Adcom GFA -565se

Discussion in 'Solid State' started by Frank Wong, Jun 9, 2017.

  1. Frank Wong

    Frank Wong Member

    Messages:
    77
    Hey Alan,

    Just read some reviews on it not that great like you say. But power for power no clipping can we really hear the difference between pure class A & AB. Obviously the bsc amp went into clipping as reviewer said sound became hard as volume went up.
     
  2. Frank Wong

    Frank Wong Member

    Messages:
    77
    I'm leaning more & more towards Adcom. Great Thd specs but none posted for 2 ohms.
     
  3. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area
    I think you can hear the difference between class A vs class AB ( generally) at low volume IF the class AB amp has a small class A region.

    Even in class AB, at idle, both the top and bottom transistor both conduct and is in class A mode, people set a idle bias current, let's say it's 100mA idle current. When there is a signal, the output swing up and down. When the current demanded by the load is twice the idle current ( 200mA), one transistor totally turns off and only one transistor driving the load. This is when the amp goes into class B operation. Say for 8ohm load, 200mA can swing the 8ohm load up to 1.6V( V=I*R=0.2A X 8ohm=1.6V) before the amp goes from class A to class B. So if you listen very soft and output never swing beyond 1.6Vpeak, you are running the amp in class A even though the amp is class AB. It's only if you turn it up, then the amp works in class AB.

    That said, good amp use heavy bias current of 0.5A to 1A to get a bigger class A region. So at normal listening level, you really have a class A amp.

    So to answer your question, yes, there should be a difference between class A amp and class AB amp. BUT, if you choose a good class AB amp, you can almost get a class A amp at normal listening level.

    The main problem of class AB amp is the crossover distortion when one transistor turns off the other take completely over. It is not signal amplitude dependent. That means the lower the signal level, as long as you cross into class B, the crossover distortion is constant. Crossover distortion affect the most at lower signal level when the amp cross into class B. But there are ways to mitigate the problem.

    1) increase the bias current to get a bigger class A region. The price to pay is more heat and need bigger heatsink

    2) Have more transistor pairs. double the number of output transistor almost half the crossover distortion. So it you have two pairs instead of one pair, you almost half the distortion. If you have 4pairs, distortion goes down to 1/4. 8 pairs reduce to 1/8.............That's the reason high end amps have so many pairs of transistors.

    That's the reason why I have been talking about the amp worth their weight in gold, that I look at the size of the heatsink, the transformer and the number of transistor as the first judgement of an amp. The emitter resistors of Adcom looks to be 0.33ohm, if they adjust it according to Oliver's condition, you have about 0.8A of idle current. So you should have about 5W of class A power into 4ohm. That's quite good.

    The Adcom 565 has so many pairs I lost count!!!! If I read correctly, it has 10 pairs!!!!............one pair more than my amp. Adcom is a low THD design as I described before with all the features. I am very impressed with the Adcom. I would love to have a side by side comparison with my amp. The Adcom must be a heavy and big amp, I bet it's about 60 to 80lbs. With all the output transistors and big heat sink, your Adcom can drive 2ohm!!!! My God, how much more transistor pair do you want!!!! That's a huge amp!!!!

    Class A amp waste too much power, generate too much heat, I just cannot see I even need to crank that high level to warren a pure class A. I design my amp to have about 1A idle current to get about 8W of class A into 4ohm, I think I am good for my listening level. I am like you, I don't crank it up.
     
    Last edited: Jun 19, 2017
  4. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area
    I was curious and read more about BSC :http://bscaudio.com/Pages/technologyCircuitry.php

    I don't agree with a lot of what they say.

    1) I don't agree complementary is better. When I designed my amp, I made it modulize, that is I have input and gain circuitry on separate pcb from the output section so I can experiment with different topologies. I have complementary and single ended, I tried both and the single end kicked the butt of the complementary both in THD measurement and listening. I have scientific data to back up my claim on this.

    2) The round corner of pcb they claim is absolutely full of it. As an engineer designing with transistor circuits for 30 years, I always do my own layout. I designed microwave circuit and layout microwave pcbs. You ONLY round corners for circuit way way over 100MHz. I know all about stripline design and all the high speed signal integrity. For audio frequency and even to 1MHz, there is absolutely not different whether you round or use sharper bends.

    3) I don't agree through hole components are better. Yes, for high power components, you have no choice as you need the size to dissipate the power. But I use as much service mount components as possible for the lower power area. For one, the size is small, it's easier to layout without worrying about cross talk because the signal path is a lot shorter using SMD.

    They even proudly show the picture of their layout:
    BSC layout.JPG

    They are using 2 layers signal trace pcb, they can't even put a ground and power planes with two layers. In all my boards, I have full ground planes and power planes. Even the output signal run on a plane to lower the resistance in the signal paths. This is amateur layout. From looking at this, I can lay it out so much more compact and shorten the signal path so much to avoid crosstalk and interference. I was the manager of EE, if the pcb designer give me this layout, he/she'll be fired.


    EDIT:

    Here is my layout that I just sent out for my next amp:

    OPS Layout.JPG

    This is a two layer pcb, you can see full ground plane ( blue on the left for output section, red on the right for input and gain section) to prevent noise pickup and radiation.

    You see separate +V, -V rail planes for the output section, the middle is the output signal, it's a wide plane. Signal traces is as short as possible. Look at the low power section, you see very few traces as most components are point to point connected and you don't even see the traces.
     
    Last edited: Jun 19, 2017
  5. FootFungus

    FootFungus Super Member

    Messages:
    1,040
    Location:
    Dallas
    Have you thought of bi-wiring the Martin Logans, assuming they can? You can get a decent high powered pro amp for the bottoms, and pick a lower powered finesse amp for the top.
     
  6. Frank Wong

    Frank Wong Member

    Messages:
    77
    Thanks for the great read Alan! Soundwise how would you describe the amp you built?
     
  7. Frank Wong

    Frank Wong Member

    Messages:
    77
    Unfortunately the new motion esl are single wire!
     
  8. Frank Wong

    Frank Wong Member

    Messages:
    77
    Hey Alan what you think of the topology used in the Schiit Freya preamp?
     
  9. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area
    Compare to the Nakamichi Stasis, my amp has more separation, you can hear different instrument clearer and more detail. It has a bigger sound stage, more of an echoy sound like in a bigger hall. Also it has m, ore attack.

    My amp is not the fastest amp, but I have about 30V/uS slew rate which is already very fast. Frequency response is over 200KHz ( I really never measure in closely as I think the slew rate and rise time is more important). Rise time is a little over 1uS.

    I am still new in this audiophile design, my first amp is emphasizing on ultra low THD which I achieved. There is a school of people that believe lowest THD is not necessary the best, I cannot dispute that as I don't know. I only have the Acurus and Nakamichi to compare to. The Acurus is way lower, it's more like the Adcom 555 quality and the THD measurement shows. I intentionally bought the Nakamichi because it's designed by the great Nelson Pass and is almost identical to his famous Threshold S300 ( each has their strong and weak points). I pick this Stasis design because it does not have a global negative feedback and he use the so called CFP output stage. The design is not for low THD. I bought this to serve as a reference for me to compare and beat. I did not expect to beat the amp, but I was really surprise I can hear the difference and my amp sounds more real and dynamic. I gave my wife the double blind test and she could pick out my amp has more separation. She described she can hear things clearer with my amp also.

    This is only my opinion and I am at the process of proving it: I think people that claim low THD is not important needs to take a second look at their speaker cables. I believe if they have very good cable that does not introduce distortion, they will hear the difference and the ultra low THD shines.

    I did extensive testing in THD vs speaker cables and I compare quite a bit and make different cables. A 10ft 12 gauge monster cable that measure 0.006%THD right at the output of the amp, INCREASES to 0.14% at the load end with a resistor that is not that linear. Doubling up the cable half the THD to 0.07%. But still it's no where close to 0.006% at the output of the amp. I then experiment with different ways of making cables and I got down to about 0.035% or so with the same length. It is my opinion the speaker cable has to be as short as possible, monoblock is the way to go. I am building a second identical amp so I can run monoblock. I talked about all these in the two thread I gave the links to you. The key is to have multiple cables in parallel. You cannot have one big cable, that does not work. My current cables in the system uses 5 pairs of 16 gauge cables in parallel together to drive the mid and tweeter. Then I just use two pairs of monster 12 gauge cables to drive the woofer.......For each speaker!!!! If you read the thread on my amplifier, you'll see pictures of my setup and how huge those cables are. They really really make a different. The different is as big as between my amp and the Nakamichi.

    Even if you have very nice cable, try experiment parallel more cables. Don't have to waste money buying expensive cables, try the way I described, it's only going to cost you like $20 to build two cables. What do you have to loose but $20. I don't think I can hear the difference between the two amps if I don't have the new cables.


    I don't even want to guess and comment on Schiit Freya preamp. It's tube which I have not work with in audiophile ( I did successfully design two guitar amps that sounded very good, but that's different animal). Now a days, people use CD and DVD or other device that are line level signal. You really don't have to do a lot on the preamp. Hell, if you put the volume pot on the power amp, you don't even need any electronics, just a volume pot to drive straight into the power amp. That gives you the most natural sound. When comes to audiophile, less is more, the simpler the circuit, the better it sounds. Any electronics degrades the sound. The ideal amp is a wire with gain!!!! I designed my own preamp and I have two that I am using. It's simple minded circuit, just a relay switching matrix to route different inputs to the control board, a remote motorized volume pot and a buffer amp to drive the coax.
     
    Last edited: Jun 20, 2017
  10. Frank Wong

    Frank Wong Member

    Messages:
    77
    Hey Alan! I don't doubt your amp sounds superb! If you were to sell it how much would you charge?

    BTW I stumbled across Aliexpress.com where some Chinese companies are putting out clones of Dartzeel & Accuphase class A amp. Google it they have big pics. What you think?

    BTW the Freya preamp is a three in one presmp.

    1. Passive using 128 step antenuators
    2. Jfet buffer
    3. Tube 6SN7 buffer
     
  11. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area
    Ha ha, I am not in business in selling amps!!! I don't even want to deal with making money at this point. I put in my 30 years already, this is all fun and game for me. If you are interested, I can sell you the bare pcb at cost and I can help you on the forum to build it. I ordered 20pcs of the pcb so I can spare a few. But this is way too premature, I just sent out for fabrication, I have to test and verify it is up to standard before anything else. As I said, I am willing to selling at cost, which is $10 each pcb IF everything test out good and get the low THD. But forewarn, it's not cheap to build, the pcb is the least of your expense.

    Do you have experience in electronics? Where do you live? I live in Sunnyvale California, if you are close by, come and listen to it first, talk is cheap, all the THD spec and all doesn't mean anything unless you like the sound. If you come by and listen and give me your honest opinion, I give you the two boards free!!!!

    Regarding to the clone of the Dartzeel, I looked at the site, it's supposed to be 99.99% copy of a very expensive amp, so I don't dare to criticize it. I just don't like that idea it uses 1 pair of output transistors only. I sure would not pay big money for the real one, but the clone is only $600 or so.

    Clone.JPG

    If you listened to the real thing and like it, then by all means, buy the clone.
     
  12. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area
    Wait, something really does not add up with the clone amp. Go look at some pictures of the internal of Accuphase amps, they have many pairs of transistors, big heatsink, lots of filter caps, typical of a very high end amp. This clone looks NOTHING like that. It's empty.

    Something is really not right in the pictures of the clone, it only has one pair of output transistors MJL4281 and 4102, there is no way two transistor can drive 200W into 4ohm without burning up quick. It never say it's class A. There's no way it's class A. I don't trust the clone. Don't trust me, go google internal pictures of Accuphase and you'll see the huge difference.

    I think they lie.
     
  13. Ds2000

    Ds2000 All About every cool stereo component. Subscriber

    Messages:
    2,915
    Location:
    New York City
    The pictured amp is a DartZeel clone, here is the read deal DartZeel NHB-108. I'll take the 108, thank you very much.
    IMG_1077.JPG
     
  14. Ds2000

    Ds2000 All About every cool stereo component. Subscriber

    Messages:
    2,915
    Location:
    New York City
    Also, my comment above is part of the thread jack gong on, so my apologies to the OP.
     
  15. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area
    The 108 has bigger caps, but still one pair of output transistor. I got the link of DIYaudio and the schematic, Doesn't look that hot either.http://www.diyaudio.com/forums/solid-state/284517-dartzeel-based-ebay-kit.html

    This is the schematic:

    Dartzeel .JPG

    Not crying out high quality. Like comment on diyaudio, definitely not low THD and low damping factor. Of cause you have to listen to it to judge, but my money is on 565.
     
  16. Frank Wong

    Frank Wong Member

    Messages:
    77
    Hey Alan,

    I'm no where near you unfortunately! Also I'm ok with soldering iron as I use to build my own speaker crossovers! But I couldn't handle putting the board together! Haha

    I won't get the clone as everything seems less smaller and lighter. You should see the Mark Levinson ML2 replica. What a joke. No way you get 25w class A. Yeah I'm leaning towards Adcom. I like Emotiva too but mono blocs too big. I will go to store to listen to a Bryston to get a handle on their sound. I demoed a 3St many years ago and the highs were not refined. I'm looking for a sweet refined sound I still have fond memories of a all Mark Levinson system 100w per ch driving Theil CSi5! Sound was holographic and layered. Refined with slight dark tint. Another system was Audible Illusions L2 preamp with Classe ca-100 driving Magna planers. On the Chesty ultimate demonstration dusk on sax solo. The clickty clack of Sax keys was so papalable. Even on my system now not even close. Another system was Apogee Slant 6 driven by YBA 2 electronics. Singer Sara K was so enchanting like singing to me.
     
  17. Frank Wong

    Frank Wong Member

    Messages:
    77
    Yeah I wouldn't touch the clone unless I could hear them side by side covered up so I didn't know what amp was in the chain.
     
  18. Frank Wong

    Frank Wong Member

    Messages:
    77
    My favorites if I could swing it would be the Coda S12.5. And of Course pass X amps ( too big and hot though). Also I remember a Threshold T5 driving Martin Logan Aerius which did a great clog on Hotel California from cd He'll Freezes Over. Alan if you have the cd can you see on that song from 49 secs to about a minute into the song where you hear the drum beat near the left channel. Around 50s mark the bass should change pitch was very audible on my old system but can't hear it on my system. I don't know what in the chain is not letting me hear it. It's probably the woofer in the esl which is the weak link in the chain.
     
  19. Alan0354

    Alan0354 Super Member

    Messages:
    2,676
    Location:
    Bay Area

    I wish I have the space for the Martin Logan or Magnapen. The way my set up is the two speakers are on each corner between the wall and the entertainer hudge( huge). Until I get rid of the hudge, the speaker is confined. I am waiting for the old tv to die, I still have an old Mitsubishi 72" projection tv, I am born cheap, if it's not broken, I don't have the heart to dump it....and it's is lasting forever!!!!! I want to get an 80+ inch next time and it won't fit the hudge, I have an excuse to dump the hudge together with the old tv at that time. Then..........I will have more room for the speaker to breath. Those panel speakers need a lot of room to breath. I have a pair of JM Lab Spectro 913.1 speakers which is not bad, but it would be nice to try out something else.

    Is Theil good, I never have a chance to listen to one? One brand of speaker I like is Sonus Faber, I listened to a pair of Concertino ( not even expensive) with Cary SE monoblock, it was heavenly. I bought the JM Lab because it was a demo sale and I got it for $3000 instead of original price of over $5000.

    You don't like monoblock? I am building a second amp just to make it monoblock!!!!


    Yeh, the more I look, the more I don't trust those clone amps. My friend has a Jolida tube amp he loves, but it won't drive panel speakers, he has a pair of Magnapan and he can't use them. He is diehard tube head, only tube amp is good enough for him.
     
  20. Frank Wong

    Frank Wong Member

    Messages:
    77
     

Share This Page