, 08:43:50 pm »
I see they are used in oscilloscopes, even my 1980s analog scope. How is it actually cleaned up enough, to not interfere with something as sensitive as a scope ? Or do they use circuits that can cancel it out or ignore it ?

You have two slightly wrong assumptions here:

1) Oscillosscopes being extra sensitive. Most are not; think about the fact that most digital scopes digitize at only 8 bits! The timebase accuracy is more important, most scopes more or less suck at measuring low-level noise, or getting very accurate voltage readings. There are exceptions on the high-end market, of course.

Oscilloscopes and some other test instruments are sensitive and take special precautions when switching power supplies are used.  Careful attention is paid to magnetically shielding the CRT of an analog oscilloscope and to preventing ground loops.  In addition, unconventional switching topologies were often used in the past.

Tektronix had a steady progression of series resonate and push-pull soft switching and current driven topologies from the 70s through the 90s.  The sensitive parts of the differential signal paths included plenty of common mode isolation.  Remove the shield from around the CRT and the display will be rendered completely useless; Tektronix did not use mu-metal shields without reason.

An unusual example of the attention Tektronix paid to these issues is shown below.  What is the extra transformer winding and C908 doing?  Apparently, they introduce a counter current between two points in the chassis to cancel EMI.  Also notice that there is no secondary side regulation but there are secondary side LC filters.  Feedback for regulation only comes from the primary side but later they switched to feedback from the secondary side and added linear post regulation.



Modern analog instruments like digital voltmeters and function generators often use linear power supplies.  I suspect this is because of excessive common mode noise and poor common mode isolation in commonly available regulators.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: When are SMPS not used these days?
« Reply #12 on: June 25, 2019, 06:59:29 am »
A lot of consumer audio equipment still uses mains transformers and linear power supplies.

Yes, because audio amps don't need voltage regulation. Regulated linear supply is highly inefficient, heatsinking cost alone is considerable. In an audio amp, you don't need that, it's just the transformer, rectifier and filter caps. Simple, highly efficient. More cost in copper and iron, but in the big picture, not too much difference.

Another reason is that audio equipment has mostly stayed the same or very similar, and made by the same companies, using their legacy designs, because why not, they work and have a good market case. (Of course there are totally new audio product segments, like wireless bluetooth speakers, but they do not indeed use linear power supplies.)
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20360
  • Country: gb
  • 0999
Re: When are SMPS not used these days?
« Reply #13 on: June 25, 2019, 10:20:14 am »
A lot of consumer audio equipment still uses mains transformers and linear power supplies.

Yes, because audio amps don't need voltage regulation. Regulated linear supply is highly inefficient, heatsinking cost alone is considerable. In an audio amp, you don't need that, it's just the transformer, rectifier and filter caps. Simple, highly efficient. More cost in copper and iron, but in the big picture, not too much difference.

Another reason is that audio equipment has mostly stayed the same or very similar, and made by the same companies, using their legacy designs, because why not, they work and have a good market case. (Of course there are totally new audio product segments, like wireless bluetooth speakers, but they do not indeed use linear power supplies.)
Strictly speaking, a transformer with a rectifier and capacitor is not a linear power supply. A linear power supply has a pass device acting a resistor to dissipate the difference between the input and output voltage across it.
 

Online tszaboo

  • Super Contributor
  • ***
  • Posts: 8218
  • Country: nl
  • Current job: ATEX product design
Re: When are SMPS not used these days?
« Reply #14 on: June 25, 2019, 10:27:51 am »
I had sensitive analog circuits (few ppm accuracy, few KHz bandwidth)  running from switching power supply. Once you know your supply, you can filter out the noise and that's it. You can always exchange a little bit of efficiency for precision analog, so so there is no real reason to use it. In fact filtering 100KHz and harmonics is a lot simpler than 50Hz.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: When are SMPS not used these days?
« Reply #15 on: June 25, 2019, 03:21:56 pm »
some modern regulators regulate at MHz or tens of MHz ranges just fine

Do you have a few part numbers as example?
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 4004
  • Country: nl
Re: When are SMPS not used these days?
« Reply #16 on: June 25, 2019, 04:48:56 pm »
MP1584 runs at 1.5MHz and indeed it is not an exceptional high frequency.
LT8650 runs at 2MHz.

2 unrelated links about "silent switcher".
https://www.mouser.com/pdfdocs/ACTSilentSwitcher.pdf
https://linearaudio.nl/silentswitcher

The first link is from a presentation from Analog Devices.
It combines extremely good PCB layout with on chip features that make the loop area (and thus stray magnetic fields) of the "hot" components very small. Syncrhonous switching also eliminates external diodes, which eliminate ringing and again, also results in still smaller loop area.

Also:
A long time ago I had an old 400MHz analog scope, with a big bundle of twinax as a delay line wound around the CRT.
It was quite a beast.

It had 2 SMPS boards moulded in some half hard epoxy / rubbery compound, and when they failed I replaced them with 2 toroidal transformers.
After that the traces on the CRT were all wobbly because of the stray magnetic field of those transformers. Even though the transformers were in a separately shielded compartment ( More than 1mm thick aluminum between the transformers and the CRT's).
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 4034
  • Country: us
Re: When are SMPS not used these days?
« Reply #17 on: June 25, 2019, 06:10:31 pm »
After that the traces on the CRT were all wobbly because of the stray magnetic field of those transformers. Even though the transformers were in a separately shielded compartment ( More than 1mm thick aluminum between the transformers and the CRT's).

Of course.  Aluminum hardly shields low frequency magnetic fields at all.  You would need a few cm thick plate to do much of anything.  Or: a thin layer of steel or mu-metal or both.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: When are SMPS not used these days?
« Reply #18 on: June 25, 2019, 10:54:49 pm »
MP1584 runs at 1.5MHz and indeed it is not an exceptional high frequency.
LT8650 runs at 2MHz.

Oh, I thought we are talking about linear regulators with bandwidth beyond megaherz... I'm not sure that high frequency helps. I'd say the opposite, as it requires sharper edges to achieve same efficiency. Although, smaller frequency would require bigger components, which in turn creates more parasitics....

Frankly, I'm yet to build a switcher myself, but I saw other people doing it :). Harmonics can go well into 500MHz+, filtering such frequencies is difficult. Even probing is a challenge, a typical scope with passive probes cannot do that. So, I'm a bit skeptical hearing that "filtering is not a problem", etc. I don't buy this :).

I think the reason smps everywhere, including point-of-load, is that most circuits are not that sensitive to broadband noise. They also often use post-regulator. But post-regulators have just so much bandwidth, much less than most people think. I'd say even more, most regulators don't do anything beyond a few hundreds of kHz, all the high-freq filtering/PSRR is solely due to input/output caps... Except, may be some special rf regulators. Although, I just picked one I have, rt9193, and PSRR is not so good.

Anyway, please post part numbers/circuits with low noise, I'm happy to dive deeper into the subject.

PS as of noise from a toroidal tranny, I don't think it's an issue. I can imagine why CRT is sensitive to it, but most circuits are not. Even if there is induced noise, it's just 50/60Hz. My DIY power supply I built has a very good performance despite unshielded tranny. Well below 1mV of noise (limits of my oscilloscope sensitivity), thanks to lt3080. I actually don't have tools to measure its noise. May be I'll build a noise amplifier one day...
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: When are SMPS not used these days?
« Reply #19 on: June 26, 2019, 12:28:08 am »
I still see linear supplies in things like microwave ovens and the occasional LED clock radio although those are less common all the time. When you want a high reliability device that will run for a decade or more, a linear supply is cheaper than an equivalently reliable switcher. They are also nice in radio gear because they tend to be quieter from an RF standpoint. Yes it's possible to make a very quiet switcher but you'll usually eat up your cost savings in doing so.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: When are SMPS not used these days?
« Reply #20 on: June 26, 2019, 02:45:42 am »
After that the traces on the CRT were all wobbly because of the stray magnetic field of those transformers. Even though the transformers were in a separately shielded compartment ( More than 1mm thick aluminum between the transformers and the CRT's).

Of course.  Aluminum hardly shields low frequency magnetic fields at all.  You would need a few cm thick plate to do much of anything.  Or: a thin layer of steel or mu-metal or both.

My Tektronix 7603 has the fan modification which was *not* used in the other 76xx oscilloscopes.  The fan has a standard shaded pole motor but Tektronix bolted a heavy folded piece of steel around the stator to shield the CRT from leaking flux.
 

Offline digsys

  • Supporter
  • ****
  • Posts: 2209
  • Country: au
    • DIGSYS
Re: When are SMPS not used these days?
« Reply #21 on: June 26, 2019, 03:29:14 am »
I've designed many linear, fanless powers supplies for mining and intrinsic locations, where they are almost exclusively used.
Also, we still get a lot of orders from critical hospital locations. They are mandated not just for noise, but also are less likely to HV spark, especially in a fault condition.
Haven't seen any talk of S/Mode being approved in these areas to date. Also, still get orders from other "high risk" places.
Hello <tap> <tap> .. is this thing on?
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: When are SMPS not used these days?
« Reply #22 on: June 26, 2019, 03:49:33 am »
I did a multichannel SMPS some time ago, under 1mV noise over an unusually wide bandwidth.  Straightforward design: tight layout, extra filtering on the input and outputs, and a little faffing with snubbers.  (The customer is in this thread, actually; I'd better not talk about it in too much depth. ;D )  I don't think it can be much better without actually shielding the circuit board.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: exe

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: When are SMPS not used these days?
« Reply #23 on: June 26, 2019, 07:20:01 am »
Straightforward design: tight layout, extra filtering on the input and outputs, and a little faffing with snubbers.

How many layers did your board have? Did you use post-regulators? 1mv of noise was p-p or rms value? :)
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: When are SMPS not used these days?
« Reply #24 on: June 26, 2019, 11:19:55 pm »
Multilayer (>=4).  No postreg.  Peak voltage, at 100kHz RBW (measured with spectrum analyzer).

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: exe


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf