Unexpected DC offset at preamp input

Discussion in 'General Electronics Chat' started by Robartes, Jan 23, 2015.

  1. Robartes

    Thread Starter Member

    Oct 1, 2014
    57
    13
    I wanted to build a simple audio amplifier for use in my hobby room (which the rest of the house occupants refer to as 'cellar') and decided to use two Velleman kits to do so:

    K4003 power amplifier
    K8084 Preamplifier (schematic here at the end. Please note that the kit I was sent contains NE5532 opamps instead of the TL072 in the schematic. Also, you got to love the phrase 'comes complete with knobs'. It does, actually)

    I plan to build a discrete transistor amplifier later to learn, but as I needed something that actually would work *now*, I used the kits for this one. The power amplifier works beautifully, but the preamp has thrown up an issue that has me stumped.

    The basic signal path is Sonos ZP80 line out -> preamp -> power amp -> speakers. At low volumes, the system sends a loud buzz to the speakers until I turn the volume pot on the preamp to about 75% clockwise. With the pot between 75%-100% clockwise the buzz goes away and the music comes through. At first I thought I was looking at ground hum, but I don't think it was that. Well, I do have ground hum, but that is only audible when the volume pot is beyond 75% clockwise and no music is playing, so it seems to be something distinct from the loud buzz at lower volumes.

    In trying to find the source of this problem, or whether it is even a problem at all, I have been looking around the preamp with a function generator and scope and found something I do not understand. This could be a real issue, an issue with my test setup, or maybe it's not an issue at all and it's supposed to be that way, I don't know. This post is to get some input on this.

    In my test setup, the preamp board is powered with a split 14V (-14V, 0V, +14V) DC lab supply. A 1V peak-to-peak, 1kHz sine wave from a function generator is fed into both the left and right channels, and I have some power resistors on the output (though I tried without those as well, and the effects are the same). The scope is probing various input pins on the first NE5532 (IC1a):

    overview.jpg

    With the volume pot turned all the way to the left, this is what I get on pin 3 of the opamp (non-inverting input, connected to the wiper of the volume pot):

    pic_50_6.gif

    Channel 1 (yellow) is the sine wave input, probed directly at the crocodile clips of the function generator lead, channel 2 (blue) is pin 3 of IC1. It looks pretty much as I expected - flat line for pin 3 as the wiper of the pot is more or less directly connected to GND.

    Now watch what happens as I turn the pot to its middle position:

    pic_50_7.gif

    Channel 1 & 2 are still probing the same spots. Instead of the signal appearing as I would (perhaps wrongly) expect, I see a constant DC offset. Turning the pot lowers or raises the amount of offset, until I turn the pot to about 3/4 of its full range. Then I see the signal appear, with still 2V of DC offset:

    pic_50_9.gif

    Again, same probe spots.


    I'm surprised at the lack of signal and presence of a DC offset on the wiper of the volume pot for most of its range. From my understanding, I would expect the signal presented to the opamp to be symmetric around ground (no DC offset), and the amplitude to increase and decrease with the turning of the pot. Is the behaviour I am seeing instead expected for this circuit? Am I looking at an actual problem or is this an issue with my test setup?

    Thanks for looking!
     
  2. MagicMatt

    Member

    Sep 30, 2013
    117
    14
    That makes me wonder if your connection is good to the hot signal line and the grounded shield. I don't know much about your circuit I'm afraid, but the sounds you describe are exactly the behaviour you get from a PA system if somebody starts plugging in stuff to the desk without muting the channel first, and it usually happens when the signal line touches the ground on the desk.
     
  3. bertus

    Administrator

    Apr 5, 2008
    15,645
    2,344
    Hello,

    The links the kits do not work for me.

    How is the preamp powered?
    In the schematic is given a centertap transformer.

    Bertus
     
  4. Robartes

    Thread Starter Member

    Oct 1, 2014
    57
    13
    Which links - all three seem to work for me?

    In the test setup, the preamp is powered from a DC lab supply with +14V and -14V (to account for the two diode drop in the rectifier), in the full system it is powered with a toroidal center tapped 24V 60W transformer (actually two 12V secondary windings in series).

    Edit: corrected 60A to 60W for the transformer specs
     
  5. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,233
    I don't see where the second probe is but I might guess the output of the preamp.
    Make sure you have -14 at the negative of the op amp.

    Edit:
    You should have 3 leads from your power supply to the board - +/- 14 and ground.
     
  6. Robartes

    Thread Starter Member

    Oct 1, 2014
    57
    13
    For all of the pictures above, the second probe is on the non inverting input of the opamp. On the output, I just see a flat line:

    pic_50_8.gif

    In this picture, channel 2 (blue) is at the output of the opamp.

    Good point wrg to the opamp's supply voltage. I'll check that later.
     
  7. Robartes

    Thread Starter Member

    Oct 1, 2014
    57
    13
    And ronv nailed it! The negative supply of the opamp was at -0.3V. Looking at the schematic, it should be at around -9v (regulated by Zener diode ZD2 and pass resistor T2). Measuring the values of T2's terminals I got -13V for the collector, -0.26V for base and -0.31V for the emitter. Now I have a bad track record regarding transistor bias, but even I know that's wrong :).

    To confirm, I bypassed T1 and T2 entirely and put +9 resp -9V on their emitters and the circuit started behaving perfectly.

    It looks like T2 came DOA from the kit. I'm sure I have on lying around - let's replace it and see.

    Thanks!

    Edit: had collector & emitter measured values switched around for T2. Corrected.
     
    Last edited: Jan 23, 2015
Loading...