Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Audio amplifier

  1. Jun 18, 2013 #1
    Hello,

    I am trying to make an audio amp on my breadboard, for fun I guess.
    But it's not working! I am running off of this circuit of the lm1875.

    http://www.google.com/imgres?imgurl...=KBTBUZqGFLPp0QH3xIAg&ved=0CDEQ9QEwAA&dur=189

    So I put the circuit together many times with different components on different breadboards and it just makes a HORIFIC sound. For the power I have a variac or 1-110v output transformer, then I am plugging in a 24v transformer on the output. (I do this so I can safely control the voltage.)
    I have the rectefier hooked up and everything on the breadboard. as the ground I am running a line strait from the ground on my variac to all the points in the circuit which must be grounded.


    any thoughts?
     
  2. jcsd
  3. Jun 18, 2013 #2

    davenn

    User Avatar
    Science Advisor
    Gold Member

    You had better show us YOUR construction
    I particularly want to see your power supply
    Why are you using a variac ???
    you should be using a proper mains transformer with a 220/240 VAC primary (110VAC for relevent countries) and a secondary that is a ~ 20 - 0 - 20 centre tapped secondary

    This is because you need to provide the chip with a +VDC - 0 - -VDC supply

    I suspect you havent done this
    supply some good sharp and well lit pics of your construction :)

    Variacs are VERY dangerous devices as there is NO isolation from mains voltages
    get it out of there before you kill yourself and get the correct transformer as I suggested further up

    Dave
     
  4. Jun 19, 2013 #3

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    You do make them sound really scary devices, Dave. The OP is using a legit transformer as well, it appears.
    I remember a guy who I worked with, years and years ago. He used to do a lot of what we called "home office" in the lab (very low pressure of work in a government - type establishment). He would always use a variac and gradually wind up the volts to full mains. Some power circuits really don't like 'brown out' conditions and he was known to have produced smelly smoke doing it that way.
    Variacs are not that deadly as long as you remember that the few volts you get out could shoot up to full mains if the neutral wire comes off - but that's mains devices for you.
     
  5. Jun 19, 2013 #4
    Yeah. Use a battery.
     
  6. Jun 19, 2013 #5

    davenn

    User Avatar
    Science Advisor
    Gold Member

    MY earlier comments stand .....the whole problem is that he doesnt really seem to know what he's doing with low voltage supplies, how much worse is it if he's connecting the 0V rail of the amp to the GND of the variac ???? bypassing any isolation protection his incorrect other transformer gave him.
    its a cloud of smoke and a possible zapping waiting to happen


    I agree, tho in this case he needs 2 batteries in series so he can produce a -V, a 0V, and a +V rail

    a couple of 12V gelcells would give enough voltage at least to do basic testing, then once the circuit is working he can progress to a proper centre tapped secondary transformer and rectifier etc PSU

    Dave
     
    Last edited: Jun 19, 2013
  7. Jun 19, 2013 #6

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    What we have here is a failure to appreciate difference between "circuit common" and "ground".

    With dual supply made from just a 24volt transformer and rectifier, where is power supply common?

    I'll bet it DOES hum !

    see this thread:
    https://www.physicsforums.com/showthread.php?t=677676
     
  8. Jun 19, 2013 #7

    davenn

    User Avatar
    Science Advisor
    Gold Member

    hey Jim

    yup exactly, there isnt one

    I hope the OP comes back with the pics I asked for so we can see what has been done and so he can be guided correctly from there :)

    Dave
     
  9. Jun 19, 2013 #8

    davenn

    User Avatar
    Science Advisor
    Gold Member

    OK
    taking the drawing from the datasheet, have added to the image a couple of power supply choices

    here's a dual 12V 7A/H gel battery setup for initial testing......

    attachment.php?attachmentid=59712&stc=1&d=1371687975.jpg

    here's a suggestion for a mains supply......

    attachment.php?attachmentid=59715&stc=1&d=1371689220.jpg

    depending on how hard you want to push this amplifier, a 40V/8A centre tapped transformer would be the maximum. A 36V / 5 to 6A centre tapped transformer will power it quite well and provide a bit of current headroom

    cheers
    Dave
     

    Attached Files:

    Last edited: Jun 19, 2013
  10. Jun 20, 2013 #9

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    Wow Dave - nice explanation and diagrams !

    To Tesladude: Note Dave's two different symbols - circuit common is the three lines in an upside down pyramid, earth ground looks like a garden rake. He's shown them NOT connected.


    He could make your top diagram with two 9 volt batteries then try earthing the common and hear effect on hum. 9V batteries would make plenty of milliwatts to hear, but 1875's 100ma quiescent current will run them down in a short while.

    It'd be good for learning.
     
  11. Jun 20, 2013 #10

    rbj

    User Avatar

    doesn't the power supply ripple filter need some (small) resistors after the 4700 μF caps (in series with the VCC and VEE to the amp)? the only way to make the ripple small is make C1 so damn big that it holds the peak voltage from the rectifier which means it only charges at very thin spikes when the diodes conduct.

    in fact, you really need two C1 caps, one from VCC to common and the other from VEE to common (well, actually not the final VCC and VEE but what comes out of the bridge rectifier. it's not good enough to connect a single C1 between the + output of the bridge and the - output because it doesn't deal with any common-mode component to the ripple. both VCC and VEE could be rippling w.r.t. common yet they might seem rock solid to each other.
     
    Last edited: Jun 20, 2013
  12. Jun 20, 2013 #11

    rbj

    User Avatar

    also, i am a little concerned about the 1 Ω + 0.22 μF load on the output of the amp. sure it ain't 1 KΩ?
     
  13. Jun 20, 2013 #12
    Ok so normally I work with low voltage analog and digital circuit design, never really anything with a ground or anything past 12v for that matter! So I didn't even know the center tap is the ground in this case!
    Don't you miss the good old days when the earth ground symbol actually meant earth!

    Ok so I used the center tap as my ground with a 24v transformer, main decoupling at 4700 just for testing.
    And it was slightly amplifying my music up to maybe 1/4 of a watt until when my volume on my phone was at half and I started hearing crackling and everything bad.

    When I took pictures my memory card was not even in the camera so I will upload another time.
     
  14. Jun 21, 2013 #13

    davenn

    User Avatar
    Science Advisor
    Gold Member

    no, its correct, as per datasheet, actually it can be anything from 1E to 10E

    Dave
     
  15. Jun 21, 2013 #14

    davenn

    User Avatar
    Science Advisor
    Gold Member


    Hi tesladude

    am pleased you are still with us, I thought you may have given up on us all ;)

    with the way that circuit is wired, you cannot really use a non centre tapped transformer
    The way I am reading that comment is that your transformer DOESNT have a centretap secondary .... it has just 2 wires out ?? please confirm)

    if that is so, you really need to get a centre tapped transformer else you really are going to struggle to make it work properly. For reasonable performance you could use any of the following secondary tapping options ....

    12 - 0 - 12
    15 - 0 - 15...
    up to 20 - 0 - 20
    and rated for ~ 5 or 6 Amps

    cheers
    Dave
     
  16. Jun 21, 2013 #15

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award



    Sounds like you're making progress. That's how we learn - iterating toward success.
    Isn't audio fun? I love it. Back in 1963 I built a tube amp with about same power as this LM1875. Used 6BQ5 pentodes.
    It was voted best sounding stereo in the dorm.


    ................................... To your project:

    I'd say rbj has a point there - two filter caps, one for V+ and one for V- is your next step.

    Ditto to dave on the 1 ohm resistor, it's for high frequency stability. The .22 uf cap limits low frequency current through it.

    Now THIS is interesting ; TI's datasheet gives a schematic and pcb layout for single supply like you tried at first ----
    but you used the split(dual) supply schematic instead.

    For single supply see fig 2 here, and PCB layout is fig 15.:
    http://www.ti.com/lit/ds/symlink/lm1875.pdf
    Note how they biased the + input to middle of supply voltage,
    provided C3 so amp's output can drive the - input to same DC level as + input,
    and added C6 to block the DC out of speaker.



    old jim
     
    Last edited: Jun 21, 2013
  17. Jun 21, 2013 #16

    davenn

    User Avatar
    Science Advisor
    Gold Member

    Thanks rbj :)
    I read this earlier and scratched my head .... then suddenly realised the error with my drawing
    DOH, I should have known better, I have built enough dual rail supplies haha

    OK here's the corrected diagram

    attachment.php?attachmentid=59756&stc=1&d=1371817394.jpg

    cheers
    Dave
     

    Attached Files:

  18. Jun 21, 2013 #17

    rbj

    User Avatar


    okay, so now my question is don't you need something (a small resistance) between where VCC comes outa the supply and before it goes to your audio amp circuit? (and similarly for VEE?) otherwize, the only way that your power supply ripple can be low is if the current draw is so tiny that the caps are holding the peak voltage coming outa the bridge (which means they only charge when the rectified voltages reaches this peak).
     
  19. Jun 21, 2013 #18

    davenn

    User Avatar
    Science Advisor
    Gold Member

    not something I have personally seen or done on any of my or commercial audio gear I have worked on


    Dave
     
  20. Jun 21, 2013 #19

    jim hardy

    User Avatar
    Science Advisor
    Gold Member
    2016 Award

    The ripple will be however much the filter cap decays between peaks. In tube days we had to add more stages of filtering
    but look at that LM1875's spec - 94 db of ripple rejection.

    if it's not objectionable to the listener ....
     
  21. Jun 22, 2013 #20

    rbj

    User Avatar

    i think that 94 dB of ripple rejection is the saving grace. because the more (DC) current you draw, the more the filter cap decays. those 100 μF caps on the chip (C6 and C7) help, but i think they would help a little more if there was a tiny resistance between them and the mondo power supply caps. in a tiny sense there already is some resistance.

    and i still think that 1Ω load resistance is awful small when the loudspeaker is 8Ω. even if that 0.22 μF cap is in series. at higher frequencies (i s'pose i should calculate it... looks like about 1 MHz) all the current goes down the 1Ω load. i guess we need not worry about 1 MHz, but if the spec or design guide says you can make it 10Ω, that's what i would do.

    and i might put two 4.7Ω resistors in series with the VCC and VEE supplies. at 100 mA, that would be just a 1/2 volt drop and it might really reduce the ripple that the chip sees. might make it a little cleaner.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Audio amplifier
  1. How to amplify audio? (Replies: 1)

Loading...