Switching Power Supplies in AV Receivers

Discussion in 'GENERAL AV Discussions' started by Verdinut, Apr 8, 2017.

  1. gzubeck Audioholic

    gzubeck
    Joined:
    Jan 8, 2017
    Messages:
    100
    Likes Received:
    1
  2. Johnny2Bad Full Audioholic

    Johnny2Bad
    Joined:
    Jan 16, 2010
    Messages:
    253
    Likes Received:
    79
    Capacitors in Linear Supplies are a known point of failure and are regularly replaced as gear ages. If said gear is not regularly used ... maybe an amp is put into storage for a few years ... the rate increases substantially. In some cases large Electrolytics can last 50 years, but 20 is more common and of course some fail sooner. Transformers can overheat and when they do, they self-destruct, an expensive replacement cost.

    That doesn't mean Linear Supplies are unreliable ... I don't consider them to be, personally ... but it does mean that SMPSs don't have a very high bar to overcome to equal a Linear's expected lifetime. As a metric for what to use I don't see reliability to be a negative with either approach.

    My opinion is Benchmark make products that were standards in the Professional Recording and Production field long before Audiophiles discovered them, and the people who choose products for Pro use generally choose by reputation and experience amongst a close-knit community. Broadly speaking Pro gear is more reliable than Consumer gear, and if Benchmark feels SMPSs are "good enough" for their primary customer, which is still the Pro market, then they probably are.
    • Like Like x 1
  3. Johnny2Bad Full Audioholic

    Johnny2Bad
    Joined:
    Jan 16, 2010
    Messages:
    253
    Likes Received:
    79
    The SMPS cited is not intended for use in a Power Amplifier Power Supply; it's intended for Lab Test Equipment (it's original design criteria) and is recommended as an alternative to a Linear Supply for low level signal applications (such as a Phono Pre-Preamplifier). Small levels of noise can be significant in that application.

    So were that particular supply more expensive than the Linear alternative, it would be a criteria that would be closely investigated. Since the supply is in fact less expensive, then "meet or exceed" is, in fact, "good enough".

    In a commercial product, "good enough" is usually good enough. I am not a Good Engineer, because a Good Engineer gets the most performance for the least cost (or more often, the best possible performance that meets a target cost). Even companies like Classé still build to a defined cost BoM. A $10,000 Power Amplifier is still a compromise; it still cuts corners at some acceptable area.

    I DIY, where my build criteria is different ... I can take advantage of the standard BoM ratio in the industry ... where the good Engineer has designed a product where parts count equals 20% of the retail price ... and build at, say, 50% BoM ratio and still end up with an item that performs better than equivalently priced retail devices.

    Many designs do use 10 nF as snubbers. Capacitors are there to prevent AC transformer leakage inductance combined with diode capacitance from ringing when the diode switches off.

    However in that case they are just using default values; a properly designed Linear Supply would use calculated values. The difference is in the ringing; with default values you almost certainly will get some, with calculated values you can eliminate it.

    Step one is to measure the transformer's leakage inductance and secondary capacitance at high frequency, say, 100 kHz, not a trivial procedure. Next, estimate the capacitance of your rectifier(s), which often cannot be found in your data sheets. Punch the numbers into a formula that spits out snubber values -- and then you hope it's all correct. View it with your test equipment, rinse, repeat, and *then* you're done.
  4. Johnny2Bad Full Audioholic

    Johnny2Bad
    Joined:
    Jan 16, 2010
    Messages:
    253
    Likes Received:
    79
    I have enjoyed contributing to this topic, but I've said pretty much all I am going to on the subject. Certainly there are choices that have to be made when designing an Audio product. Designers make their evaluations and come to their conclusions, and carry on from there. Someone ... usually not an Engineer ... decides how much the product should cost. I don't believe the motive is increased profit from "cheaping out" on the product parts count and pocketing the difference, instead I believe the motive is offering a competitive product at a more competitive retail price. Others may differ; I am not here to tell people what to buy. I do feel, however, that buyers should be armed with the best information available to them so they can make informed decisions.
  5. slipperybidness Audioholic Ninja

    slipperybidness
    Joined:
    Jul 24, 2011
    Messages:
    4,910
    Likes Received:
    1,074
    I agree with you 100% on the DIY comments and how you are not constrained to trying to hit a certain price point. One of the main reasons that I also DIY.

    Now, that info on the snubbers is quite interesting to me! The literature that I have read, or at least what I can remember from what I have read, ONLY seems to have mentioned the snubbers as being critical to snub out the noise produced by the diodes rapid switching action. I don't recall reading about the snubbers taking care of the combination of transformer inductance + diode capacitance at all.

    Any chance you could provide a literature reference to this phenomena? I would be very interested to read up on it.
  6. gzubeck Audioholic

    gzubeck
    Joined:
    Jan 8, 2017
    Messages:
    100
    Likes Received:
    1
    That's why I gave it a minimum ten year life to be fair and conservative to smps power supplies...but like you said more often than not a linear power supply lasts 20 years...the transformer failure I'm not sure how frequently they fail but its gotta be much smaller than capacitors.
  7. slipperybidness Audioholic Ninja

    slipperybidness
    Joined:
    Jul 24, 2011
    Messages:
    4,910
    Likes Received:
    1,074
    I can't say that I have ever seen nor heard of a transformer failure as the root of a failure. Perhaps other components failing and taking the transformer out with it, but not a true transformer failing.

    Back to DIY-- I typically use fuses on the secondary windings on my transformers, and you will rarely if ever see that in a commercial design!

    Transformer is likely the largest cost for any electronics project, so protect that investment!
  8. gzubeck Audioholic

    gzubeck
    Joined:
    Jan 8, 2017
    Messages:
    100
    Likes Received:
    1
    If you can't get satisfaction from a $10,000 amplifier or even a $5,000 amplifier it's time to work on getting better speakers and a dac (Vinyl's not my thing).
  9. gzubeck Audioholic

    gzubeck
    Joined:
    Jan 8, 2017
    Messages:
    100
    Likes Received:
    1
    Maybe...but if I was a small company and I can save $50 on parts and $25 on shipping and handling and sell 10,000 units per year that's $750,000 extra dollars in my wallet (after ten years it's more than $7.5 million). You might not care what happens to your product when it starts failing after year 7 if only 20% of the existing product fails. But if your part of that 20% it would suck as your expectations are dashed. Also, your resale value will plummet close to zero instead of getting 50% of your money back (secondary market) when people start realizing your smps is not here to last like a linear power supply amplifier.
    Last edited: Apr 21, 2017
  10. <eargiant Junior Audioholic

    <eargiant
    Joined:
    May 3, 2016
    Messages:
    95
    Likes Received:
    11
    This makes a lot of sense to me.

    I had the 37 year old amp in my avatar completely restored/refurbished with the replacement of approximately 300 discrete components. The 10 large filter caps were deemed to be in good order so they were left in place for now. The specifications on this amplifier are beyond stellar. It is also one of the best sounding and quietest I have ever heard. Even with efficient headphones, no music playing and the volume turned almost all the way up, you hear nothing as if the amp was turned off. Pure silence. Quieter than my brand new Yamaha A-S801 even when it's in Direct mode. I tried to post the schematic here for those that are qualified to review it's design but I get an error telling me the file is too large.

    [​IMG]




    I’m very interested in trying the Benchmark AHB2. I'm going to get the DAC3 so I'll probably get the AHB2 just to hear it for myself. Much of the performance goals that John Siau was striving for in the AHB2 are very similar to those of the AU-X1.


    [​IMG]
  11. PENG Audioholic Warlord

    PENG
    Joined:
    Oct 23, 2004
    Messages:
    8,968
    Likes Received:
    2,415
    Location:
    Ontario
    Transformers typically have excellent overload capability but if you don't overload them and avoid frequent switching (on/off) they can last so long that one could be led to believe they last forever, in normal ambient temperature environment.
    PENG,
  12. PENG Audioholic Warlord

    PENG
    Joined:
    Oct 23, 2004
    Messages:
    8,968
    Likes Received:
    2,415
    Location:
    Ontario
    That's an impressive looking amp, what model?
    PENG,
  13. <eargiant Junior Audioholic

    <eargiant
    Joined:
    May 3, 2016
    Messages:
    95
    Likes Received:
    11
    Last edited: Apr 22, 2017
  14. PENG Audioholic Warlord

    PENG
    Joined:
    Oct 23, 2004
    Messages:
    8,968
    Likes Received:
    2,415
    Location:
    Ontario
    Thanks, very similar specs to my equally old Marantz amp but my SM7 is a power amp and it has a huge EI core transformer.

    For 10,000 yen less per audio-database, the Sansui seems like a much better deal but my Marantz has a better look, subjectively.
    Last edited: Apr 23, 2017
    PENG,
  15. gzubeck Audioholic

    gzubeck
    Joined:
    Jan 8, 2017
    Messages:
    100
    Likes Received:
    1
    I don't know...put a glass top on it and turn it on its side and it looks meaty and delicious to those in the know...sort of like popping the hood on a special edition muscle car...
  16. <eargiant Junior Audioholic

    <eargiant
    Joined:
    May 3, 2016
    Messages:
    95
    Likes Received:
    11
    Looks like a nice unit. Power-wise they do seem close but bandwidth, speed, distortion and SNR is where they differ substantially.

    Some will say these things make no difference but they are precisely what John Siau was striving for with the AHB2. He feels they make all the difference. When I first read about the AHB2 I thought "Wow, a modern day X1! - what's old is new again." ... and with a switching power supply no less, go figure.
    Last edited: Apr 22, 2017
  17. PENG Audioholic Warlord

    PENG
    Joined:
    Oct 23, 2004
    Messages:
    8,968
    Likes Received:
    2,415
    Location:
    Ontario
    Are you just making assumptions? My Marantz manual does not show any SN and speed related specs but the THD and IMD (graphs) show similar specs to the Sansui. The printed specs just show 0.01%, that appears to be conservative, compared to what they show in the graphs. In Japanese, it does say less than 0.01%, closer match to the provided %THD and IMD vs frequency and power output graphs. Mine was a hand me down from my brother who bought it brand new with the matching preamp. He gave me the well preserved original instruction manual, that they call "Handbook of instructions".

    Anyway, all these are just academic but I am just being curious if there are more documentation about the amp/preamp that I would love to read. By the way, you sound like a good potential buyer for that Benchmark power amp:). For me, no thanks, I see contradictory statements on their website to become at least highly skeptical of what they are promoting, there is just no evidence that "Power amplifiers are a weak link in most audio systems".
    PENG,
  18. <eargiant Junior Audioholic

    <eargiant
    Joined:
    May 3, 2016
    Messages:
    95
    Likes Received:
    11
    Hi PENG, this is an interesting discussion. I hope other members allow it to remain that way before it devolves into something else.

    I'd say I was extrapolating as opposed to assuming on some of the points. Here's my logic:

    I believe an amplifiers speed and linear bandwidth are related. The SM-7 has a published bandwidth of up to 100kHz (as most amps) while the Benchmark publishes beyond 200 kHz and the Sansui AU-X1 to 500 KHz! Agreed that this does not mean that the SM-7s bandwidth is not higher than 100kHz. If I'm not mistaken the SM-7 is basically a 300DC which has an SNR of ~ 110 dB whereas the AU-X1 is 125 dB.

    I agree that less than .01% in the SM-7 can be lower as shown in the graphs but the same would go for the AU-X1 whose literature states "both channels driven, into 8 ohms from 5 to 20,000 Hz with no more than .007% total harmonic distortion and inter-modulation distortion." So using the same logic, the Sansui graph could dip below .007% but not above it at any point. If you look at the SM-7 graphs both Harmonic and IMD are well above .007% in some areas of the bandwidth.

    Here's an oscilloscope shot of an AU-X1 square wave at 60 kHz, it takes speed (slew rate- 260v/ms) for an amp to measure like that. Also, Sansui used 12 (yes there are 12 in there) unobtanium Sanken ultra high speed output transistors in the AU-X1 for a reason. Their speed is what makes them special. I purchased an extra set just to have an "insurance policy" because without them, the X1 is not an X1. I do not think the AU-X1 and SM-7 were designed with the same goals in mind. Sure, better manufacturers were conservative in their specs but I think it's a stretch to equate the measurable performance (I didn't say sound, let the user decide that) of the SM-7 with an X-1.

    A bench test would reveal the truth.

    AU-X1 @ 60 kHz
    [​IMG]

    Sanken High Speed Linear Outputs (complete set for an AU-X1)
    [​IMG]

    I have personally never heard an SM-7 but I have heard several Marantz units from the same era and feel that they were striving for a different signature. As a matter of fact, my techs primary specialty is restoring Marantz gear from that era (I don't know if he's ever worked on an SM-7). He is very good at what he does and had the courage to take on the AU-X1 for me. Most techs will shy away from this amp because it is not for the unskilled or faint of heart. Once he completed it he admitted to me that the performance and sound of this amp shocked him. Believe me, not much gear excites this guy and he has no problem telling it like it is.


    Back to the AHB2, John Atkinson said "Benchmark amplifier's specifications are close to those of my Audio Precision system, both regarding the purity of its signal generator and the dynamic range of its analyzer. Correctly characterizing the AHB2's measured performance would therefore present a challenge. And, to my embarrassment, one of the speaker cables with which I routinely measure amplifiers and my test load set to 4 ohms were also introducing small amounts of nonlinearity. These problems—and here I'm talking about the difference between 0.0003% THD+noise and 0.0005%—haven't affected the measured performance of other amplifiers I've tested, but they were detectable with the AHB2's very low intrinsic distortion and noise." He went on to say, the "Benchmark Media Systems' AHB2 is an extraordinary amplifier. Not only does its performance lie at the limits of what is possible for me to reliably test...It is truly a high-resolution amplifier."

    Speaking of Audio Precision Analyzers, it would be pretty cool if we could send the AU-X1 and the SM-7 to Gene so that he could test on his AP. ;)
    Last edited: Apr 23, 2017
  19. PENG Audioholic Warlord

    PENG
    Joined:
    Oct 23, 2004
    Messages:
    8,968
    Likes Received:
    2,415
    Location:
    Ontario
    I have to agree with your logic. Thanks for taking the time.
    PENG,
  20. Johnny2Bad Full Audioholic

    Johnny2Bad
    Joined:
    Jan 16, 2010
    Messages:
    253
    Likes Received:
    79
    "Small" companies are not going to sell 10,000 units of anything, let alone per year. Large companies probably wouldn't get that volume on any particular model that wasn't a bona fide sales leader, or series if they re-use the power supply. A company like Benchmark, who have a well known and respected product, turns over a few hundred DACs a year, and fewer of their other offerings. I remember Lexicon reported they sold exactly 32 copies of a CD player they had in the catalog during the product lifetime, and they have a robust professional installer dealer network.

    The reality is our little hobby is not that lucrative for the manufacturers save for a few non-typically successful operations. Even companies like Harmon International makes their money from things like the OEM Automotive unit versus the consumer electronics divisions. The vast consolidation of brands we've seen over the last 10 years doesn't happen in an industry when companies are making money hand-over-fist.

    Home Audio is a tough racket. Most enterprises start up because they love the product segment, not because it's a good business model. The vast majority don't even manufacture their own products. Oppo doesn't. And that is not new ... it was the model the Japanese used going back to the 1970's. Companies like Sansui and Kenwood didn't manufacture their own products even at that time. Look up "T Codes" ... the system that reveals who actually made a given product in Japan 50 years ago. It probably wasn't the name on the face plate, and it probably isn't a company you have even heard of. Audio firms are design bureaus who contract out actual manufacturing in most cases.

    Frankly, it isn't that lucrative at the best of times. One hit product, or the lack of one, makes or breaks your bottom line. Panasonic recently announced that they will shed the divisions that fail to meet their net profit goals ... 5%.

    It's also extremely competitive. $75 at manufacturer level translates to $375 at retail, even a direct-to-consumer model it would be $225. Can you sell a similar product to your competitors for that much higher price? If you can, and you're not a manufacturer right now, you're in the wrong business.
    Last edited: Apr 23, 2017

Share This Page

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
  • CEDIA