Do SMPS have a home at high end audio?

[Sch3mat1c]

Originally Posted by smoking-amp

I recently came across some very nice SMPS on Ebay for automatic test equipment that measure just 5 mV noise (100 KHZ) with 160 V output. While testing them, I kept getting large intermittant bursts of HF noise on the scope. I eventually tracked that down to my cordless mouse on the PC. Sheesh!

Ah yes, UHF+ noise is notorious for driving input stages into nonlinearity, which inevitably results in rectification (either directly, due to base-emitter junctions or grid-cathode diodes, or indirectly, due to slew rate limiting or even harmonic distortion), which results in the amplifier being a crude AM detector. Pulsed transmissions from GSM, mouse, etc. result in popping noises everywhere.

I recently had a problem with college radio station reflections getting into my amplifier (this being a solid state amplifier, probably suffering from a combination of B-E rectification and slew rate limiting). The solution? Big fat ferrite beads and ceramic chips on all inputs and outputs. Kicked it in the butt.

Tube amps should be relatively immune to RFI, due to the higher working voltages, lower transconductance, lower bandwidth and higher device linearity.

Tim

[Cassiel]

Originally Posted by Richard Ellis
....Sand nibbling around the edges, trying to get inside the chassis.....followed by his evil friend..the microprocessor.

That was funny. For me the good thing about SMPS is the weight and size but they will have to be cheap so if one blows another is ready. Now tell me about a SMPS that will stay in perfect shape after 40 years of use and abuse just like a transformer.

[thoriated]

My take on SMPS (and I've designed dozens of them) is that they're problematic at best for audio, unless a considerable amount of time and effort is spent ensuring that they do not inject noise into at least the analog I/O, AC input and even the environment. Even with an implementation that succeeds in preventing any sonic deterioration within the unit itself due to SMPS radiated and conducted electrical and magnetic fields, any equipment to which it is connected will probably be adversely affected sonically due to out of band switching noise injected through its interfaces and chassis. This is particularly true of solid state equipment which varies widely in its SQ tolerance to injected out of band noise due to factors such as internal IC slewing, intermodulation, etc, due to SMPS switching noise as well as good old fashioned RFI, one difference being that the injected switching noise from SMPS powered equipment, CD players et al. more often than not travels through the connecting cable center conductor so is hardly attenuated brather than being rejected by the shield. and SMPS noise is mostly too low in frequency (e.g .< 10Mhz) to be satisfactorily filtered using standard RFI techniques such as ferrite beads and clamps that many claim bring their own sonic problems.

[Richard Ellis]

Well Well.........I've been away for awhile & while I was away guess what happened? The SMPS for our computer screen went bye bye. (That makes four!)
Bought new in 05 & now it's a BIG collection of NON-functional circuitry.
Like I was trying to say....the current crop of SMPS' in 99% of all gear is just plain poorly designed junk!
The unit that went out is just a rectangular plastic box.....So where is the thermal management in a sealed plastic box....I'll tell you...NONE.
When one can design an SMPS that can stand to run 24/7 for thirty years I'll consider it..........won't happen. When your "old" design has one transformer(With filament winding), one tube, two chokes and two caps versus your SMPS with countless components.............How many devices exist to go wrong????
GGGRRrrrr!!!

__________________________________________________ __Rick........

[Eva]

Of course, everyone involved in audio electronics but not interested in learning about switching power conversion will tell that SMPS are a very bad thing.

If you were really interested into learning electronics, you would have repaired the first one that failed, and traced the fault to avoid it failing again.

Some SMPS produce so little heat that they can work in a sealed plastic enclosure day and night. Chinese are not always good at copying these designs... I've got some Chinese cell phone charger replacements with ugly SMPS stuff inside.

But switching power conversion with modern components is very efficient and reliable. Have you ever seen an amplifier capable of 2800W/4 ohm with a heatsink the size of a DVD movie case and no fan? My current work does that and it's about to be released It also does full music output on 2 ohms without fan This is possible not due to high operating temperature but due to very low power loss in the amplifier.

[raypalmer]

Originally Posted by Richard Ellis

Its a creeping thing.........Sand nibbling around the edges, trying to get inside the chassis.....followed by his evil friend..the microprocessor...

Should copyright that and sell it to tubedepot

While I'm never in love with putting any manner of solid state in a tube amplifier I have to acknowledge that linearity can never be what it could without some little black chips... I'm hoping some kind soul posts a build thread of an smps driven tube amp so that I may assess the lay-person-diy-potential (or; lpdiyp).

That said I agree with stalker on the reliability front. Extreme reliability and longevity has an allure to it, especially in this era of throwaway DVD players and cameras. There's not much out there that I could buy and confidently think would still work long after I'm dead. I know a few people who have their parent's, and parent's parent's old-timey radios that still work (some with original caps) That kind of persistence is appealing. I can build an amp and so long as no one harvests it for parts or throws it into a volcano it will play music for my grandkids, so long as tubes are still available...

I think 35mm mechanical SLR, mechanical watch, and tube amp enthusiasts all have something in common. We like to harken back to days of old, when men were men, transite siding was on every house, John Wayne was somehow considered a physically impressive specimen. and when something broke it could be fixed, and it made sense to fix it.... consarnit.

It's a cabalistic dark art from the days of the slide-rule and fountain pen.



P.S. Isn't the internet just a system of tubes?

[Richard Ellis]

Reliable? Reliable??? I have four SMPS' units now...dead..dead.
Add to this two Digital cameras...dead. One was VERY expensive!
Yes, this has become a truly throwaway society.
Eventually we will have five channel All in one units with 99.95% efficiency...no pass-thru losses(1st Appx) all in a compact shoebox size with output power as far as you want to go, but still throwaway devices but tubes are not about efficiency nor absolute "perfect" performance.
Back in '77' I had a Nikon F, straight manual, no Ftn head, F1.4 ,( I could hit the exposure 90% of the time), early model. I was selling cameras back then and I noticed the electronics creeping into the machines....I recall one fellow who had a Canon AE-1 that was 'dunked' in a fishtank as a joke...never to work again. And another story of an individual who dropped his Minolta SRT-101 into the Ocean...after a cleaning out W/Water , the Minolta still worked, unlike the Canon whose electronics went bye-bye rendering the camera useless. Even back then I implored customers to buy "manual" cameras as they won't break-down as easily as the emerging "electronic" ones will.
I guess I will return to film cameras soon....unfortunately the Nikon I bought for $80 is now around $500-800. Your Digital camera won't even remotely last the forty years ol' Nikons do. Your digital camera will be "recycled" a dozen times over to get to forty years.
Keep your SMPS', I'll stick to a straightforward linear supply.

__________________________________________________ _____Rick....

[raypalmer]

Ellis and I are cut from the same cloth it seems...

Crotchety old man cloth.

[Calvin]

Hi,

I think one should first look at the capabilities of a technology and then on how the implementations work.
When I have a technology at hand which is virtually free of noise within my intended working range it shurely is interesting. When I have a technology that reacts much faster than the traditional one it surely is interesting.
If I have technology at hand that comes with safety and overload protection features and accommodates to a wide range of line voltages it surely is intersting. When I ave a technology at hand that reduces weight and size of components this surely is interesting.
What -on the other hand- is so desirable about heave, sizy transformer supplies, needing oversized capacitance banks, that in return generate large current spikes in the rectifier, that create HF-noise. What is desirable about 50(60)/100(120) Hz noise. What is desirable in noise that affects 1:1 the working range against noise that is well outside the audio frequency range and that could effect it only in a indirect and attenuated way?
So far I only made positive experiences with SMPS and to my taste they are the superior technology. And please guys do us a favour and compare good transformer supplies with good SMPS. It´s no evidence in comparing a good transformer supply to a bad SMPS and vice versa.

jauu
Calvin

[richwalters]

Originally Posted by raypalmer

Should copyright that and sell it to tubedepot

While I'm never in love with putting any manner of solid state in a tube amplifier I have to acknowledge that linearity can never be what it could without some little black chips... I'm hoping some kind soul posts a build thread of an smps driven tube amp so that I may assess the lay-person-diy-potential (or; lpdiyp).

That said I agree with stalker on the reliability front. Extreme reliability and longevity has an allure to it, especially in this era of throwaway DVD players and cameras. D

It seems SMPS is getting more a "black art". A growing Lack of experience in both build and circuitry. It appears from this thread so. I've been using the technology from it's onset in the 1970's without any problems from roughest line voltages, Short circuit proof, brown out survival, overloads, efficiency and many more reality scenarios will check the designer. Personally I love SMPS design, and it has been with me now for 40 yrs.

To design a proper SMPS does require Physics approach, (no getting away from this) and a feel for all the worst operating conditions, something which many manufacturers cut corners with by flying into an asic design without a serious layout check.

I have my doubts that some topologies can be made any easier for the lay person even though there are many software design programs around. The cue lies with the actual layouts and a feel for the RF aspect and understanding what one is actually looking at..

There is nothing wrong using SMPS in audio supplies. It ought to be encouraged.
Maybe it's historical; us true Europeans are generally better at solving math problems than those of elsewhere.

richy