This Week on the Breadboard: The Barber Small Fry (PPCB Tater Tot)

Yes I like the bass knob. It could also called tight I guess. That what it does to my ears, ccw tightens up the sound. As a bass control it doesn't have a huge range and doesn't really pump up the bass. In the overall picture it is quite usefull to have that on dashboard me thinks.
 
I made one of those. Great sounding unit. Is it dumble-esque? How can I know? ;)
During the make, I observed the following things:
  • the dynamics pot will have the wrong sense of rotation as per the kit schematic. This is annoying, since we are used to CW=more of something.
  • so is presence, but since this is no performance adjustable parameter, its not so bad. Particularly if you adjust from the rear.
  • During tracing a problem, I found that the output buffer's output idles at above 7V. Since all input voltages were correct, I suspected some large input offset voltage on the 4558 I used. Indeed this was the case. Replacing it with a contemporary unit (such as a TL072) completely fixes that. I attribute this to a) the age of the unit and b) the fact that the inputs are referenced on two points to Vref. An alternative fix is to reference the inverting leg to a rail (such as GND) via a large enough electrolytic.
  • Since there is so much talk, I tried a TLC2262 that was lying around, but this seemed to amplify some strange octave buzz coming from the GE diodes (I think). It was as if the signal were overlayed with a harmonic buzz that was awful. I finally settled with the trusted TL072 and be done with it.
  • I had used re-cycled OA... something germanium diodes. Big glass bottles.. Although they measured ok, their transit voltage was about 0.1V lower than some AA types I found. It appears that the GE diodes clamp so much that their tonal contribution becomes dominant. While this is bearable on some heavy gain settings, I do not like it too much with low settings. Lowering the dynamics too makes this buzz noticeable.
  • During debugging I temporarily removed the dynamics pot. I found that the "burn" setting is by no means totally equivalent to the larger burn unit, since the GE diodes are always a little in play. It is not as if they were taken out of the system entirely.
  • I changed the tone control (treble) to SWTC, which is a superior solution in many instances. I also relocated the volume control to before the buffer. This certainly is the proper way to go technically, but I don´t know if it makes much of an improvement.
  • I did not see an urgent need for lowering the signal into the output stage in order to prevent saturation. Even with dynamics full up the signal was too small. However, since the output signal was too big for them anyway, they subjected it to a 1:2 divider. I did the same thing on the input, but since the series resistances amount, a lower upper leg resistor (22-27k) sufficed.
  • I guessed a total of 100k will not load the tone generation section noticeably, and seemingly I was right.
  • Note that I reduced the output series resistor from 47k to 330R to serve as a built-out resistor (no additional volume drop needed). The grounding resistor can be 100k.
 

Attachments

  • circuit_fragment_sfbu.jpg
    circuit_fragment_sfbu.jpg
    50.3 KB · Views: 13
I didn't like that the 2nd stage had such a high DC gain. I obtained a 4559* and it worked with that. TL072 is another good choice. The opamp's input bias current creates a voltage drop in R10, which is multiplied by the DC gain of R11 and the MID & BASS pots. This caused the large offset you observed on IC1.2's output. As you said, a proper fix would be to disconnect the left end of R11 from Vref and return it to GND thru a 10uF cap. That reduces the 2nd stage DC gain to unity.

I agree that PRESENCE is wired backwards. It's also extremely touchy below 2 (9:00).

I used ITT GD012 Ge diodes. They are allegedly equivalent to 1N34A. With BURN at minimum and S1 in the center (off) position, they clip at 300mVp-p. With BURN dimed, they clip at 485mVp-p. Input is 200mVp-p at 250Hz. Because the circuit impedance around the Ge diodes is low, diode leakage is not much of a concern. I don't bother measuring Vf at some arbitrary DC current because that doesn't tell me what will happen in the actual circuit.

On my breadboard, I can saturate the 2nd stage if BURN, BASS & LEVEL are dimed, DYNAMICS is set for max resistance and S1 is in the center position. Another solution is to keep LEVEL at the output of the 2nd stage and increase R11 a bit.

* the DC voltage at IC1-7 is 5.00V with Vref = 4.49V.
 
Great. So this can be fixed on the commercial PCBs too.
The version I sketched works very well too. It is incredible how quiet this circuit is with that amount of clipping.
Switching is very quiet too. Thumbs up!

Edit: the bass knob BTW is very usable, and indeed needed.
 
Last edited:
You must log in or register to reply here.
Back
Top