Small Audio amp questions

Discussion in 'General Electronics Chat' started by chunkmartinez, Mar 15, 2013.

  1. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    I have been playing with op amp amplifiers for headphone amplification.

    I understand most about the design, negative feedback etc...I just have a few questions.

    Below is a schematic that I generally went off except on one of mine, I use 100uF caps in the rail splitting supply. I have one in front of me on a breadboard now that I used an OPA 2134 op amp for. I am also generally use a 50k input POT and feedback voltage divider with R values of 1-2k, and 6.8k(per channel).

    [​IMG]

    So, what exactly is the purpose of the capacitorss, does it contribute to even the rail splitting out? I have read about it before somewhere...The 4.7k resistors, this is the primary rail splitting part that creates the virtual ground correct? I just want to be sure. The small NP caps are bypass caps, correct? Well, what I really want to know is, what values for caps, and resistors are necessay for a given a desired power draw, etc? Do the resistors dissipate power in the power supply reguardless of the total/idle power draw?

    Also, what is the 100k resistor for paralell to the input? Is it to set input impedance? Does the input impedance depend on a comination of the op-amp gain, and this resistor in relation to the input devices output impedance? My laptop that I use measured 60 something ohms per chanel, any notes or equations to deal with setting in/out impedances? Is going off my output devices measured resistance at the end of the cord a good thing to go off, or does the amplifiers gain on the device count too? I have been learning a bit lately that I never knew before and I think I'm getting it down. Is the gain of a transistor it's equivalent input/output impedance?

    Lastly, I have been playing with this op amp to push pull circuit below:
    [​IMG]

    I love the idea of beeing able to source much more current from an op amp. If you have seen, some designs use a diode between each transistor on the input(the base) of them. I am using a bc547B and bc557B for these, to-92 package. Btw, my cheap DMM gives me an hFE reading of 300 when I measured my bc547. I would really like notes on what I can add or do with this circuit to optimize it? I have complimentary tip41/tip42 transistors (to-220 package) that I really want to add on the back of these to-92s to further increase current supply so I can try and shoot for around 15W. Question, how do I optimize the circuit, and also how do I implement adding the last larger current amplification stage, just like the first? And for 15W, what sort of batts would be able to supply this, 4 6 volts(relatively large cells) for a dual 12v supply? What about for 2 chanels? I also realize that I will need to add negative feedback to the last stage aswell but would this be okay or will it have issues with the first neg feedback stage?

    One more, with the use of biasing diodes in the push pull stage, dosn't this mean that the amp is a class A/B, or close? The definition of class A/B is constant partial bias, right? Also, with the way I have it setup rightnow for me I have a low voltage gain op-amp(300mV max) connected directly to my to-92 PP stage with nothing else like biasing diodes, etc, and I am def using negative feedback on the output. When I use the negative feedback, I don't seem to get cross-over distortion on my o-scope. I have looked at music, and sine waves(50hz). So is the biasing diodes necessary? Will it depend? When I listen to the amp it sounds surprisingly great but except there is a TINY high frequency distortion with the voice sometimes it seems. It is so hardly noticeable but there. I am not for sure if it is constant and I only sometimes notice it or if it comes and goes. But it is hardly there.

    Sorry for the long post but i'm craving for the knowledge and really enjoy this stuff. Any notes or explainations would be awesome.
     
  2. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Chunk, maybe this is just me, but I get overwhelmed when too many questions are asked at the same time. I could give you good answers to most, or maybe all your questions, but I don't have the energy to tackle all of them.
    I counted 18 question marks, and I might have missed a couple.
    Maybe you could prioritize them.:D
     
  3. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    I know there's alot of questions. Why don't I restart and maybe ask one by one.

    Question: On the pic in my original post, I want to add higher powered transistors along with small to-92 transistors(bc547/557). I have made a thread where I learned alot about the limitation of a circuit that I wanted to build. It was impractical. I would like to have a class B or semi class B that is driven by an op amp input stage. I have larger complimentary transistors I want to put behind my to-92s since an op amp trying to drive large low hFE transistors is impractical. The to-92s will be used to ocomplish this. I have been reading up alot and I want to try a darlington configuration with them.

    What I would really be interested in asking is, what modifications to the power supply in pic one can I do to be able to supply upto 10-15watts? I know when rail splitting it can be hard to keep the rails balanced when dealing with high currents, what battery power options do I have when dealing with upto 15 watts? I have read about some but I will probly need into the ampere range.
     
  4. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    15 Watts of sinusoidal power into 8 ohms requires

    V=√(P*R)
    V=√(15*8)=√120=10.9V RMS
    Vpk=Vrms*√2=15.5V
    The peak current is 15.5V/8Ω=1.94A

    So, your audio amp output has to swing 2*15.5V=31V p-p.

    15W into 4 ohms requires
    V=√15*4=√60=7.75V RMS
    Vpk=7.75*√2=10.95V
    The peak current is 10.95V/4Ω=2.74A

    Does this sound like what you expected?

    If you want to proceed with the split supply, it can be done, and we can help you with it.
    Also, we'll try to help with the output stage.
     
    chunkmartinez likes this.
  5. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    I am wanting to use the amp for a portable boombox. It will use a pair of efficient full range speakers in stereo(I will use two seperate channels) to work okay off the the limitation of power I will have. I think I can deal with a split 12V supply at 8 ohms to make things on the simple side. I will be using 4 6V batts, not the large heavy ones for scooters but the "household" ones that are still decently large. If anyone has experience with these batteries and think that they will be alright with the current draw at a 4ohm speaker load let me know. Now that I think about it, I would need to double the speaker impedance per chanel to split the power in half so it would be 8-16 ohms per chanel. About 7.5W per speaker.
     
  6. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Check ebay for "Class D amplifier module" they are common and cheap, and use digital switching tecvhnology to make a lot of sound power from very little battery drain (as you mentioned portable).

    You just connect power and speakers to it and have a ready made amp!
     
  7. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    I don't really want to use a class D module for this project.
     
  8. nigelwright7557

    Senior Member

    May 10, 2008
    487
    71
    If you are using 2 batteries then you already have +, ground and -9 volts so you don't need a rail splitter.
    However you will need to switch both rails with the switch.

    The 100k biases the input correctly around zero volts.
     
  9. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    On this later project I will be using +/- 12V actually...So, theres no need to worry about balancing the rails when using two power sources vs one?
     
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Your power supplies don't have to be equal (in magnitude) to make a good audio amp. Unbalance just affects which peak (positive or negative) clips first.

    I'm working on a design.
    Are you locked into a particular op amp?
    What is your input source?
    Do you need adjustable gain? Over what range?
     
    chunkmartinez likes this.
  11. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    Problem is that the OP design has no adjust for DC offset so it will be running current through the speakers. If the output is DC coupled, it needs an offset trimmer.
     
  12. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    I was addressing the case where he has positive and negative supplies, and therefore would not need the resistive divider for creating a "virtual ground". See post #9.
     
  13. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    I am not stuck an a particular op amp...I will probly use a tl082 or similar..but I am not stuck on using a particular model.

    My input source is either a laptop, or ipod.

    My laptop has a max voltage of around a volt on music. When I measured this, it was on my handheld oscilloscope....does an oscilloscope show peak to peak voltage as the instantenous voltage displayed? This is what I went off of.
     
  14. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    btw, here is the page that I read about the balancing of the rails: http://tangentsoft.net/elec/vgrounds.html

    It says the higher the current, the harder it is to keep the rails balanced? This is why I figured that using a split power supply would be hard when dealing with a 15W amplifier. Would it be easier to use a single ended supply and then AC couple the output

    EDIT: Wait, I forgot, it was mentioned that using a power source per rail(batt) avoids this correct? I wouldn't have to use a resistor divider with two batts either? wow I wasted time(and batt power) using them on one of my projects lol..
     
    Last edited: Mar 25, 2013
  15. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    When you use two batteries, the node where they are joined is your ground.
    The reason I asked about your choice of op amps is that most of them don't swing rail-to-rail on the output (e.g., TL082), and this limits your output power. If squeezing every watt possible out of your amplifier is important, then a rail-to-rail op amp (e.g., LT1677) is in order.
     
  16. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    Sorry I havent replied for a while I have been a little busy...

    Yea I have been ommiting the resistors and using the middle node as the GND. I knew it would work before but though there was an issue but after re-reading that write-up I see that it isn't as big of a deal as I thought.

    Okay so, the LT1677 swings all the way from rail to rail or very close? If the price isn't bad and the specs are good enough I don't have a problem using that op amp as my choice.
     
  17. timescope

    Member

    Dec 14, 2011
    298
    44
    You could also experiment with TDA2030. It is a complete amplifier in a 5 pin package and performs very well.

    Timescope
     
  18. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    Thanks for the suggestion but this is a project that I wan't to learn from and an IC amp is sort of too simple and straight forward.
     
  19. timescope

    Member

    Dec 14, 2011
    298
    44
    That's good.

    Here is another way of adding power transistors to an op-amp to make a TL081 12 watt audio amplifier.

    This circuit uses the quiescent current of the op-amp to bias the power transistors and eliminates the problem of the op-amp not swinging rail to rail.

    Timescope
     
    chunkmartinez likes this.
  20. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    Edit: When I first looked at this circuit I didn't notice the design untill I actually really looked at it...

    Awesome, this is the sort of stuff I like to learn about...

    In reality guys, I am not that stuck on full efficiency and swinging rail to rail and getting a full, or close, 15W. However I wouldn't mind learning how to do it since I want to learn anything I can that is practical to circuit design. I am to the point in electronics where I want to know how to properly and fully design circuits. I wouldn't mind the circuit Timescope posted. I can use a tl072/82 for stereo operation. Any other design explainations or tips to this circuit?
     
    Last edited: Mar 31, 2013
Loading...