Professional audio went to class D amps and switching supplies a while ago because of the huge weight savings. Not to mention less heat and power consumption. These amps have to be lugged around, setup and tore down for each show. They are well designed for reliability and low noise. The sensitive speakers they are used with would make noise apparent. Lots of opportunity for noise with all the cables running around at a venue. The EEs at these companies must be on top of their game.
@@ranbymonkeys2384 Majority of home users are probably using Sony, Panasonic and Japanese/Chinese equipment, which will use a standard off the shelf chipsets. PSU's are more likely to be adapter based switch mode PSU these days. Mainly due to meeting international mains requirements.
@@johnclarke2997 That's fascinating but I was wondering for example 60% of the audio is professional and 40% personal ect..... about what percentage of the market is pro vs personal. Way to switch the subject.
Excellent rant. Thank you! I design software defined radios in my small company and I can say our industry still suffers from the same diseases you detailed for audio companies. The parallel was interesting.
Cool bunch of 'work' stories. I reach retirement age in 2 yrs, a career as an electronics technician. It came from my obsession hobby of electronics as a kid & instead of going to Tech Institute i went from secondary school direct to being employed as a Trainee for a Radio Station where i had to study as i worked & qualifed as a Tech. So often i regret not doing a Degree at Uni but that was my young guy attitude way back when. It's then i have to remind myself that even as a Tech i have been all over the World working in so many countries mostly broadcast gear (cameras etc), repairing, redesigning to customer's requirement even teaching customer's maintenance techs for a product my employer sold to a broadcaster... In some ways if i'd done a Degree I am certain my experiences would have been quite different & i'd never would've had what was in effect a damn great time....
Also spend 1/2 of your career paying student loans. What you did was have somebody pay you to learn. Now days a degree is paying $300,000 for somebody to tell you that you can be a man or a woman and the earth isn't going to be here in 9 years.
…YOU KNOW WHAT GRINDS MY GEARS!!??… The fact that Eddie, Amir (ASR) weren’t a part of my life🥹 when I bought my first (and last) pair of B&Ws. I’ve now graduated to two pairs of DBR62s (second set are spares😂). Seriously though, thank you Eddie and Amir/ASR for reigniting a long dormant obsession with one of my favorite things- Sound/Psychacoustics, oh yeah, and music. The fact the you are giving away years worth of practical experience, knowledge, and wisdom away the way you are, are how you go about it are truly a class act from where I furtively perch. Thank you again Eddie/Kiss Analog. Coming from an aircrew/avionics tech. background, I think it’s a travesty that these “high-end” audio companies aren’t scrambling to hire people such as yourself. Well, if the notion quality/reliability is part of said companies ethos. The whole audiophile/high-end concept reeks of douche/conspicuous consumption/ King-Wears-No-Clothes mentality to me. Audiophool-free zone here. Thanks Eddie, subscribed.
Thanks so much! I appreciate you! I think one problem is that these companies might have an owner or person in charge with insight - but those old cronies that are working under him will not carry his flame or if they do have to hire a power supply guy because of this owner - the management that they then give that engineer is not there and it is reported back up stream that the experiment didn't work... At least that's one story I've seen play out.
I've stopped reading Hi-Fi magazines for many years, a trustworthy writer stopped writing, I like he test all equipment with a science explanation without subjectively say something is good. So I love to watch video that you can prove the good side of switching power supply, and I respect you paid your effort to design a good switching power supply.
Thank you for the video. The deeper I venture into audio world the more misinformation that I realized that's floating around. And your analysis and testing of the analogue circuitry definitely fills a void in my understand on some of the subjects. I'm sure that lots of folks over at diyaudio and ASR will also appreciate your content. Not sure if I missed any of these in your videos. But could you recommend some SMPS models for some of these Class D boards? I'm currently running my Aiyima 4 Channel TPA3255 board (for DSP and stereo active crossover) with a 48V LSR series MEAN WELL SMPS (choosen mostly because of brand recognition and accessiblility). They sound very clean to me at least. Also the topic of balanced and fully balanced DAC/amp still confuses me. Seems to be a lot of marketing shenanigans around this.
I have just finished designing an analog saturation circuit using +/-12 and 5v. After watching your videos I am going to try a mean well SMPS. Thanks for all your insight.
Thanks for the feedback! It sounds like the SMPS might be a Flyback as they typically do need a load to regulate, and it probably would be the 5V output. They also can be more noisy.
In the mid-80's Sony experimented with receivers where the iron-and-copper power supplied were replaced by a switching supply in a sealed sub-chassis. They actually made these receivers with two different model numbers, one for the switcher and one for the linear supply unit. Otherwise they were identical. You could always tell the switchers by the fact that the receiver weighed half its normal amount, and the weight was evenly distributed instead of all on one side. Repair techs (like me) hated them because if you were repairing an amp channel, you couldn't bring it up on a variac to test it, and even dimbulb testers could cause problems. But they did work well otherwise.
Great video man. I’m a 3rd year electrical engineering undergrad trying to get into PSU design for amps but a lot of people are uninterested in power engineering. I think linear supplies are easy to learn and just make sense to a lot of beginners. But to be honest, getting the proper and necessary transformer can be hard, and of course, bulky as hell at any respectable VA rating. SMPS were such a revelation to me, mainly because of how size efficient and cost effective they are. I’m more of a car audio guy so the SMPS boost converters that take 12V to rail level are so incredibly small even for 1000W subwoofer amps. I’m definitely going to be checking out more of your videos to learn more about power/audio engineering.
"It's the way it always has been" is the queue to question the status quo and set out to prove the new facts. Of course, the people they sell to are the one that perpetuate the old thoughts and myths and never question the words marketing people tell them.
Wow, this stuff is pure gold! I learned from your previous video that there is no such thing as a linear power supply, just low frequency switchers and high frequency switchers! From this, the conclusion follows directly: obviously it is better from a noise point of view as well as efficiency to have the power supply switching frequency well above the audio band! From this one I got confirmation of what I've always suspected, that class D amplifiers are just switch mode power supplies too! But taking this to its logical conclusion, why maintain the distinction between power supply and amplifier at all? Why not design a class D amplifier that works directly off the mains? At the very least, integrating the control algorithm would allow the PSU output voltage to be adjusted with the gain setting so the optimal PWM range of the output stage can be mapped to the dynamic range of the audio even at low volume levels. Maybe the real reason for the popularity of linear PSUs and amplifiers (and also of tube amplifiers) is directly related to their inefficiency? They produce enough waste heat to raise the temperature of the listening room and the hotter, denser, dryer air propagates sound better!
ICE Power Amplifiers (available at Part sExpress) do exactly that. Mains in/Audio out. Fully balanced audio circuits too ! A bit grainy sounding but great for Subwoofers.
Great question John! Many Class D amps do use SMPS but to convert line power to the Class D stage does take a conversion process so by definition a converter stage. So there are at least 2 stages if not 3 three. The PFC stage, isolation stage and Class D stage.
This is one of my favorite youtube videos you've made Eddie. "They don't know what they don't know" - brings on the Dunning Kruger Effect quite often. I got a good chuckle out of one of your stories where you were talking about how someone in a company thought the low voltage level measured between the neutral and ground at the equipment was some great achievement. It reminded me of something I had to deal with in my work. I used to do work for a large IBM manufacturing facility that had a special manufacturing safety department who would write their own electrical safety rules / codes. My tasks during those days in part was to deal with high voltage systems. One of the rules was that no where on the raised floors where the workers carried out manufacturing tasks was the voltage difference between neutral and ground to be greater than 3 vac. They would measure random electrical outlets and where the level was exceeded they would call me in to solve the problem. The loads were supplied by large 3 phase distribution panels that would have heavy cabling that sometimes ran 100s of feet back to the distribution switches that were connected to 480 to 208 transformers and that's of course where the neutrals and grounds were bonded together. Just as you stated of course the amount of current carried and the distance the wire had to travel would determine the delta voltage between neutral and ground. I was never able to explain to them that this was quite normal and all the neutral did was carry the current from the phase to phase load imbalance back to the panel and the capacity of the neutral to do this was far greater than the load would ever require. "Nope, the voltage had to had to be dropped". My solution was to measure the loads on individual single phase 120 volt breakers and move them around to try and get the panel as close to balance between all three phases as possible. Funny stuff. I learned that you could never tell the people at IBM anything. But that was ok because they paid well, and merely presented amusing challenges to solve based on 'their playing field'.
8:16 If your house use the TN-S power net like mostly used here in Europe, your RCD will trip if you short Neutral to Ground behind the RCD. As in matter a fact, with a TN-S power net, between Neutral and Ground, there will always be a low voltage present, this voltage rise the further away you are from the sub station, and during day when there is more loading of the net in the neighborhood. Here in Belgium at my home, this voltage is useally between 0,5V to 2V (the times that i have measured it) but i have read articles, that with a TN-S net, even up to 5V is possible between Neutral and Ground.
I know exactly what you mean about enjoying Paul McGowan's videos while cringing at some of his claims and technical explanations. His AC regenerator claims drive me nuts as well. It's outrageously expensive and does nothing for you that you can't get by using a simple filter from a handful of passive components. Also, it uses power which is another waste. He has another "power cleaning" product that's full of nonsense claims and pseudoscience.
@@KissAnalog I drank the PS Audio Kool-Aid for a few years too. But their DirectStream Sr. DAC is the best put there to my ears and Wallet. Yes, the AC Regenerators make me Cringe too and NEVER bought into them. I've used the Stellar SGCD DAC/Preamp and M700/S300 Amplifiers and they are great sounding products. Hypex Switching Supplies with their audio section ripped out and the beautiful sound ing PS Audio (Darren Myers) designed jfet input stage feeding the Class D Gain Block of the Hypex. Paul's book 99% is quite the interesting read if you ever get the chance. Never met the man but toured their factory and sound rooms. Great place with great people, but some of his videos are highly educational and others do make me cringe.
Thanks for the great feedback! I agree with you but the quality products, so I does make me wonder why you'd mix snake oil in when all it does it makes people wonder... I'll have to look into his book - I didn't know he had one;)
@@KissAnalog 99% True. It’s a hilarious and sometimes dark book. How he lost his company and then regained it is incredible. Lots of Pot stories and Army stories too !
@@joeygsaudiochannel3972 -- The cordial and affable Mr. McGowan is not a formally credentialed engineer in any discipline. He's a self-taught audio enthusiast who's picked up some technical knowledge of the course of a long career in the industry, but he's basically more of an entrepreneur and marketing guy.
I recently bought an Aiyima A07 (3255) and a simple Mean Well 48V 9.5A industrial supply and a pair of new Q Acoustics 3030i speakers, holy crap it plays beautifully!
@@KissAnalog LOVE your channel, you're great at explaining things and good fun to listen to! So a very big thanks for inspiration and education! Got a new better ventilated chassis for both PSU and amp with some components for modding the amp on the way, upgraded power inductors, caps and a log 24 step attenuator rather than the linear pot that came with it. So trying to improve where i find they messed up a little :) It's a really fun project!
Thanks Eddie, great chat! Audio is voodoo. You have to run an amp for 100 hours before it works right? Can you imagine if Keysight sent out a meter with a tube rectifier and then said you have to run it in for 100 hours before it works right! This is how I look at it, everything is an instrument, doesn't matter whether it's a scope or an audio amp, the only difference is the end function. It doesn't matter to what I'm designing.
Thanks for your great feedback! I agree with you. I think of the oscilloscopes that have very high bandwidth and 24 bit ADCs - and they all use switching power supplies. Audio companies have small design teams and just simply don't pay well. They also just don't have the best engineering - but don't tell the audiophiles that;) I am speaking in general terms - there can be the outlier...
I know a engineer who designs the consoles used in the music recording studio and the consoles use switch mode PSU's with FPGA devices with advanced DSP techniques. So the equipment used in the studio is now using this technology, many of the old analogue being reserved to specialist work.
Excellent rant, feel free to do so again whenever you have something to rant about. I'm also the type of guy that doesn't do things a certain way just because, today it is rare that a linear power supply makes sense, you certainly can't beat their simplicity, but when power goes past a certain point linear becomes inconvenient and expensive.
Love your channel! I own an integrated Class D amp and am beyond happy with it. One of my main considerations in favor of Class D was simply… much higher efficiency (= less energy consumption) and consequently less heat generational . Plus, the sound was actually better than a highly rated A/B amp I compared to (via listening session). Obviously, acoustic flavors vary and depend on individuals’ systems an other components. But looking forward to eventually add a tube pre, as to get -arguably- the best of both worlds for some of my listening.
If the class D sounds better than A/B (which it doesn't) then why would you need to be looking for "the best of both worlds" you must want to get rid of the efficiency.
@@ranbymonkeys2384 There are sports cars, family sedans, SUVs, vans, pick-up trucks, gasoline, electric, etc. None of them are ‘the best’ without throwing them into their intended environment they’ve been designed for, i.e. roads, tracks, off-road terrain, inner city and so forth. The same goes for audio equipment. We don’t listen to the same music or genres all the time either, so sometimes it might be nice to have -as for example in my case- the holographic sound and warmth of the tubes for certain dedicated listening, hence my ‘best of both worlds’ statement. You might disagree.
@@sveneisenhauer1 Pre amps always make the power amp better in most cases sound better. So let me ask you, if you are riding your dirt bike or cruising in the towncar, is that pre amp going to be used in both situations? Your turn...
Thanks so much Sven! I appreciate you! I think more black box testing would show people how great Class D is. Just like in any class of amp - there’s good and bad;)
Some years back my brother built one of those boxes to use an audio signal to create the 120v power. I think it was more of a novelty than anything. It was fun to use it as a VFD to play around with AC motors, and to play music into incandescent lights, but there really weren’t any practical applications. Up until now, I didn’t realize that people actually used something like that as a method to create “clean” AC power.
I agree with you that if you are going to be a company that design's your own equipment it is essential that you have a good power supply engineer or at least have a power supply consultant on call to help out when problems arise because eventually, they will arise. The company I worked for had a consultant his name was Sam B., and he would come in and he would take a few measurements and give his recommendations and sure enough the problem was resolved.
Thanks for this great feedback! I was once referred to as the 'hired hand' the guy they sent to solve the difficult problems - that needed a solution once and for all. ;) One manager said that I was the tool in the box that they reached for when no other tool worked;) I miss working for large companies that have lots of programs and need help once in a while. Unfortunately, a lot of companies can not afford that guy full time:(
Again another great video, thank you. Is it true that SMPS returns a lot of noise to the line and thus disrupts the work of other devices, as "audiophiles" claim and if so, what can be done to prevent it?
I love this question - thank you!! First, as I’ve shown in videos just prior to this, the amplifier as a thing called PSRR (Power Supply Rejection Ratio). This is where the amplifier shows its immunity to noise at the input. So, every device has an input EMI filter which both protects from the device putting noise into the system and protects from noise coming into the product it powers;) So, I call BS on this;)
Sorry for late reply but from a helicopter perspective it works like this; devices sold to be used on the power grid has to be compliant, there are (in Europe) basically 3 categories. 1. Industrial use, Home use, medical/life saving use. Each class has different limits as to how immune they must be and how much crap they are "legally" allowed to feed back to the grid. So industrial devices has to be able to withstand a metric crap-ton of line noise and voltage surges and magnificent short circuits etc. but they are also allowed a little relaxed rules on what they feed back. (Testing against extreme standards is exciting, quite often things go 'pop' as photonicinduction says) Home use are mid/mid on immunity and emission. Medical devices has the strictest rules; They have to be high on immunity tolerance and low on emissions. I've never plowed down into UL(American/Canada use) as I've been working in Europe only, so I'm somewhat familiar with EN/DIN (resposible for european norms and German industrial norms: both of them are largely homogenized today due to EU and such)
I have a custom 2.1 class D amplifier with a switching power supply, the induced noise indicators are excellent, I think complaints about switching power supplies can be justified if the release is poor
I really enjoyed this video! I worked in high end audio before going back to school to get my EE degree. I used to think that you shouldnt use switchers because it would be difficult to keep up with the switching speed of the amplifier (Needing higher frequency to keep from drooping "Cycle by cycle" vs the power amp switching frequency and also to keep the frequencies from beating against each other) I now see that you can use multiphase switchers in order to increase the frequency much higher in the power supply to overcome these problems. Do switchers have a shorter life span because electrolytics might fail at higher frequencies faster than at low frequencies (Not saying they do or dont - I am asking the question)? I would love to see a video about longevity between linear and switch mode power supplies.
concerning the longevity of switchers, consider how the company SeaSonic makes computer power supplies with a 10 to 12 years warranty. obviously making electrolydic capacitors inside the power supply last so long is probably not easy nor cheap, but it can be done. apparently.
Thanks for the great subject! I’ll do a video on expected life. Switchers are used in aircraft as they generally do have a much longer life. Partly do to using fewer aluminum electrolytic capacitors.
Nice M-T-M monitor in the background. Yup! there is nothing wrong about switching power supplies. There isn't more distortion from them. Period. But in Audio, the audiophiles are cousins of the Flat Earthers. I'm an EE. So I know about switching power supplies, and electronic design. Recall, Technics put out a high quality mini system with a streamlined class A/B amplifier powered by a switching power supply. It put out a very clean 50W/ch. Ran cool, very low in noise and distortion, and simply sounded great. But it didn't weigh a ton, and wasn't massive. So it was rejected like a red-headed stepchild by audio purists. Too bad. Digital amplification came out in 1998 from Texas Instruments. But thanks to stubbornness and ignorance, Class D amplifiers didn't take off until 20 years later. I'm glad luddites are dying off.
Thanks for that great feedback! I think you are so right about the Flat Earthers;) Maybe we can make a little noise on this channel and get people to think;)
What I don't understand is why the smaller niche audio folks just don't use an off the shelf switching power supply. They don't necessarily have to hire an expensive switching power supply engineer! Not sold on Class D yet. But have been a switcher advocate since the mid 90's.
Great question! It really isn't easy to find an off the shelf power supply wiith the right voltages and current ratings. Also, there are 2 cost considerations, the NRE (non re-occuring engineering) which is what you pay for the engineering development, and then the re-occuring engineering cost for every product that is built. Buying a power supply for each unit you build is expensive compared to the raw cost of parts that it cost to make your own.
The Amplifiers and other equipment from Linn Products in Scotland use SMPS. They’ve been using SMPS since the 1990’s and continued to improve since then, The latest version called Utopik is an improvement on their previous ‘Dynamik’ SMPS. Their power amps too also use SMPS including their latest 800W Monoblock
the icepoweraudio ASX series look interesting. it's an integration of a high power class D amplifer and SMPS in a single pcb module. Also, there's Quilter Amplifiers for guitar. Class D and SMPS and some circuitry to emulate tube amp response, like applying current feedback to give lower damping factor. i'm working on a guitar amp project now using a class D power section and tube preamp and SMPS for both power amp and high voltage for preamp.
What about output current of amplifiers with SMPS vs linear? I've noticed that almost every Class D amp has a max output current of about 30 A, Class AB have no problem with 45, 60, 90, even 180 A for something like the Parasound JC1+ monoblocks. When you're trying to control big speakers with lots of drivers that current can come in handy.
Thanks for the great question! I'll use the Parasound as an example of how those specs (peak current of 135A) is not possible. It is more marketing BS. Pro sound uses switching power supplies on concert stages driving those big speakers without a problem. There's so much misinformation.
i'm using an old sony integrated amplifier with one of their "pulse power supplies", a ta-ax6. i have several other models with them too. sony really wanted to do that back in the late seventies. i wonder how successful they were with it? they seemed to go back to linear for the eighties.
interesting. i've had this 40 year old ta-ax6 for probably a decade and it seems to be doing well. i've actually got two of them. and a ta-f70. and a ta-f45 and 55. and a ta-n86b. they all have the sony pulse power supply, are all over 40 years old, and all work, aside from some dirty switches and blown bulbs. that seems pretty reliable. @@dongwarrenmusic
Great rant! Of course, nowadays you can build a world-class hifi amp out of professionally-designed OEM modules for both SMPS and Class D output stage, leaving one major marketing challenge: product differentiation via cosmetics or a supposedly "audiophile" input buffer designed in-house. This is yet another reason why many of the struggling little "high-end" audio companies cling to obsolescent technology like linear PSUs and Class AB -- or even Class A -- output stages.
Thanks Bruce! Well said! I plan to design a Class D with SMPS power supply that will have a sweet front end on the amp and the PS will have an active power factor correction front end;)
@@KissAnalog -- Great project, Eddie. Looking forward to learning what's involved in designing such a sophisticated SMPS. Are you going to buy a power amp module or roll your own?
Disclaimer: I am NOT an EE! For high current amplifiers, a [large] transformer has something a switching PSU does not have: lots of iron and the "flywheel effect'. Same idea behind low-frequency inverters w/ large transformers -vs- cheapeer high-frequency inverters. It is my understanding that those older style inverters (with large heavy transformers) can spin up dificult (inductive) loads with ease - while the newer high-frequency ones always struggle. The electro-magnetic force stored up in large chunks of iron act as a large flywheel and seem to deliver more "punch" during large instantaneous demands. I am sure modern Switch-Mode PSU's have ways of getting around this - but that has always been my understanding.
Thanks for the great question! Yes you are right about the large iron power supplies. The problem with the switchers is that they have a much better control. So, when an over current event happens - they protect quickly. Problem is - they don't want to over load themselves. Solution - get a inverter rated for your surge current and don't blame the switcher for doing what it is supposed to do. So, rule of thumb - most of these cases can be solved by buying an inverter with - maybe twice the nominal power - or buying one that is designed to handle large surge currents. The new switchers will be better in every way if they are sized appropriately;)
Not sure about the title but amplifier brands do use switching power supplies. Depends on the brand, model..etc and not just class D amps that use switching supplies. Its more popular in pro amps to use switching supplies but it happens otherwise too - just not as common to linear supplies.
Liked the rant and your discussion on power supplies vs. switching supplies. I also liked your mention of Amir at Audio Science Review. I've been following Amir for sometime now and like his no b.s. reviews.
Audio amps do use switching supplies. I have several and they are all noiseless and distortionless. And do remember that so-called linear PSUs are in fact switchers. They just switch at 100/120Hz instead of a few hundred kHz. But they work exactly the same way - dumping charge into a capacitor from a variable mark/space ratio source.
The first switching supplies were introduced in the high end audio field by Meridian with their MCA and Chord with their 1000 model. I was lucky enough to see the prototypes at the time in the early 1990s. We were cynical that these would perform but they tested immaculately. The Chord designer/owner was a power supply designer for the Ministry of Defence in the UK working on Tornado aircraft amongst other things. Boothroyd/Stuart were amazing designers. The best audio designers were all engineers and all the great advances came from science. Bose springs to mind in the modern day who went off at a tangent - But all pure science. The greatest engineers historically in audio were Blumlein and Armstrong who between them invented almost everything that laid the foundation of the industry. No pseudo science has any place with the great designers in audio.
Thanks for this amazing feedback! Really cool information! I’d love to see one of those Meridian units on my bench;) Pseudo science has no place for real engineers;)
Does a class D audio amplifier just have a switching power supply to create the DC power rails, or does it also use switching to produce the audio output amplification?
Yes, the Class D is made with switching - see this video I did to try and explain: How Class D Amplifiers Work part 1 with EPC GaN FETs ruclips.net/video/GCA3t67yOLc/видео.html
@@KissAnalog I just watched that video - that answered my question perfectly! Really good with the whole setup and scope of the input vs output. Thanks!
Back in the early 80's switching power supplies were just making there appearance. We were using them in Jerrold CATV amps. August was lightning season and we would replace 15 or 20 in a night. We started noticing that as the linear power supplies were the only ones getting changed.
IDK if you remember BASH amplifiers in the 90's, but they used switching power supplies and were really compact / high output, and very clean sounding. That company was short lived, because soon after, everyone went to cheap chinese stuff.
@@KissAnalog ELAC makes a BASH amp right now. There is some technical information on the setup if you goto the ELAC website. The company in Canada was Indigo who had ST micro electronics fab up the IC's for B.A.S.H. I forget the abbreviation for the technical term bash. It was over 20 years ago so the company went under, and the IC's were still protected with a patent until recently. Looks like ELAC had ST microelectronics re fire up the IC's since the patent expired.
I really liked this video. To be honest I even wouldn't call it a rant. This was informed background info from a professional. Always something to be learned here. Thanks Eddie.
You are right sir Thanks for this video I watched your full video I watched pal audio video I watched audio science video Thanks for putting all in one video
Well Manley didn't just hire "some" power supply guy. Bruno Putzeys might be one of the most influential Class-D technology audio engineers on the planet in IMO.
I have seen similar company behavior myself in the past. They stick to what they know best (and have been able to charge to customers successfully for ). But then often the market shifts under their feet. Some of those company do not survive. Were the Carver's "Magnetic field Power Supply" some kind of switching power supply? Yamaha lost a patent suite when they were copying the idea (e.g. with the A-760). Were those somehow SMPS ? More important : any advise on where to start if we were to try SMPS on our diy audio gears ? Some hints would make a great video ! Thank you for your content !!! Always instructive.
I think you are right - many companies have not made it. I'll have to go over the Carver design in a video. I think it was very interesting. I appreciate you!
I own two REVOX B251 amplifiers built in the early eighties with switching mode power supplies. I always found it was sounding weak and dull thinking the SMPS was the problem. Now you make me wonder if it's because of SMPS design instead... I love so much my REVOX stack and don't want to get rid of the B251. Here I go again...
A couple of years ago I started messing around in audio by replacing electrolytic caps with film caps, stuff like that. Sometimes I can improve the sound and other times I mess the whole thing up. Well, I have to be soldering something or I start shaking and go through withdrawals. How do I get into a job where I can do something like that?
@@JasonW. I was actually a welder fabricator for 14 years maybe that's where it comes from, electronics I think is more personal, not to mention cleaner.
I generally agree with your take in reference to Amir's (Audio Science Review) criticism of PS Audio's device - - the thing is, at least for me, and yes I'll say up front that I enjoy Paul and yes he is an engineer and has been responsible for many audio design improvements over the years and admire Amir as well - but the thing is, clearly Paul's device provides a better sine wave and the primary thing I took from Amir is that the noise floor actually increases from that of an unprocessed outlet - and, rightfully so, Amir was also critical about the marketing aspect of how PS Audio presents their product - that their marketing doesn't live up to what Amir measured including the implications of impendences - however, whether or not the device actually has an impact on sound is never covered by either - Amir dismisses it out of hand telling us it's a waste of money but doesn't give it a sound test and neither did Paul in his follow up to Amir's review - - all too bad - - imho, if the impedance hit isn't great enough to impact the gear being used, obviously the noise floor, although worse, is still low enough to have a minimal impact which leaves me wondering whether the more perfect sine wave is indeed a benefit - Amir claims it doesn't matter in that a well designed power supply can overcome a less than ideal sine wave - - so, I get his science, but where's the listening tests? Why don't we get both measurements and listening test? J. Atkinson has been testing for years and years with equipment similar to Amir's - I would bet that John would still imply that measurements don't always correlate with what's presented as a final musical presentation - that's the catch for me - I like and want measurements but, to this very day that doesn't mean one can "always or necessarily" predict sound from specifications.
btw - I'm running class D and haven't looked back - of course I freely admit that there's plusses and minuses to all three approaches (Class A, Call A/B, Class D) and yeah - I don't have experience with some of the other exotic approaches like Bob Caver has marketed.
I developed semiconductors for the audio industry, and it is indeed full of pseudo science. In fact, we used to joke that some customers would like AC coupled inputs because it gave them a chance to change from red caps to blue caps and "improve the soundstage", while another customer would do the opposite. We used big Agilent SMPS lab supplies to evaluate our power amplifier chips, by the way.
Lol, I believe it. I have to say, the best sounding amps I've heard to date were designed around VFETs (Static Induction Transistors). Could always just be a coincidence.
An ideal power supply should have no noise, plus the output impedance should be 0 ohms throughout the frequency range and sudden loads should not cause ringing, small slow variations in voltage are for most amplifiers unimportant. I'm missing Data for SMPS output impedance , I can't find this data either at Hypex or Cornex, do you have a link to such data.
Great feedback and questions. You are correct that no noise would be ideal - but doesn't exist - with any power supply. Some think that switchers are more noisy - but their noise is in the high frequency range that is out of the audio range. So called linears (which are really slow frequency switchers as few are actually 'linearly' regulated have noise in the audio range (60Hz or 50Hz which also results in 120Hz and 100Hz). You are also correct that output impedance ideally would be zero ohms, but no power supply is zero ohms. Actually the low frequency switchers (not linearly regulated) have higher output impedance than switching power supplies. Switching power supplies are regulated so will show much lower impedance than using just bulk capacitors off of rectifier diodes.
Linear can be understood by nearly anyone. Easy to design, fix and upgrade. They looks good and can be cheaper. Not always better. If space is limited switch mode can be more dynamic.
I had a Fender 100 watt GTX with ICE power class D. Very low heat, almost no heat sink. Was impressed. BUT, the class D power section did not sound as good as my MOSFET amp. I would like to build a MOSFET and try the switch mode power supply. I remember TI was working on a good Class D amp about 12 years ago, the engineers said they wearing quit there yet, or even close. Which brings me to your video, let's sell it even though it's really not ready, lol. No class D for me, yet.
and, I could learn a lot from you too. thx for this video. I realy had to hear other opinion about wich PS is better. I think that if smps's would work on higher frequency (more than 100kHz, maybe MHz...ishh) and garantee big spikes, thay would be great option for power amps forever. I don't have almost no knowledge in this area but I ask myself, how can a smps do what a big ass cap can do? and realy still get a big efficincy...
The customer base for high end audio products is conditioned to accept subjectivity. As long as they spend their money, don't rock the boat. There are renegades, but the market - opinionated and willing to spend money - demands what it demands. We need more renegades.
Nice video I wonder if smps have reserve power like those power supplys based on very large capacitors,around 20,000 mf and up which act like powe reserve bank that can deliver high current when needed, like music passages of kick of a drum, combined with tough loads that can go low in impedance (some loudspeakers can go down below 3 ohm in certain frequencies), it's not all about continuous power, it's also about transient power, can they deliver?
Thanks for the great questions! So one thing to think about, the so called linear power supplies (they are slow freq switchers - unless they actually have a regulated output) are very slow, and the switchers are 100x faster so the caps can be 100x smaller as they can react so much faster;)
@@KissAnalog can it be simulated and measured in the lab these situations ,maybe making square wave load and to see on oscilloscope or other method the behavior of the output?
For the first time in a long time, I didn't feel cheated listening to a long video. Yeah, it's possible to make reliable switcher, but few can do it. That's why, often, military surplus is the best there is. I don't know too many audio companies with an excess of development funds, either.
it's alway a trade of between quality and margins, a lot of electronics and other things are of so so quality and happen to break fast which is in there interest, some things are gettig way to hot because it is tightly dimensioned.
That’s so true! Just because a circuit is efficient, there still needs to be spacing to spread what heat there is. Linear might seem to be more robust because of the fewer parts, but the large aluminum caps will wear out as they outgas from the day they are built. That’s why there is such a thing as ‘recapping’ where all the aluminum caps are replaced after time.
Yes, all my commercially made HiFi equipment are now all powered by SMPS, I was previously a linear high speed feedback based regulator fan but times have changed.
@@KissAnalog You should explain why this happening and explain the reasons. Cost and performance are reasons, maybe explain the aspect of costs, as well as performance and also the efficiency and size advantages and also the Before and Now solutions and relative performances, the FET and its appearance and also ICs. ASR does the performance but never explains how it is done. Wit ha regulated, the extra overhead of dissispation , cost and energy is removed, I'd like to know more about the different methods of SMPS and the Pro Con in Audio.
A lineair power supply is also much more basic it's reliable , switching power supply's are sensitive to ac interuptions (by the power company) and whenthe ac goes off and on fast in a row then switched mode supply's wil tend to break, what's you're opinion about this? and is it true, or is the PSU bad designed then?
Thanks for the great questions Ronald. Actually a linear with the huge caps is a bigger problem with power interrupts. The NTC typically used to slow the inrush current is hot and then can’t slow the inrush with a quick power cycle. Also, some use relays to short the NTC after the inrush - and these have so many cycles. A Switching power supply designed correctly should not care about power cycles or brown outs. Military uses switching power supplies for the reliability.
The problem you describe is common. It's basically caused by people not going for the company goal but individual goals (which in the end they'll always do), which haven't been aligned. The reason is simply a lack of communication.
Could it be because of the waight?, when you buy a amplifier that's for instance 500 watt, and it's very light , maybe people think does this hting do the 500 per channel ?, because the weight is because of the transformer.
Thanks Ronald, I think there’s truth to that. Also the big transformer and large caps look impressive when they show pictures of the inside of the box;)
Hey, while I've got your attention...if I got your attention... How about a rant on how software/digital has moved into the analog realm? By this I mean...that switching power supply controller chip might actually have some software inside of it. Right? Tell us how you feel about that! Us embedded software guys are very interested! In any case I love your videos!
I constructed a power amplifier using tda7294 ics, and a 500 watt smps. The combo sounds fantastic, and dead quiet on idle. This coming from a guy with a linear ps fetish. I think the days of linear supplies are numbered.
Something else Eddy Another thing, don't tell elon musk that the Clean Air Solution for the Electronic Car is not as clean as suggested because the energy from the power station is miles of wires to several transformers for his electric car and the power plant runs on coal. I wonder what percentage only comes from our sockets
Audiophiles hear stuff the instruments don't, it's magic. Or snake oil. I trust measurements, as such audiosciencereview is great. But I also understand that people like the character of a certain kind of sound, which is why we have vinyl and tube amp fans. In my opinion there's no shame in liking the character so if you're into it, go for it. Nowadays we just don't have to buy into the snake oil anymore, which is great.
Thank you! I totally agree with you! We should be able to characterize the sound with measurements. If someone wants a certain sound - it would be nice to say - oh look for this in the spec.
yes I wasn't sure about that ps audio regenerator. you can do the same with a site transformer and some motor run caps, transformers on the primary and secondary of 5uF one thing I wasn't sure is how many run capacitors on each section? what would be plenty 60uF or 120uF divided between the secondary and primary? got any feelings on that! and I haven't noticed but why don't they use split film capacitors on switch mode power supplies, surely this would be better approach! and the ideal capacitor. I know one company that would appreciate your talents and that's bang and Olufsen and ice have you tried? why can't you sell your own switch mode power supply for valve gear and solid-state on the open market isn't that better?? and companies will try it and realise not just enthusiasts.
Thanks for this great feedback!! The regenerator is completely unnecessary! It is almost a hoax - IMHO. Offer amps with better designed power supplies if you really believe this is an issue. That is a great idea! I should give B&O a shot. But, I really don't want to move so I'd want to work remote or travel often One thing I am doing is showing how to design these on my RUclips channel;) I have thought about offering these for sale as you suggest, and maybe I will. I'm an engineer and know that I'm not good at business, but maybe I should give it a try. I'm also bothered a little that I'm teaching these companies how to do it for free. I like the idea of teaching the diy community and those that want to be engineers. But I hate to give away to these companies for free that they wont pay for.
For world wide product distribution (sometimes called a universal supply) the only real viable power supply topology is a switching regulated supply. Add to this the feature of PFC and the power supply could have multiple switching controllers. If you look at the possible EU requirements for PFC and the wide range of line voltage and frequency world wide, with the added challenge of accommodating high and low line, and testing per IEC61000-4-11 / 4-14 / 4-28 which targets AC power line disruptions, the PFC + switching regulated supply is the only real option. But, as stated in the video, the design goals are more of a challenge due to component stress and qualification. For example, in looking at high speed diode turn on overshoot (a type of stress) the test oscilloscope needed should have a 1GHz bandwidth (to actually see the rise time characteristic) which is not only expensive, but will require the test fixture be well designed and can maintain high measurement fidelity (also not easy). Add to this the detailed study of virtually all of the passive components (also needed for high quality models for use with circuit simulation), and equally challenging, to effectively design the PCB (which should include testing for pre-compliance (conducted and radiated EMC)), and you can see the task is not easy or inexpensive.
Thanks for a very interesting presentation, we shouldn’t forget that the audio industry is partially in the sales of fantasies and despite all the advantages of SMPS the linear ones look much sexier with their large toroidal transformers and tons of electrolytic capacitors, I like to show off with the 2 large toroidal xformers and 1/4 Farad battery of electrolytic capacitors in my 22 kilos diy Pass F6 and it gives me the impression that it makes the sound better😅 Looking at another fantasy industry: who is willing to swop the 6 liters V8 (360 cubic inch for my fellow Americans) engine in his muscle car for a much smaller turbo charged 6 inline that weighs half, runs on 20% less fuel and give it’s full torque from very low RPM? The V8 is much more impressive 😊
you are right, big heavy, transformers in liniar power supplies are cool. its kinda like driving a sports car, that cosumes more fuel just to be more cool. but i have to admit that they are just worse and impractical, especialy for high power
I have one of these Sunfire TGA-7400 - From the Manual - The big breakthrough is the uncanny tracking downconverter, which uses 18 Herculean Mosfets in the TGA-7400 and the TGA-7200. This allows the amplifi er to produce phenominal power output, without the heat generated by conventional designs
I very much liked this video, well-made and told! :-) I think of the SMPS a bit like I look at the CD-players in the start, nobody knew how to make it sound good, but got it to work in the end. You can, as you say, add to that their bashing of the SMPS because of them been badly designed and everybody "knows" how much they mean for an amplifier. I may have a suggestion for a new rant: if you ask a non audio-interested electrical engineer, does every amplifier sound the same, you can take a look at the spec and determine that none of it is audible. Therefore, does a $200 SMPS amplifier sound exactly the same as an $50,000 class A amplifier, and everything you may imagine hearing in différance, is, yes, due to you thinking thee is a difference. Is that true, sort of true, or simply wrong?
Thanks so much Fried Mule;) I think there's so much mis-information and lack of testing methods to truly characterize an amplifier - that we need to start working out a way to do that now.
An unregulated linear supply is efficient enough and will last much longer than any switching power supply. It is no fun when one rail fails and you blow your speakers. They are also much easier to design and build.
Thanks Florian! Actually, if designed well, a SMPS will outlast any Low Frequency Switcher (not linear unless it is regulated). The aluminum electrolytics have a life span on them but switchers can be made without them. The International Space Station and all other SPACE born electronics are switch-mode and have to last 20 years with zero failures. You have probably heard of capacitor reconditioning or re-capping? This is because aluminum caps dry out. You are correct - SMPS do take some expertise;)
Jeff Roland makes A/B amps with Switching power supplies but their stuff is too dark. Mola Mola uses switching power supplies on all their audio products from Dac's, Preamps and amps but their stuff is analytical and their parts are a little cheap to a lot cheap. The problem is everyone has an "opinion" and aren't willing to try new things. They still love their tube amps or big Boulder sized 150 grand boulder amps. The best things I ever heard were Pilium Amps. I don't know how they do it but omg. This greek company is amazing. One of their amps is 30-50k. The guy that started it used to make treadmills of all things. So at least he was a real engineer. They use multiple Torrids that are shielded in little round cylinders. It is more of the quality of the power supply not the design and most "elite" companies use junk parts even if they are expensive.
Thanks for all that great feedback!! It is great when the management hires great engineers and lets them use the parts that they need that make the difference! I think a lot of these companies run with an amp engineer that also tries to do power supply design - because its easy - right;)
@@KissAnalog YEP even a lot of "upgraded" linear power supplies are junk! take a look at the one that is $1200 for a headphone amp from Ferrum Audio it's all junk parts! They charge you extra for it and put it in a pretty box. a pretty box always makes your external linear power supply better!
Thanks for asking. I could probably elaborate a bit more on this story, but I can't say names as this engineer had to sign a paper that he wouldn't say certain things about this company. Sad, but that's how companies protect themselves after damaging someone that they don't want others to know about.
@@KissAnalog Thanks for the reply. I was thinking the technical side when I commented. Maybe a few examples of cases where it's better to do it your friend's way. And a few good practices would be great too. Thanks.
Yes, love the rant idea! It's funny that I follow the same youtubers where audio is concerned anyway. I have been watching for some years now, trying to wade though BS and snakeoil. Also trying to learn as I go. eg I decided on the (Chi-Fi) Topping D50s DAC I have after following ASR, I'm sure there is always something better, but at the price it cost, 2nd hand, I am not disappointing! It was as much an experiment ie, do I really need this, will it make as much difference as "THEY" say? My opinion, is it very much depends on how you listen. I like to spend an hour or so just relaxing in the dark, or at least dimly lit room, sat in the sweet spot listening to some of my favourite music ( whatever it is on the day). In this scenario it made a good difference. The other scenario, where the music is on, but you are walking about doing housework, cooking, whatever, a lot of these add-ons or improvements they want you to buy seem to make no damn difference other than emptying the wallet. You answered a lot of questions I had more diplomatically than I could ask them lol. Thanks Eddie.
I believe that Benchmark AHB2 power amp is using SMPS inside. It’s lite and also not even class D (it’s based on THX AAA feed forward, i.e combination of class A + class AB design)
TBH i dnt know why PSU arent regarded as the most important thing in an amplifier rather than the signal chain. thats what were actuallt modulating. 3 months learning to make my own tube amps made me realize this.
Thanks so much! I think the big reason is cost - and audiophile would rather spend mucho dollars on cables than ask for a properly designed audio amplifier;) The industry has done a great snow job!
Just over 25 years ago I remember listening to my first D class amp and it sounded horrible switch mode power supplies and D Class amps technology has come a long way and now are better than AB class amps because they are more reliable and don’t dissapate heat and they are more efficient and are friendly to the back.
great to hear your stories. trying to find my way in my late start of work life. i am learning as i go and don't want to be someone that believes in falsehoods. didn't know someone as knowledgeable as you would be so underappreciated! working in a work place with others that want to also show what they know is hard! how to go about it, and wanting to show what i know too while not telling people.
Thanks for the great question! At times they could - but even then they want to have their name on the stuff inside their box so they can charge what they want;) Also, it is usually less expensive in parts to make your own, so buying one it would cost too much in parts.
Professional audio went to class D amps and switching supplies a while ago because of the huge weight savings. Not to mention less heat and power consumption. These amps have to be lugged around, setup and tore down for each show. They are well designed for reliability and low noise. The sensitive speakers they are used with would make noise apparent. Lots of opportunity for noise with all the cables running around at a venue. The EEs at these companies must be on top of their game.
What percentage do you think professional audio is compared to home (personal) in the form of overall need and use in the market?
Yes, QSC, (Quilter Sound Company, name of founder like KA said) did ABX testing when they switched to switching.
@@ranbymonkeys2384 Majority of home users are probably using Sony, Panasonic and Japanese/Chinese equipment, which will use a standard off the shelf chipsets. PSU's are more likely to be adapter based switch mode PSU these days. Mainly due to meeting international mains requirements.
@@johnclarke2997 That's fascinating but I was wondering for example 60% of the audio is professional and 40% personal ect..... about what percentage of the market is pro vs personal. Way to switch the subject.
@@ranbymonkeys2384 No pun intended
Excellent rant. Thank you! I design software defined radios in my small company and I can say our industry still suffers from the same diseases you detailed for audio companies. The parallel was interesting.
Thanks so much! Great feedback! I realize now that this story must be true for many in parallel universes:)
Linear power supplies have more nostalgia, and they keep you warm in winter.
LOL So true;)
Imagine how big a linear PSU would have to be for a modern computer. 😂
LOL, my laptop would not be a laptop;)
Cool bunch of 'work' stories. I reach retirement age in 2 yrs, a career as an electronics technician. It came from my obsession hobby of electronics as a kid & instead of going to Tech Institute i went from secondary school direct to being employed as a Trainee for a Radio Station where i had to study as i worked & qualifed as a Tech. So often i regret not doing a Degree at Uni but that was my young guy attitude way back when. It's then i have to remind myself that even as a Tech i have been all over the World working in so many countries mostly broadcast gear (cameras etc), repairing, redesigning to customer's requirement even teaching customer's maintenance techs for a product my employer sold to a broadcaster... In some ways if i'd done a Degree I am certain my experiences would have been quite different & i'd never would've had what was in effect a damn great time....
Also spend 1/2 of your career paying student loans. What you did was have somebody pay you to learn. Now days a degree is paying $300,000 for somebody to tell you that you can be a man or a woman and the earth isn't going to be here in 9 years.
Sounds like an amazing career!! @peteb2
Great point! I paid my loans and it wasn't easy at times! OJT is what we called it in the military (On the Job Training;)
…YOU KNOW WHAT GRINDS MY GEARS!!??…
The fact that Eddie, Amir (ASR) weren’t a part of my life🥹 when I bought my first (and last) pair of B&Ws. I’ve now graduated to two pairs of DBR62s (second set are spares😂).
Seriously though, thank you Eddie and Amir/ASR for reigniting a long dormant obsession with one of my favorite things- Sound/Psychacoustics, oh yeah, and music.
The fact the you are giving away years worth of practical experience, knowledge, and wisdom away the way you are, are how you go about it are truly a class act from where I furtively perch.
Thank you again Eddie/Kiss Analog. Coming from an aircrew/avionics tech. background, I think it’s a travesty that these “high-end” audio companies aren’t scrambling to hire people such as yourself. Well, if the notion quality/reliability is part of said companies ethos.
The whole audiophile/high-end concept reeks of douche/conspicuous consumption/ King-Wears-No-Clothes mentality to me. Audiophool-free zone here.
Thanks Eddie, subscribed.
Thanks so much! I appreciate you! I think one problem is that these companies might have an owner or person in charge with insight - but those old cronies that are working under him will not carry his flame or if they do have to hire a power supply guy because of this owner - the management that they then give that engineer is not there and it is reported back up stream that the experiment didn't work... At least that's one story I've seen play out.
I've stopped reading Hi-Fi magazines for many years, a trustworthy writer stopped writing, I like he test all equipment with a science explanation without subjectively say something is good. So I love to watch video that you can prove the good side of switching power supply, and I respect you paid your effort to design a good switching power supply.
Thanks for your great feedback Johnson Lam!
Well, that's you off of Paul McGowan's Christmas card list.
LOL Thanks;)
@@KissAnalog The funny thing is that I wrote the comment once you started talking about audio snakeoil. Before you even mentioned PS audio.
Now that's really awesome;)
For a "rant video" this is very pleasant and doesn't get me down. It's educational with real word people examples. I appreciate that a lot
Thanks so much!
Thank you for the video. The deeper I venture into audio world the more misinformation that I realized that's floating around. And your analysis and testing of the analogue circuitry definitely fills a void in my understand on some of the subjects. I'm sure that lots of folks over at diyaudio and ASR will also appreciate your content. Not sure if I missed any of these in your videos. But could you recommend some SMPS models for some of these Class D boards? I'm currently running my Aiyima 4 Channel TPA3255 board (for DSP and stereo active crossover) with a 48V LSR series MEAN WELL SMPS (choosen mostly because of brand recognition and accessiblility). They sound very clean to me at least. Also the topic of balanced and fully balanced DAC/amp still confuses me. Seems to be a lot of marketing shenanigans around this.
Thanks so much! Have you seen my recent Class D videos?
I have just finished designing an analog saturation circuit using +/-12 and 5v. After watching your videos I am going to try a mean well SMPS. Thanks for all your insight.
Thanks for the feedback! It sounds like the SMPS might be a Flyback as they typically do need a load to regulate, and it probably would be the 5V output. They also can be more noisy.
In the mid-80's Sony experimented with receivers where the iron-and-copper power supplied were replaced by a switching supply in a sealed sub-chassis. They actually made these receivers with two different model numbers, one for the switcher and one for the linear supply unit. Otherwise they were identical. You could always tell the switchers by the fact that the receiver weighed half its normal amount, and the weight was evenly distributed instead of all on one side. Repair techs (like me) hated them because if you were repairing an amp channel, you couldn't bring it up on a variac to test it, and even dimbulb testers could cause problems. But they did work well otherwise.
Thanks Scott! That is very interesting! I appreciate you sharing that story!
Great video man. I’m a 3rd year electrical engineering undergrad trying to get into PSU design for amps but a lot of people are uninterested in power engineering.
I think linear supplies are easy to learn and just make sense to a lot of beginners. But to be honest, getting the proper and necessary transformer can be hard, and of course, bulky as hell at any respectable VA rating.
SMPS were such a revelation to me, mainly because of how size efficient and cost effective they are. I’m more of a car audio guy so the SMPS boost converters that take 12V to rail level are so incredibly small even for 1000W subwoofer amps.
I’m definitely going to be checking out more of your videos to learn more about power/audio engineering.
Thanks so much! Here's a great video for a power guy;) ruclips.net/video/fQSlo1zmAlA/видео.html
The thing to do for new grad is look at available jobs where you want to live and maybe concentrate on what they are looking for.
You can always move around to find the job you want to have.
"It's the way it always has been" is the queue to question the status quo and set out to prove the new facts.
Of course, the people they sell to are the one that perpetuate the old thoughts and myths and never question the words marketing people tell them.
So true! Thanks for the great feedback!
Wow, this stuff is pure gold! I learned from your previous video that there is no such thing as a linear power supply, just low frequency switchers and high frequency switchers! From this, the conclusion follows directly: obviously it is better from a noise point of view as well as efficiency to have the power supply switching frequency well above the audio band!
From this one I got confirmation of what I've always suspected, that class D amplifiers are just switch mode power supplies too! But taking this to its logical conclusion, why maintain the distinction between power supply and amplifier at all? Why not design a class D amplifier that works directly off the mains? At the very least, integrating the control algorithm would allow the PSU output voltage to be adjusted with the gain setting so the optimal PWM range of the output stage can be mapped to the dynamic range of the audio even at low volume levels.
Maybe the real reason for the popularity of linear PSUs and amplifiers (and also of tube amplifiers) is directly related to their inefficiency? They produce enough waste heat to raise the temperature of the listening room and the hotter, denser, dryer air propagates sound better!
ICE Power Amplifiers (available at Part sExpress) do exactly that. Mains in/Audio out. Fully balanced audio circuits too ! A bit grainy sounding but great for Subwoofers.
Great question John! Many Class D amps do use SMPS but to convert line power to the Class D stage does take a conversion process so by definition a converter stage. So there are at least 2 stages if not 3 three. The PFC stage, isolation stage and Class D stage.
This is one of my favorite youtube videos you've made Eddie.
"They don't know what they don't know" - brings on the Dunning Kruger Effect quite often.
I got a good chuckle out of one of your stories where you were talking about how someone in a company thought the low voltage level measured between the neutral and ground at the equipment was some great achievement. It reminded me of something I had to deal with in my work.
I used to do work for a large IBM manufacturing facility that had a special manufacturing safety department who would write their own electrical safety rules / codes. My tasks during those days in part was to deal with high voltage systems.
One of the rules was that no where on the raised floors where the workers carried out manufacturing tasks was the voltage difference between neutral and ground to be greater than 3 vac. They would measure random electrical outlets and where the level was exceeded they would call me in to solve the problem. The loads were supplied by large 3 phase distribution panels that would have heavy cabling that sometimes ran 100s of feet back to the distribution switches that were connected to 480 to 208 transformers and that's of course where the neutrals and grounds were bonded together. Just as you stated of course the amount of current carried and the distance the wire had to travel would determine the delta voltage between neutral and ground.
I was never able to explain to them that this was quite normal and all the neutral did was carry the current from the phase to phase load imbalance back to the panel and the capacity of the neutral to do this was far greater than the load would ever require. "Nope, the voltage had to had to be dropped". My solution was to measure the loads on individual single phase 120 volt breakers and move them around to try and get the panel as close to balance between all three phases as possible.
Funny stuff. I learned that you could never tell the people at IBM anything. But that was ok because they paid well, and merely presented amusing challenges to solve based on 'their playing field'.
Thanks for this fantastic story - awesome feedback! I appreciate you!
8:16 If your house use the TN-S power net like mostly used here in Europe, your RCD will trip if you short Neutral to Ground behind the RCD.
As in matter a fact, with a TN-S power net, between Neutral and Ground, there will always be a low voltage present, this voltage rise the further away you are from the sub station, and during day when there is more loading of the net in the neighborhood. Here in Belgium at my home, this voltage is useally between 0,5V to 2V (the times that i have measured it) but i have read articles, that with a TN-S net, even up to 5V is possible between Neutral and Ground.
Thank you for your great feedback!
@@KissAnalog , You're welcome. grtz
I know exactly what you mean about enjoying Paul McGowan's videos while cringing at some of his claims and technical explanations. His AC regenerator claims drive me nuts as well. It's outrageously expensive and does nothing for you that you can't get by using a simple filter from a handful of passive components. Also, it uses power which is another waste. He has another "power cleaning" product that's full of nonsense claims and pseudoscience.
Thank you for saying this so great! It is like you want to like him and his company - but then you he gives so many opportunities of total cringe;)
@@KissAnalog I drank the PS Audio Kool-Aid for a few years too. But their DirectStream Sr. DAC is the best put there to my ears and Wallet. Yes, the AC Regenerators make me Cringe too and NEVER bought into them. I've used the Stellar SGCD DAC/Preamp and M700/S300 Amplifiers and they are great sounding products. Hypex Switching Supplies with their audio section ripped out and the beautiful sound ing PS Audio (Darren Myers) designed jfet input stage feeding the Class D Gain Block of the Hypex.
Paul's book 99% is quite the interesting read if you ever get the chance. Never met the man but toured their factory and sound rooms. Great place with great people, but some of his videos are highly educational and others do make me cringe.
Thanks for the great feedback! I agree with you but the quality products, so I does make me wonder why you'd mix snake oil in when all it does it makes people wonder... I'll have to look into his book - I didn't know he had one;)
@@KissAnalog 99% True. It’s a hilarious and sometimes dark book. How he lost his company and then regained it is incredible. Lots of Pot stories and Army stories too !
@@joeygsaudiochannel3972 -- The cordial and affable Mr. McGowan is not a formally credentialed engineer in any discipline. He's a self-taught audio enthusiast who's picked up some technical knowledge of the course of a long career in the industry, but he's basically more of an entrepreneur and marketing guy.
I recently bought an Aiyima A07 (3255) and a simple Mean Well 48V 9.5A industrial supply and a pair of new Q Acoustics 3030i speakers, holy crap it plays beautifully!
Wow - thanks for the great feedback!!
@@KissAnalog LOVE your channel, you're great at explaining things and good fun to listen to! So a very big thanks for inspiration and education!
Got a new better ventilated chassis for both PSU and amp with some components for modding the amp on the way, upgraded power inductors, caps and a log 24 step attenuator rather than the linear pot that came with it. So trying to improve where i find they messed up a little :)
It's a really fun project!
Thanks Eddie, great chat!
Audio is voodoo. You have to run an amp for 100 hours before it works right? Can you imagine if Keysight sent out a meter with a tube rectifier and then said you have to run it in for 100 hours before it works right!
This is how I look at it, everything is an instrument, doesn't matter whether it's a scope or an audio amp, the only difference is the end function. It doesn't matter to what I'm designing.
Thanks for your great feedback! I agree with you. I think of the oscilloscopes that have very high bandwidth and 24 bit ADCs - and they all use switching power supplies. Audio companies have small design teams and just simply don't pay well. They also just don't have the best engineering - but don't tell the audiophiles that;) I am speaking in general terms - there can be the outlier...
I know a engineer who designs the consoles used in the music recording studio and the consoles use switch mode PSU's with FPGA devices with advanced DSP techniques. So the equipment used in the studio is now using this technology, many of the old analogue being reserved to specialist work.
Thanks John for the great feedback!
Excellent rant, feel free to do so again whenever you have something to rant about.
I'm also the type of guy that doesn't do things a certain way just because, today it is rare that a linear power supply makes sense, you certainly can't beat their simplicity, but when power goes past a certain point linear becomes inconvenient and expensive.
Thanks so much! I appreciate you!
I agree with you. Linear is great at low power.
Love your channel! I own an integrated Class D amp and am beyond happy with it. One of my main considerations in favor of Class D was simply… much higher efficiency (= less energy consumption) and consequently less heat generational . Plus, the sound was actually better than a highly rated A/B amp I compared to (via listening session). Obviously, acoustic flavors vary and depend on individuals’ systems an other components. But looking forward to eventually add a tube pre, as to get -arguably- the best of both worlds for some of my listening.
If the class D sounds better than A/B (which it doesn't) then why would you need to be looking for "the best of both worlds" you must want to get rid of the efficiency.
@@ranbymonkeys2384 There are sports cars, family sedans, SUVs, vans, pick-up trucks, gasoline, electric, etc. None of them are ‘the best’ without throwing them into their intended environment they’ve been designed for, i.e. roads, tracks, off-road terrain, inner city and so forth. The same goes for audio equipment. We don’t listen to the same music or genres all the time either, so sometimes it might be nice to have -as for example in my case- the holographic sound and warmth of the tubes for certain dedicated listening, hence my ‘best of both worlds’ statement. You might disagree.
@@sveneisenhauer1 Pre amps always make the power amp better in most cases sound better. So let me ask you, if you are riding your dirt bike or cruising in the towncar, is that pre amp going to be used in both situations? Your turn...
@@ranbymonkeys2384 I don’t have a dirt bike OR towncar. You tell me. 😜
Thanks so much Sven! I appreciate you! I think more black box testing would show people how great Class D is. Just like in any class of amp - there’s good and bad;)
Some years back my brother built one of those boxes to use an audio signal to create the 120v power. I think it was more of a novelty than anything. It was fun to use it as a VFD to play around with AC motors, and to play music into incandescent lights, but there really weren’t any practical applications.
Up until now, I didn’t realize that people actually used something like that as a method to create “clean” AC power.
Thanks Chris! I have a power generator in my storage I think;) Could be great as an AC power supply. Definitely not needed in a hifi system;)
I agree with you that if you are going to be a company that design's your own equipment it is essential that you have a good power supply engineer or at least have a power supply consultant on call to help out when problems arise because eventually, they will arise. The company I worked for had a consultant his name was Sam B., and he would come in and he would take a few measurements and give his recommendations and sure enough the problem was resolved.
Thanks for this great feedback! I was once referred to as the 'hired hand' the guy they sent to solve the difficult problems - that needed a solution once and for all. ;) One manager said that I was the tool in the box that they reached for when no other tool worked;) I miss working for large companies that have lots of programs and need help once in a while. Unfortunately, a lot of companies can not afford that guy full time:(
Again another great video, thank you.
Is it true that SMPS returns a lot of noise to the line and thus disrupts the work of other devices, as "audiophiles" claim and if so, what can be done to prevent it?
I love this question - thank you!! First, as I’ve shown in videos just prior to this, the amplifier as a thing called PSRR (Power Supply Rejection Ratio). This is where the amplifier shows its immunity to noise at the input. So, every device has an input EMI filter which both protects from the device putting noise into the system and protects from noise coming into the product it powers;) So, I call BS on this;)
Sorry for late reply but from a helicopter perspective it works like this; devices sold to be used on the power grid has to be compliant, there are (in Europe) basically 3 categories.
1. Industrial use, Home use, medical/life saving use. Each class has different limits as to how immune they must be and how much crap they are "legally" allowed to feed back to the grid.
So industrial devices has to be able to withstand a metric crap-ton of line noise and voltage surges and magnificent short circuits etc. but they are also allowed a little relaxed rules on what they feed back. (Testing against extreme standards is exciting, quite often things go 'pop' as photonicinduction says)
Home use are mid/mid on immunity and emission. Medical devices has the strictest rules; They have to be high on immunity tolerance and low on emissions. I've never plowed down into UL(American/Canada use) as I've been working in Europe only, so I'm somewhat familiar with EN/DIN (resposible for european norms and German industrial norms: both of them are largely homogenized today due to EU and such)
That was world record stuff Ed. I loved thr bathroom-related camera cuts hajahahahha
LOL thanks Joey! I can always count on you to make me smile;)
@@KissAnalog I'm not wrong is the thing lol
@@KissAnalog I still want to know if you own a soldering iron. Your whole adoring public actually lol
LOL, I might have shown those skill during my Class A 20Watt build;)
I have a custom 2.1 class D amplifier with a switching power supply, the induced noise indicators are excellent, I think complaints about switching power supplies can be justified if the release is poor
Thanks for your feedback! Yes Class D amplifiers powered by switching power supplies are taking over. It has already started:)
What was the fried egg problem on the ISS power supply?
Yeah! I want to know.
I really enjoyed this video! I worked in high end audio before going back to school to get my EE degree. I used to think that you shouldnt use switchers because it would be difficult to keep up with the switching speed of the amplifier (Needing higher frequency to keep from drooping "Cycle by cycle" vs the power amp switching frequency and also to keep the frequencies from beating against each other) I now see that you can use multiphase switchers in order to increase the frequency much higher in the power supply to overcome these problems. Do switchers have a shorter life span because electrolytics might fail at higher frequencies faster than at low frequencies (Not saying they do or dont - I am asking the question)? I would love to see a video about longevity between linear and switch mode power supplies.
concerning the longevity of switchers, consider how the company SeaSonic makes computer power supplies with a 10 to 12 years warranty. obviously making electrolydic capacitors inside the power supply last so long is probably not easy nor cheap, but it can be done. apparently.
Thanks for the great subject! I’ll do a video on expected life. Switchers are used in aircraft as they generally do have a much longer life. Partly do to using fewer aluminum electrolytic capacitors.
What are the spec’s/requirements for the SMPS for Power Amplifiers?
Is there difference for the SMPS for servo drive and audio?
Great questions! I think an SMPS for a power amplifier is like many others - maybe a little more filtering.
@ SMPS for a computer or LED lightning is different than for AF and power amplifier.
Yes they are different but many similarities;) Mostly feedback is different.
@@KissAnalogwhat makes the Hypex SMPS (Bruno Putzeys) so good?
How do the SMPS from Micro Media compare to the Hypex and lets say Mean Well?
I've just learned today that blackstar the guitar amp company is doing some new lightweight tube amps with SMPS. Very interesting stuff!
That's great info - thank you!
Nice M-T-M monitor in the background. Yup! there is nothing wrong about switching power supplies. There isn't more distortion from them. Period. But in Audio, the audiophiles are cousins of the Flat Earthers. I'm an EE. So I know about switching power supplies, and electronic design. Recall, Technics put out a high quality mini system with a streamlined class A/B amplifier powered by a switching power supply. It put out a very clean 50W/ch. Ran cool, very low in noise and distortion, and simply sounded great. But it didn't weigh a ton, and wasn't massive. So it was rejected like a red-headed stepchild by audio purists. Too bad. Digital amplification came out in 1998 from Texas Instruments. But thanks to stubbornness and ignorance, Class D amplifiers didn't take off until 20 years later. I'm glad luddites are dying off.
Thanks for that great feedback! I think you are so right about the Flat Earthers;) Maybe we can make a little noise on this channel and get people to think;)
What I don't understand is why the smaller niche audio folks just don't use an off the shelf switching power supply. They don't necessarily have to hire an expensive switching power supply engineer! Not sold on Class D yet. But have been a switcher advocate since the mid 90's.
Great question! It really isn't easy to find an off the shelf power supply wiith the right voltages and current ratings. Also, there are 2 cost considerations, the NRE (non re-occuring engineering) which is what you pay for the engineering development, and then the re-occuring engineering cost for every product that is built. Buying a power supply for each unit you build is expensive compared to the raw cost of parts that it cost to make your own.
The Amplifiers and other equipment from Linn Products in Scotland use SMPS. They’ve been using SMPS since the 1990’s and continued to improve since then, The latest version called Utopik is an improvement on their previous ‘Dynamik’ SMPS. Their power amps too also use SMPS including their latest 800W Monoblock
Wow - thank you for this feedback! I'll have to look on the used market to see what I can find;)
the icepoweraudio ASX series look interesting. it's an integration of a high power class D amplifer and SMPS in a single pcb module. Also, there's Quilter Amplifiers for guitar. Class D and SMPS and some circuitry to emulate tube amp response, like applying current feedback to give lower damping factor. i'm working on a guitar amp project now using a class D power section and tube preamp and SMPS for both power amp and high voltage for preamp.
Wow thanks for the great feedback Larry!
Nice vid Ed!!!!
I would love to see a regular rant every couple weeks. I work for a audio company as analog HW eng and I do get what you said.
cheers!
Thanks so much for the feedback! I appreciate you!
What about output current of amplifiers with SMPS vs linear? I've noticed that almost every Class D amp has a max output current of about 30 A, Class AB have no problem with 45, 60, 90, even 180 A for something like the Parasound JC1+ monoblocks. When you're trying to control big speakers with lots of drivers that current can come in handy.
Thanks for the great question! I'll use the Parasound as an example of how those specs (peak current of 135A) is not possible. It is more marketing BS. Pro sound uses switching power supplies on concert stages driving those big speakers without a problem. There's so much misinformation.
i'm using an old sony integrated amplifier with one of their "pulse power supplies", a ta-ax6. i have several other models with them too. sony really wanted to do that back in the late seventies. i wonder how successful they were with it? they seemed to go back to linear for the eighties.
Wow they were way ahead of their time! But then they did make a lot of leading tech back then;) Thanks for sharing that!
Reliability. SMPS are not reliable at all. Trust me on this. Haha
interesting. i've had this 40 year old ta-ax6 for probably a decade and it seems to be doing well. i've actually got two of them. and a ta-f70. and a ta-f45 and 55. and a ta-n86b. they all have the sony pulse power supply, are all over 40 years old, and all work, aside from some dirty switches and blown bulbs. that seems pretty reliable. @@dongwarrenmusic
The power supply noise rejection ratio of many circuits is high at low frequency, and low at high frequency. See any OP amp date sheet.
Thanks - so true! That's why switching power supplies can work so well with Audio;)
Great rant!
Of course, nowadays you can build a world-class hifi amp out of professionally-designed OEM modules for both SMPS and Class D output stage, leaving one major marketing challenge: product differentiation via cosmetics or a supposedly "audiophile" input buffer designed in-house. This is yet another reason why many of the struggling little "high-end" audio companies cling to obsolescent technology like linear PSUs and Class AB -- or even Class A -- output stages.
Thanks Bruce! Well said! I plan to design a Class D with SMPS power supply that will have a sweet front end on the amp and the PS will have an active power factor correction front end;)
@@KissAnalog -- Great project, Eddie. Looking forward to learning what's involved in designing such a sophisticated SMPS. Are you going to buy a power amp module or roll your own?
Disclaimer: I am NOT an EE! For high current amplifiers, a [large] transformer has something a switching PSU does not have: lots of iron and the "flywheel effect'. Same idea behind low-frequency inverters w/ large transformers -vs- cheapeer high-frequency inverters. It is my understanding that those older style inverters (with large heavy transformers) can spin up dificult (inductive) loads with ease - while the newer high-frequency ones always struggle. The electro-magnetic force stored up in large chunks of iron act as a large flywheel and seem to deliver more "punch" during large instantaneous demands. I am sure modern Switch-Mode PSU's have ways of getting around this - but that has always been my understanding.
Thanks for the great question! Yes you are right about the large iron power supplies. The problem with the switchers is that they have a much better control. So, when an over current event happens - they protect quickly. Problem is - they don't want to over load themselves. Solution - get a inverter rated for your surge current and don't blame the switcher for doing what it is supposed to do. So, rule of thumb - most of these cases can be solved by buying an inverter with - maybe twice the nominal power - or buying one that is designed to handle large surge currents. The new switchers will be better in every way if they are sized appropriately;)
Not sure about the title but amplifier brands do use switching power supplies. Depends on the brand, model..etc and not just class D amps that use switching supplies. Its more popular in pro amps to use switching supplies but it happens otherwise too - just not as common to linear supplies.
Thanks! Yes SMPS are found in pro gear but Audiophiles resist or misunderstand them. Class D amps and SMPS are starting to make a strong appearance.
Liked the rant and your discussion on power supplies vs. switching supplies. I also liked your mention of Amir at Audio Science Review. I've been following Amir for sometime now and like his no b.s. reviews.
Thanks Walter! Amir is really great;)
Audio amps do use switching supplies. I have several and they are all noiseless and distortionless. And do remember that so-called linear PSUs are in fact switchers. They just switch at 100/120Hz instead of a few hundred kHz. But they work exactly the same way - dumping charge into a capacitor from a variable mark/space ratio source.
Oh boy you are singing my song;) I hope you become a subscriber and that I hear from you more;) I appreciate you!!
Amen brother great video. I think the answer to the question is snake oil has a large markup.
LOL I think you are so right David! Thank you!
The first switching supplies were introduced in the high end audio field by Meridian with their MCA and Chord with their 1000 model. I was lucky enough to see the prototypes at the time in the early 1990s. We were cynical that these would perform but they tested immaculately. The Chord designer/owner was a power supply designer for the Ministry of Defence in the UK working on Tornado aircraft amongst other things. Boothroyd/Stuart were amazing designers. The best audio designers were all engineers and all the great advances came from science. Bose springs to mind in the modern day who went off at a tangent - But all pure science.
The greatest engineers historically in audio were Blumlein and Armstrong who between them invented almost everything that laid the foundation of the industry.
No pseudo science has any place with the great designers in audio.
Thanks for this amazing feedback! Really cool information! I’d love to see one of those Meridian units on my bench;) Pseudo science has no place for real engineers;)
@@KissAnalog Glad I found your channel. Subscribed.
Enjoyed your "rant" and really like your workbench! I worked in the defense industry as well and relate to a lot of what your presented.
Thanks so much! I appreciate your feedback;)
Does a class D audio amplifier just have a switching power supply to create the DC power rails, or does it also use switching to produce the audio output amplification?
Yes, the Class D is made with switching - see this video I did to try and explain: How Class D Amplifiers Work part 1 with EPC GaN FETs
ruclips.net/video/GCA3t67yOLc/видео.html
@@KissAnalog I just watched that video - that answered my question perfectly! Really good with the whole setup and scope of the input vs output. Thanks!
Thanks for the feedback! i appreciate you!
Back in the early 80's switching power supplies were just making there appearance. We were using them in Jerrold CATV amps. August was lightning season and we would replace 15 or 20 in a night. We started noticing that as the linear power supplies were the only ones getting changed.
Thanks for the great feedback! Very interesting!
IDK if you remember BASH amplifiers in the 90's, but they used switching power supplies and were really compact / high output, and very clean sounding. That company was short lived, because soon after, everyone went to cheap chinese stuff.
Thanks for that info! I'll look them up. It is too bad that a lot of small companies are so fragile:(
@@KissAnalog ELAC makes a BASH amp right now. There is some technical information on the setup if you goto the ELAC website. The company in Canada was Indigo who had ST micro electronics fab up the IC's for B.A.S.H. I forget the abbreviation for the technical term bash. It was over 20 years ago so the company went under, and the IC's were still protected with a patent until recently. Looks like ELAC had ST microelectronics re fire up the IC's since the patent expired.
I really liked this video. To be honest I even wouldn't call it a rant. This was informed background info from a professional. Always something to be learned here. Thanks Eddie.
Thanks so much Marcel! I appreciate you!
You are right sir
Thanks for this video
I watched your full video
I watched pal audio video
I watched audio science video
Thanks for putting all in one video
Thanks so much!
Well Manley didn't just hire "some" power supply guy. Bruno Putzeys might be one of the most influential Class-D technology audio engineers on the planet in IMO.
You are right - they paid the money to get a great power guy;) Not many audio companies do that.
I have seen similar company behavior myself in the past. They stick to what they know best (and have been able to charge to customers successfully for ). But then often the market shifts under their feet. Some of those company do not survive.
Were the Carver's "Magnetic field Power Supply" some kind of switching power supply? Yamaha lost a patent suite when they were copying the idea (e.g. with the A-760). Were those somehow SMPS ?
More important : any advise on where to start if we were to try SMPS on our diy audio gears ? Some hints would make a great video !
Thank you for your content !!! Always instructive.
I think you are right - many companies have not made it. I'll have to go over the Carver design in a video. I think it was very interesting. I appreciate you!
I own two REVOX B251 amplifiers built in the early eighties with switching mode power supplies. I always found it was sounding weak and dull thinking the SMPS was the problem. Now you make me wonder if it's because of SMPS design instead... I love so much my REVOX stack and don't want to get rid of the B251. Here I go again...
LOL Thanks Wally for the great feedback!
A couple of years ago I started messing around in audio by replacing electrolytic caps with film caps, stuff like that. Sometimes I can improve the sound and other times I mess the whole thing up. Well, I have to be soldering something or I start shaking and go through withdrawals. How do I get into a job where I can do something like that?
Plumber with copper?
But I know your feeling
@@JasonW. I was actually a welder fabricator for 14 years maybe that's where it comes from, electronics I think is more personal, not to mention cleaner.
I generally agree with your take in reference to Amir's (Audio Science Review) criticism of PS Audio's device - - the thing is, at least for me, and yes I'll say up front that I enjoy Paul and yes he is an engineer and has been responsible for many audio design improvements over the years and admire Amir as well - but the thing is, clearly Paul's device provides a better sine wave and the primary thing I took from Amir is that the noise floor actually increases from that of an unprocessed outlet - and, rightfully so, Amir was also critical about the marketing aspect of how PS Audio presents their product - that their marketing doesn't live up to what Amir measured including the implications of impendences - however, whether or not the device actually has an impact on sound is never covered by either - Amir dismisses it out of hand telling us it's a waste of money but doesn't give it a sound test and neither did Paul in his follow up to Amir's review - - all too bad - - imho, if the impedance hit isn't great enough to impact the gear being used, obviously the noise floor, although worse, is still low enough to have a minimal impact which leaves me wondering whether the more perfect sine wave is indeed a benefit - Amir claims it doesn't matter in that a well designed power supply can overcome a less than ideal sine wave - - so, I get his science, but where's the listening tests? Why don't we get both measurements and listening test? J. Atkinson has been testing for years and years with equipment similar to Amir's - I would bet that John would still imply that measurements don't always correlate with what's presented as a final musical presentation - that's the catch for me - I like and want measurements but, to this very day that doesn't mean one can "always or necessarily" predict sound from specifications.
btw - I'm running class D and haven't looked back - of course I freely admit that there's plusses and minuses to all three approaches (Class A, Call A/B, Class D) and yeah - I don't have experience with some of the other exotic approaches like Bob Caver has marketed.
Thanks for the great feedback Skip! I appreciate you! I'd like to come up with more meaningful ways to characterize an amplifier.
I developed semiconductors for the audio industry, and it is indeed full of pseudo science. In fact, we used to joke that some customers would like AC coupled inputs because it gave them a chance to change from red caps to blue caps and "improve the soundstage", while another customer would do the opposite. We used big Agilent SMPS lab supplies to evaluate our power amplifier chips, by the way.
Thanks frammis! LOL that is really funny!
Lol, I believe it. I have to say, the best sounding amps I've heard to date were designed around VFETs (Static Induction Transistors). Could always just be a coincidence.
An ideal power supply should have no noise, plus the output impedance should be 0 ohms throughout the frequency range and sudden loads should not cause ringing, small slow variations in voltage are for most amplifiers unimportant.
I'm missing Data for SMPS output impedance , I can't find this data either at Hypex or Cornex, do you have a link to such data.
Great feedback and questions. You are correct that no noise would be ideal - but doesn't exist - with any power supply. Some think that switchers are more noisy - but their noise is in the high frequency range that is out of the audio range. So called linears (which are really slow frequency switchers as few are actually 'linearly' regulated have noise in the audio range (60Hz or 50Hz which also results in 120Hz and 100Hz).
You are also correct that output impedance ideally would be zero ohms, but no power supply is zero ohms. Actually the low frequency switchers (not linearly regulated) have higher output impedance than switching power supplies. Switching power supplies are regulated so will show much lower impedance than using just bulk capacitors off of rectifier diodes.
Linear can be understood by nearly anyone. Easy to design, fix and upgrade. They looks good and can be cheaper. Not always better. If space is limited switch mode can be more dynamic.
Thanks for your feedback! I agree with you;)
I had a Fender 100 watt GTX with ICE power class D. Very low heat, almost no heat sink. Was impressed. BUT, the class D power section did not sound as good as my MOSFET amp. I would like to build a MOSFET and try the switch mode power supply. I remember TI was working on a good Class D amp about 12 years ago, the engineers said they wearing quit there yet, or even close.
Which brings me to your video, let's sell it even though it's really not ready, lol. No class D for me, yet.
Love the rant, keep it coming. It will great to hear some of your troubleshooting GOTCHA stories.
Thank you! I'll have to do this;)
and, I could learn a lot from you too. thx for this video. I realy had to hear other opinion about wich PS is better. I think that if smps's would work on higher frequency (more than 100kHz, maybe MHz...ishh) and garantee big spikes, thay would be great option for power amps forever. I don't have almost no knowledge in this area but I ask myself, how can a smps do what a big ass cap can do? and realy still get a big efficincy...
Great question! I'll do a video very soon to explain how the switcher can replace the large caps;)
The customer base for high end audio products is conditioned to accept subjectivity. As long as they spend their money, don't rock the boat. There are renegades, but the market - opinionated and willing to spend money - demands what it demands. We need more renegades.
Nice video
I wonder if smps have reserve power like those power supplys based on very large capacitors,around 20,000 mf and up which act like powe reserve bank that can deliver high current when needed, like music passages of kick of a drum, combined with tough loads that can go low in impedance (some loudspeakers can go down below 3 ohm in certain frequencies), it's not all about continuous power, it's also about transient power, can they deliver?
Thanks for the great questions! So one thing to think about, the so called linear power supplies (they are slow freq switchers - unless they actually have a regulated output) are very slow, and the switchers are 100x faster so the caps can be 100x smaller as they can react so much faster;)
@@KissAnalog can it be simulated and measured in the lab these situations ,maybe making square wave load and to see on oscilloscope or other method the behavior of the output?
Yes this can be simulated;)
@@KissAnalog can you make a video about this topic?
For the first time in a long time, I didn't feel cheated listening to a long video. Yeah, it's possible to make reliable switcher, but few can do it. That's why, often, military surplus is the best there is. I don't know too many audio companies with an excess of development funds, either.
That’s fantastic! Thank you, and I agree with you!
Man I would love to replace the typical input transformer for a hifi tube amplifier. That would be such a great package
That would be interesting;)
What about a valve rectified power supply compared to switching and linar PS?
I do have an MC30 so that would be fun to do. Thank you!
it's alway a trade of between quality and margins, a lot of electronics and other things are of so so quality and happen to break fast which is in there interest, some things are gettig way to hot because it is tightly dimensioned.
That’s so true! Just because a circuit is efficient, there still needs to be spacing to spread what heat there is. Linear might seem to be more robust because of the fewer parts, but the large aluminum caps will wear out as they outgas from the day they are built. That’s why there is such a thing as ‘recapping’ where all the aluminum caps are replaced after time.
Yes, all my commercially made HiFi equipment are now all powered by SMPS, I was previously a linear high speed feedback based regulator fan but times have changed.
Thanks Michael! Yes, you are right, times are changing;) SMPS are taking over - and probably for good reason;)
@@KissAnalog You should explain why this happening and explain the reasons. Cost and performance are reasons, maybe explain the aspect of costs, as well as performance and also the efficiency and size advantages and also the Before and Now solutions and relative performances, the FET and its appearance and also ICs. ASR does the performance but never explains how it is done. Wit ha regulated, the extra overhead of dissispation , cost and energy is removed, I'd like to know more about the different methods of SMPS and the Pro Con in Audio.
Thanks, I do have videos comparing linear PS to SMPS. But I have more to come:) I like your suggestions and I’d be happy to hear more;)
A lineair power supply is also much more basic it's reliable , switching power supply's are sensitive to ac interuptions (by the power company) and whenthe ac goes off and on fast in a row then switched mode supply's wil tend to break, what's you're opinion about this? and is it true, or is the PSU bad designed then?
Thanks for the great questions Ronald. Actually a linear with the huge caps is a bigger problem with power interrupts. The NTC typically used to slow the inrush current is hot and then can’t slow the inrush with a quick power cycle. Also, some use relays to short the NTC after the inrush - and these have so many cycles. A Switching power supply designed correctly should not care about power cycles or brown outs. Military uses switching power supplies for the reliability.
The problem you describe is common. It's basically caused by people not going for the company goal but individual goals (which in the end they'll always do), which haven't been aligned.
The reason is simply a lack of communication.
Thanks for your feedback!
Thank god Manley went to a switch mode power supply!
Yes - they are very brave and very smart!!
Could it be because of the waight?, when you buy a amplifier that's for instance 500 watt, and it's very light , maybe people think does this hting do the 500 per channel ?, because the weight is because of the transformer.
Thanks Ronald, I think there’s truth to that. Also the big transformer and large caps look impressive when they show pictures of the inside of the box;)
Hey, while I've got your attention...if I got your attention... How about a rant on how software/digital has moved into the analog realm? By this I mean...that switching power supply controller chip might actually have some software inside of it. Right? Tell us how you feel about that! Us embedded software guys are very interested! In any case I love your videos!
Thanks for the great lead. I like the idea!
Wow. I really appreciate you sharing your experiences and knowledge. Thank you.
Thanks so much! I appreciate you!
I constructed a power amplifier using tda7294 ics, and a 500 watt smps. The combo sounds fantastic, and dead quiet on idle. This coming from a guy with a linear ps fetish. I think the days of linear supplies are numbered.
Thanks Bill! Great info! I'd have to agree - SMPS is the way of the future;)
I’m sure that there is a mentality that “If it weighs more then it must sound better...”
😬 I suprised even myself as I actually said "WHOA!" out loud upon hearing "the relay spechul". That's a no-no ideed!!
Thanks very much!
Manley audio products are bloody amazing! #WANT! I love her approach and product designs.
I am in full agreement;)
Something else Eddy Another thing, don't tell elon musk that the Clean Air Solution for the Electronic Car is not as clean as suggested because the energy from the power station is miles of wires to several transformers for his electric car and the power plant runs on coal. I wonder what percentage only comes from our sockets
LOL Bert;) I think those that care will get their power from the sun;)
Audiophiles hear stuff the instruments don't, it's magic. Or snake oil. I trust measurements, as such audiosciencereview is great. But I also understand that people like the character of a certain kind of sound, which is why we have vinyl and tube amp fans. In my opinion there's no shame in liking the character so if you're into it, go for it. Nowadays we just don't have to buy into the snake oil anymore, which is great.
Thank you! I totally agree with you! We should be able to characterize the sound with measurements. If someone wants a certain sound - it would be nice to say - oh look for this in the spec.
yes I wasn't sure about that ps audio regenerator. you can do the same with a site transformer and some motor run caps, transformers on the primary and secondary of 5uF one thing I wasn't sure is how many run capacitors on each section? what would be plenty 60uF or 120uF divided between the secondary and primary? got any feelings on that!
and I haven't noticed but why don't they use split film capacitors on switch mode power supplies, surely this would be better approach! and the ideal capacitor.
I know one company that would appreciate your talents and that's bang and Olufsen and ice have you tried?
why can't you sell your own switch mode power supply for valve gear and solid-state on the open market isn't that better?? and companies will try it and realise not just enthusiasts.
Thanks for this great feedback!! The regenerator is completely unnecessary! It is almost a hoax - IMHO. Offer amps with better designed power supplies if you really believe this is an issue.
That is a great idea! I should give B&O a shot. But, I really don't want to move so I'd want to work remote or travel often
One thing I am doing is showing how to design these on my RUclips channel;) I have thought about offering these for sale as you suggest, and maybe I will. I'm an engineer and know that I'm not good at business, but maybe I should give it a try. I'm also bothered a little that I'm teaching these companies how to do it for free. I like the idea of teaching the diy community and those that want to be engineers. But I hate to give away to these companies for free that they wont pay for.
For world wide product distribution (sometimes called a universal supply) the only real viable power supply topology is a switching regulated supply. Add to this the feature of PFC and the power supply could have multiple switching controllers. If you look at the possible EU requirements for PFC and the wide range of line voltage and frequency world wide, with the added challenge of accommodating high and low line, and testing per IEC61000-4-11 / 4-14 / 4-28 which targets AC power line disruptions, the PFC + switching regulated supply is the only real option.
But, as stated in the video, the design goals are more of a challenge due to component stress and qualification. For example, in looking at high speed diode turn on overshoot (a type of stress) the test oscilloscope needed should have a 1GHz bandwidth (to actually see the rise time characteristic) which is not only expensive, but will require the test fixture be well designed and can maintain high measurement fidelity (also not easy). Add to this the detailed study of virtually all of the passive components (also needed for high quality models for use with circuit simulation), and equally challenging, to effectively design the PCB (which should include testing for pre-compliance (conducted and radiated EMC)), and you can see the task is not easy or inexpensive.
Thanks for a very interesting presentation,
we shouldn’t forget that the audio industry is partially in the sales of fantasies and despite all the advantages of SMPS the linear ones look much sexier with their large toroidal transformers and tons of electrolytic capacitors, I like to show off with the 2 large toroidal xformers and 1/4 Farad battery of electrolytic capacitors in my 22 kilos diy Pass F6 and it gives me the impression that it makes the sound better😅
Looking at another fantasy industry: who is willing to swop the 6 liters V8 (360 cubic inch for my fellow Americans) engine in his muscle car for a much smaller turbo charged 6 inline that weighs half, runs on 20% less fuel and give it’s full torque from very low RPM? The V8 is much more impressive 😊
Thanks David! You make an excellent point! ;)
you are right, big heavy, transformers in liniar power supplies are cool. its kinda like driving a sports car, that cosumes more fuel just to be more cool. but i have to admit that they are just worse and impractical, especialy for high power
Yes, I think you are right;)
I have one of these Sunfire TGA-7400 - From the Manual - The big breakthrough is the uncanny tracking downconverter, which uses 18 Herculean Mosfets in the TGA-7400 and the TGA-7200. This allows the amplifi er to produce phenominal power output, without the heat generated by conventional designs
I very much liked this video, well-made and told! :-)
I think of the SMPS a bit like I look at the CD-players in the start, nobody knew how to make it sound good, but got it to work in the end. You can, as you say, add to that their bashing of the SMPS because of them been badly designed and everybody "knows" how much they mean for an amplifier.
I may have a suggestion for a new rant: if you ask a non audio-interested electrical engineer, does every amplifier sound the same, you can take a look at the spec and determine that none of it is audible. Therefore, does a $200 SMPS amplifier sound exactly the same as an $50,000 class A amplifier, and everything you may imagine hearing in différance, is, yes, due to you thinking thee is a difference.
Is that true, sort of true, or simply wrong?
Thanks so much Fried Mule;) I think there's so much mis-information and lack of testing methods to truly characterize an amplifier - that we need to start working out a way to do that now.
Hi Eddie, Awesome rant, good job. Thank you
I appreciate you!
An unregulated linear supply is efficient enough and will last much longer than any switching power supply. It is no fun when one rail fails and you blow your speakers. They are also much easier to design and build.
Thanks Florian! Actually, if designed well, a SMPS will outlast any Low Frequency Switcher (not linear unless it is regulated). The aluminum electrolytics have a life span on them but switchers can be made without them. The International Space Station and all other SPACE born electronics are switch-mode and have to last 20 years with zero failures. You have probably heard of capacitor reconditioning or re-capping? This is because aluminum caps dry out. You are correct - SMPS do take some expertise;)
Jeff Roland makes A/B amps with Switching power supplies but their stuff is too dark. Mola Mola uses switching power supplies on all their audio products from Dac's, Preamps and amps but their stuff is analytical and their parts are a little cheap to a lot cheap. The problem is everyone has an "opinion" and aren't willing to try new things. They still love their tube amps or big Boulder sized 150 grand boulder amps. The best things I ever heard were Pilium Amps. I don't know how they do it but omg. This greek company is amazing. One of their amps is 30-50k. The guy that started it used to make treadmills of all things. So at least he was a real engineer. They use multiple Torrids that are shielded in little round cylinders. It is more of the quality of the power supply not the design and most "elite" companies use junk parts even if they are expensive.
Thanks for all that great feedback!! It is great when the management hires great engineers and lets them use the parts that they need that make the difference! I think a lot of these companies run with an amp engineer that also tries to do power supply design - because its easy - right;)
@@KissAnalog YEP even a lot of "upgraded" linear power supplies are junk! take a look at the one that is $1200 for a headphone amp from Ferrum Audio it's all junk parts! They charge you extra for it and put it in a pretty box. a pretty box always makes your external linear power supply better!
I would love to know the story around 19:00 in full details.
Thanks for asking. I could probably elaborate a bit more on this story, but I can't say names as this engineer had to sign a paper that he wouldn't say certain things about this company. Sad, but that's how companies protect themselves after damaging someone that they don't want others to know about.
@@KissAnalog Thanks for the reply. I was thinking the technical side when I commented. Maybe a few examples of cases where it's better to do it your friend's way. And a few good practices would be great too. Thanks.
Yes, love the rant idea! It's funny that I follow the same youtubers where audio is concerned anyway. I have been watching for some years now, trying to wade though BS and snakeoil. Also trying to learn as I go. eg I decided on the (Chi-Fi) Topping D50s DAC I have after following ASR, I'm sure there is always something better, but at the price it cost, 2nd hand, I am not disappointing! It was as much an experiment ie, do I really need this, will it make as much difference as "THEY" say? My opinion, is it very much depends on how you listen. I like to spend an hour or so just relaxing in the dark, or at least dimly lit room, sat in the sweet spot listening to some of my favourite music ( whatever it is on the day). In this scenario it made a good difference. The other scenario, where the music is on, but you are walking about doing housework, cooking, whatever, a lot of these add-ons or improvements they want you to buy seem to make no damn difference other than emptying the wallet. You answered a lot of questions I had more diplomatically than I could ask them lol. Thanks Eddie.
Thanks for the great feedback! I appreciate you!
sir can you make a video on smps and simple power supply for amplifier which one is best
I believe that Benchmark AHB2 power amp is using SMPS inside. It’s lite and also not even class D (it’s based on THX AAA feed forward, i.e combination of class A + class AB design)
TBH i dnt know why PSU arent regarded as the most important thing in an amplifier rather than the signal chain. thats what were actuallt modulating. 3 months learning to make my own tube amps made me realize this.
Thanks so much! I think the big reason is cost - and audiophile would rather spend mucho dollars on cables than ask for a properly designed audio amplifier;) The industry has done a great snow job!
Just over 25 years ago I remember listening to my first D class amp and it sounded horrible switch mode power supplies and D Class amps technology has come a long way and now are better than AB class amps because they are more reliable and don’t dissapate heat and they are more efficient and are friendly to the back.
This is a briliant video! Congrats and keep up the good work.
Thanks so much!
great to hear your stories. trying to find my way in my late start of work life. i am learning as i go and don't want to be someone that believes in falsehoods. didn't know someone as knowledgeable as you would be so underappreciated! working in a work place with others that want to also show what they know is hard! how to go about it, and wanting to show what i know too while not telling people.
Thank you! I appreciate you!
Couldn't these amplifier manufactures just use proven off-the-shelf SMPS and tie them in to their design?
Thanks for the great question! At times they could - but even then they want to have their name on the stuff inside their box so they can charge what they want;) Also, it is usually less expensive in parts to make your own, so buying one it would cost too much in parts.