1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  
     
    TalkBass.com has been uniting the low end since 1998.  Join us! :)

DIY Preamp: Incomplete

Discussion in 'Effects [BG]' started by EricssonB, Jul 20, 2012.


  1. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    So I thought I was on the right track with ripoff of a 3EQ preamp project, but I seem to have gone afoul somewhere. Figured I'd set up my out layout that goes in a mostly-clockwise rotation for easy of my brain. Spliced it together with a power section from a different pedal that I had built that works just fine with a power supply (for +/- DC), but I'm thinking that's where the problem lies -- it doesn't work with this one, seems to bleed a few volts to the output.

    Pushes out a reduced signal when the power is cut, but no signal other than an electric buzz when on. Difference between this project and the other is that the other uses basic +/- power, this has a bias section.

    Anyone who can stare at a layout and figure out the problem?

    c0004062_2322510.
    FaepboexIIIRender.
     
  2. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    Update: IT WORKS GREAT.

    Didn't connect pin 3 of the IC to the bias power. Oops.

    Yet another reason why I need to cease ALL WORK after midnight. Here's another.
     
  3. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    Works great, but hasn't hit the 95% threshold that'll declare it a success.

    After ~20 seconds of it being switched on, starts to distort. Switch it off, then back on, then it clears up. Any takers?
     
  4. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.


    Audio clip with all pots on full. Downloads enabled in case you want the subtleties not provided by the audio stream.

    Switched on at about 7 seconds.

    Just beyond the mid-point time warp that skips a minute you can hear how gated and distorted it gets. Not sure what that's about yet. Switched off for a bit then switched back on -- everything's okay.

    Any takers?

    Side note: this thing is pretty hot. May need to throttle it down a bit.
     
    Last edited by a moderator: Apr 16, 2014
  5. General Electro

    General Electro

    Mar 28, 2012
    Boston
    I don't have an answer, but maybe I can make some useful comments.

    The 2M2 and 1mF in the bias supply have a time constant of several seconds, perhaps related to the thing dying soon after being turned on.

    As a rule of thumb, resistors in such a bias supply should be ten times smaller than the other resistor that use the bias supply. The 560k resistor might really tend to supply a current that would totally dominate over the 2M2 resistor. Of course, the TL072 doesn't have a lot of bias current, so I can't say that's the problem. But I would suggest shrinking the 2M2 's to 56k, and making the associated electro caps 100 mF.

    Your build diagram has 4 electro caps, while the schematic has 3. So you haven't built the schematic. I can't really untangle the diagram.

    Surely you have a small multimeter? I would suggest watching the bias voltage, and also pin 1 and 7, and seeing if they drift. If pin 1 starts to drift, pin 7 will follow, but if the drift is in pin7, pin 1 might be just fine; could enable you to see where the error is occurring.
     
  6. I find it odd that you have it the input going into both inverting and noninverting. I've never actually seen that before. I'd suggest just connecting the noninverting to bias only.

    Also, I believe there should be a resistor in the negative feedback loop equivalent to the input resistor on the inverting input to set it to unity. I don't know if that's what it causing the problem but I think it's not helping either.

    For simplicity, a dual supply might be easier so you don't have to use half the supply voltage for biasing. Rather, you can then just use ground as a bias point. A dual supply is similar to setting the bias voltage for single but I personally prefer it for simplicity.
     
  7. Skrogh

    Skrogh

    Jan 28, 2010
    Denmark
    Darn.. wrote a long reply, then mistakenly hit F5. :[
    Well long post short: Gated sound is usual for a badly biased op-amp.
    Try lowering the resistors in the bias-network to something like 10k or 56k.
    Increase the capacitor in the bias network, 100uF or 220uF or just 47uF should be just fine. As General Electro said, try monitoring the bias voltage with a multimeter.

    Another point is: Get a breadboard, it will save you a lot of time, you can prototype your circuits, without soldering, then smack the whole thing together, and solder-bridges will be your most common reason for failure (and they are easy to get rid of, once you find them)
    Yet another point is: Register at http://www.diystompboxes.com/smfforum
    - While I quite enjoy threads like these at TB, I think you would get many more useful replies over there. :)

    This is a great place to start, there even is a very good article on breadboards: http://www.beavisaudio.com/

    Cheers and happy building!
     
  8. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    Also, noticed that pin 3 has 500k and a 10k between it and the bias and pin 5 has nothing -- are they supposed to equal each other, or are they independent?

    Seems like a natural next step.

    When it's working, it works great. Just can't go around switching it off to kill that buildup every minute or so.

    Edit: I bassically copied this guy's schematic, as I mentioned earlier. His relies on 9V, but I'm using a 12V wall wart that I just measured. Perhaps this one is touchier than my other project which was schematic'd at 15V.

    *looks around for a proper 9V*
     
  9. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    Makes no difference.
     
  10. boomertech

    boomertech Frank Appleton Supporting Member Commercial User

    Apr 8, 2009
    Syracuse, NY
    Designer/Owner of FEA Labs
    Your schematic should work fine with 9, 12, 15 up to 36V as long as the caps are rated for the voltage seen across them. You really don’t need the 10K resistor on pin 3 (you could replace it with a wire), but it wont change the operation of the HP filter if you leave it in there. Pin 5 is correct and the schematic is correct.

    Measure the DC bias voltage at pin 3 to ground and pin 5 to ground… they both should be ½ the supply. The outputs at pins 1 & 7 to ground should also have ½ the supply voltage. Measure the supply voltage for the op-amp. I think you have something floating.

    -Frank
     
  11. delta7fred

    delta7fred

    Jul 3, 2007
    England
    My guess is that you have connected the 12v to Gnd and +ve pins on your board and left the -ve pin unconnected. It should go to -ve and +ve.

    The input stage on the first opamp forms a 12db/oct HPF (can't remember off hand what type) and works just fine. The MM 3 band is one of the few preamps with a built in HPF.
     
  12. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    Without any instrument cables, I checked the voltages from -V and +V on the bottom right:
    -- +/-V is a cool'n'steady 16.38V

    -- +V to pin 3 starts out low then grooves on up to ~4.3V, sometimes as high as 5V.
    -- -V to pin 3 shows a difference of ~7.9V

    -- +V to pin 5 heads on up to ~7.5V
    -- -V to pin 5 can get up to ~-6.5V

    None of these are very steady. Seems like a problem.

    Could you please explain this a little more clearly?
     
  13. delta7fred

    delta7fred

    Jul 3, 2007
    England
    Ignore what I posted before, I have just spotted a mistake on your perfboard.

    The BIAS potential divider resistor, R11, is connected to GND, it should be to -ve.

    I have not worked out what purpose GND serves.
     
  14. boomertech

    boomertech Frank Appleton Supporting Member Commercial User

    Apr 8, 2009
    Syracuse, NY
    Designer/Owner of FEA Labs
    Good catch… I didn’t look at the perf board.

    Connect the GND and –V together and it should be good.

    -Frank
     
  15. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    So slide down R11 so'z it's between ground and -V, or just add another of the same value?
     
  16. delta7fred

    delta7fred

    Jul 3, 2007
    England
    Remove C10 and replace with a link, job done.
     
  17. boomertech

    boomertech Frank Appleton Supporting Member Commercial User

    Apr 8, 2009
    Syracuse, NY
    Designer/Owner of FEA Labs
    Remove your C10 and put a jumper wire there = connect -V and GND.

    -Frank
     
  18. delta7fred

    delta7fred

    Jul 3, 2007
    England
    Looks like Frank was posting as I was editing my post (I'm not fast on a keyboard).
     
  19. EricssonB

    EricssonB

    Apr 5, 2011
    CoSpgs, CO.
    Well by golly, this thing works just fine! Thanks guys!

    Now which IC to use on the finished product?

    Have TL0-62/72/82 on hand as well as NE5532. Sounds like I'll just have to figure out which one I think sounds best.
     
  20. boomertech

    boomertech Frank Appleton Supporting Member Commercial User

    Apr 8, 2009
    Syracuse, NY
    Designer/Owner of FEA Labs
    If you are keeping the 2M2 bias resistors I suggest you use the FET op-amps (062/072/082).

    -Frank
     

Share This Page