1. Rest in Peace Paul (Kegger) If you would like to help the family in this time of great sorrow and need, you may donate on their GoFundme page: https://www.gofundme.com/mckechnie-medical-and-funeral-fund?
    Dismiss Notice

My amplifier

Discussion in 'DIY' started by Alan0354, Mar 29, 2017.

  1. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area

    Hold your horses!!!


    Let me build and test and make sure the circuit works, do the FFT to prove the THD is same as the last one before you spend money. They are NOT on sale, they both been around like forever, so there is no advantage of buying it first.

    That's said, The chassis is definitely big enough, according to the dimension, the heatsink is at least as big if not bigger than my chassis.

    I bought the exact same transformer also. The chassis and the transformer can definitely do the job.

    I can spare two blank pcb, I put in SMD components on 4 pcb, two for you and two for Dan, it's a lot of work to build two extra as I do production line building 10 boards at a time. There are a lot of SMD components as the whole input and voltage gain section are SMD other than the transistors. I just do this for you two because I afraid SMD might be hard for people that never do it before and I am good at it.
     
    kjello likes this.
  2. kjello

    kjello AK Subscriber Subscriber

    Messages:
    210
    Location:
    Trondheim, Norway
    Ill wait for the green light. And thanks. Again.
     
  3. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    Now one thing I want to make sure. I can only present the THD data and my opinion comparing with the Nakamichi and Acurus, I cannot guaranty how you like it. I have no experience shopping around for amps particular in the last 20 years. I did this all according to engineering, and from the two books of Cordell and Self, that I met their published result. I cannot say anything more than that. So you build it at your own risk.

    I'll test one board on the test fixture to do the stability test ( very important) by hanging caps at the output to make sure it is stable, then the THD test. After that, I'll send the pcbs out. Then I am going to build the amp together with you two and talk you through in this thread step by step as we build.
     
    kjello and I LIKE MUSIC like this.
  4. I LIKE MUSIC

    I LIKE MUSIC Super Member

    Alan, I like your straight forwardness.
     
  5. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    I am not in business nor even interested in making any money on this, I definitely not trying to market my amp. It's all fun and game for me. I just want to make sure people are happy.

    If you are interested, I still have a few bare boards, Since Dan and Kjello want to build my amp, I did ordered 20pcb( not much more than 10), I only work on 10, Kjello want two more, so I still have a few left. But, it's going to be bare board though as it would be too much work to start another run. I put in the SMD parts for them on this run as I am stuffing my own boards, it's a lot faster to do all at once in production style.
     
  6. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    Hi Dan

    I am looking at your chassis, is it possible to move the heatsink and transformer like the way I draw the arrows:

    Dans chassis.JPG

    That is moving the power transformer to the center front of the chassis and move the two heatsink out to each side as indicated. There is a good reason, I need to establish a central STAR ground point. It is very important for the return of the speakers to be on the star ground as short as possible. Also, it is vitally important for the pcb output to the speaker as short as possible also. Star ground point has to be one central point that all ground is as close as possible. If you look at most amps and also my first amp, they are all pretty much layout the same way to prevent ground hum also.

    It is very important how the grounding is, I drew the schematic on how to do the wiring in post #49. The wires from the ground of the filter board to the speaker return connector has to be short, the output of the pcb to the speaker has to be short. That's the reason the filter cap should stay in the middle close to the back pannel where the output connectors are.



    EDIT:

    To Dan and Kjello

    It would be very helpful for you to get the book by Bob Cordell. https://www.amazon.com/Designing-Au...F8&qid=1499300299&sr=1-1&keywords=bob+cordell

    I reference a lot of my design from this book, you can see my train of thoughts in the book. I use a lot of ideas from the book and I add more of my own in my amp. I'll explain more as time comes.

    Also the book talk about the grounding, what is very important. You can get hum if you are not careful. I have experience with system grounding, it is really a physical layout. I have less hum than both the Nakamichi and Acurus.......hell, I have no hum!!!! I go more steps than in the book to pull this out.
     
    Last edited: Jul 5, 2017
    kjello likes this.
  7. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    To elaborate a little, I want to show how my grounding:

    Star ground1.JPG

    I have separate filters for the front end and the output section. the bottom pcb is the filter for output section OPS. I have multiple black wires from the bottom board labeled R-RET and L-RET to the output connector. This is where the major current flows. the top pcb is filter for the frontend ( IPS and VAS). The ground on the top pcb also serve as the star ground. the ground of the lower board ONLY connect to the star ground by the two standoff labeled A and B.


    Star ground.JPG

    This picture is the finish amp that I am using. You see the top board is the filter board of the IPS and VAS and serves as the star ground. You see multiple 16 gauge wires from the ground of the bottom filter board ( hide under) go to the black banana speaker connectors. You see how short they are. Also, you see the output of the OPS board connect to the red banana connectors by multiple short wires.


    These are very important detail that does not show in a schematic. It's all these details that make or break the amp. I spent a lot of years designing system ground to do this. I try to draw the way I do the grounding in the schematic in post #49. Should read that carefully.
     
    nyhifihead likes this.
  8. kjello

    kjello AK Subscriber Subscriber

    Messages:
    210
    Location:
    Trondheim, Norway
    I ordered the book
    It will be the most expensive book I've ever owned LOL
     
  9. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    Buy used, I never buy new books, I have very good luck with used on Amazon.

    It's worth every bit of your money. If you build my amp, you'll learn so much reading the book, you can see the history of my amp blow by blow. It explain what I am trying to do. My book is falling apart and I am planning to buy another one. There is nothing new in my amp, I just have a lot more.

    Most( say all) amps don't adapt all the stuffs talked about in the book. My guess is the worry of instability and also cost. I am not into business, so money is of no concern. I went all out. I even make my pcb 2mm thick so it's not easy to crack and is very strong. I have a lot of experience in taming oscillation and instability and I work on the amp to make it stable. People avoid having high feedback gain as it can get unstable and hard to tame. Also, people avoid 3EF output stage because they tends to oscillate, I use 3EF.

    Because I want to have high bias current to have a big class A region, I have to use lower rail voltage to keep the heat down. I cannot afford to loss headroom. I use the FOLDED 3EF to gain 1.4V head room on each rail. But Cordell's "Diamond 3EF" is not stable in my first trial, I use simple Folded 3EF instead and I tame it. There goes to show you can't trust the book blindly either. I put in my anti sticking circuit to prevent the output from sticking to the rail when clipped.

    You read the book, you'll see how I design the amp and how I change the amp. Both me and Cordell had been an analog IC designer before, power amp is just a big discrete opamp. I talked to him a lot in DIYaudio before and we see things very much the same way. these circuit is nothing new, just like designing IC opamp!!!!
     
  10. Harryconover

    Harryconover Active Member

    Messages:
    418
    Location:
    Charlotte NC
     
  11. Harryconover

    Harryconover Active Member

    Messages:
    418
    Location:
    Charlotte NC
    Right "just" like designing a op amp , you do realize how rare that is , and how fue have done that !
     
  12. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    Yes, not many people done that, that's why Cordell is uniquely qualify to explain the power amp in that light. I really appreciate his book because I designed IC before and designed opamps. I actually checked a lot of his formulas using the book by Grey and Mayer, it's accurate.

    That's the difference in my opinion the difference between the book By Self and Cordell. Self does not have nearly the insight into the amp as Cordell.

    BTW, you asked question about inductance and others and I explained, what do you think, I like to hear your opinion.

    I was thinking about your comment on over current protection, if you look at the schematic on page 1 of this thread, I had the short circuit protection to limit the current, I just think because of the 9 stages, even with protection, the short circuit current is extremely high, it's not going to protect whatever the short is, it's going to be a big bang!!! So why put it in. So I deleted the protection in this new run of pcb. If you have a good reason, I can just hand wire the circuit back in, it's easy.

    If you look at my new schematic, you'll see the fuse on the rail ON the pcb after all the big filter caps, I use 5A and I have no problem. If that burn during normal operation, I can increase the fuse to 6A. I think that's good enough to protect everything.
     
    Last edited: Jul 7, 2017
  13. kjello

    kjello AK Subscriber Subscriber

    Messages:
    210
    Location:
    Trondheim, Norway
    I need to build for 230VAC, is that a problem ?
     
  14. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    No, you just put the two primary coil in series instead of we put them in parallel. I already bough the exact same 30V-30V 800VA transformer, if you look at the datasheet, you see the two primary coils.
     
    kjello likes this.
  15. Harryconover

    Harryconover Active Member

    Messages:
    418
    Location:
    Charlotte NC
    I love good textbooks that have good liner notes I had been workstudy at the bookstore at CU one of the best parts is that you got frist shot on texts if you picked a good ex owner you could youse there notes
     
  16. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    Ha ha, I like to make my own mess in my books. I write so much notes only I can read it. All my books look like crap, but I would not trade it away for anything.

    There are times I can't even fit what I want to write, I actually glue a piece of paper onto the page and write notes on the piece of paper.
     
    I LIKE MUSIC likes this.
  17. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    I just want to show an internal circuit of a popular IC opamp: MC4558. This is as common opamp as it gets:

    IC Opamp.JPG

    See the close resemblance? You have the Input stage IPS differential input pair. Then you have the voltage amplifier stage VAS that amplifies the gain to get high openloop gain. The you have the output stage OPS with bigger transistors to drive the output. You have the current mirror generator to generate the constant current source.

    You literally can build a power amp using this schematic.

    There is nothing new in most of the power amps, all are just tweaking of this basic design, adding circuit to improve and get lower distortion.

    I studied the book by Malvino long time ago, I could swear the book explain the operation of opamp internal circuit to certain extend.

    You can even see the short circuit current limiting transistors hiding between the two output transistors. BTW, the top transistor for short circuit protection is wrong. It should be NPN
     
  18. Harryconover

    Harryconover Active Member

    Messages:
    418
    Location:
    Charlotte NC
     
  19. Harryconover

    Harryconover Active Member

    Messages:
    418
    Location:
    Charlotte NC
    Yes looking at someone's books gives me a good idea of were there st in there "trip of life" I to have a well worn copay of self on audio complete with my notes , that most can't read My secratrys allwise referred to my penmanship as "Harrygriphics" , my enclosed foam a fi as to transmit meaning not check spelling ! My comment on unductance referred to that presented as a load and a large inducto
    Say 10 my whith a D.C. Current of a amp upon breakdown of that magnetic field the a mount of stored energy can be quict large , a large full sized bass reflex for a 15" can store a lot of energy becase of the mass of the driver .so my point was how dose the circuit deal with the retuned power ?
     
  20. Alan0354

    Alan0354 Super Member

    Messages:
    1,469
    Location:
    Bay Area
    Thanks for clarifying.
    I reply to you in post #99 in the last page, if you look at the schematic, I have 4 diodes circled in red color that goes from output to both rails to prevent the inductor load from kicking the voltage above either rail voltage. That is the diodes clamp the output from going 0.7V beyond the rail voltage to protect the output transistors and the solid state relay transistors. I use 1N5408 3A big diode to clamp the voltage, they are big diodes.

    Also, you were concern about the current limiting, if you look at the schematic in post #99, I have fuse on the power input to the pcb ( on top of the main AC fuse, that would be way too slow to respond). Since the fuse is on the pcb already, that is after all the big filter caps that store the major portion of charge. The only filter caps after the fuse is only about 100uF. So if there is a short, the output transistors only have to discharge the 100uF before pulling large current and burn the fuse. My guess it will burn the fuse in less than 1sec.

    Also, if you look at my output stage, the first EF is folded and current limited to about 10mA only. This mean the driver transistor can only supply about 0.5A DC current ( assume beta=50), that mean the max DC current from the output is 0.5A X 50= 25A only. That's not going to burn the transistors in the 9 stage output. But yes, it will still make a big arc and burn things at the place of short.

    Note that I talking about DC current during a short. Audio signal is AC, the AC current capability of the output stage is higher, more like 40 to 50A minimum.

    Tell me what you think.
     

Share This Page