Missing Build Docs / Schematics

Vsupply is the max operating potential.

Vs=(V+)-(V-)

TL07X devices have a Vs_max of 42V. The 36V spec refers to the max on the input pins (i.e., common mode voltage and voltage differential), not the supply absolute.
I think I understand now. If pin 4 Vcc neg is connected to ground in the circuit, the input pins can handle the 36V.

Because Vs = 36 -(-0) = 36V

Is this on the right track?
 
Pins 8 and 4 are your positive and negative supply pins. The Vs measurement is the absolute sum of the difference. The input pin max deals with the common-mode voltage or differential voltage measurement. These are essentially measurements of the difference in voltage seen at the positive and negative inputs. In most circuits we’re dealing with (i.e., voltage followers and inverting and non-inverting amplifiers) this is not really a design concern since the two pins are typically equal.

NOTE: I need to correct my previous post. I thought something looked odd, so I went back to the datasheet. The 36V figure is for M-suffix (military), NS, and PS packages. Typically the differential figure is around -0.5–0.5V (which is the spec for the standard TL072).
 
I breadboarded the two schematics (Skeptical Buffer and my best guess at the Eunuch Buffer) and soldered up the Skeptical Buffer - and all circuits worked. The breadboarded Skeptical Buffer sounded similar to the PedalPCB board and the Eunuch sounded like the Skeptical Buffer with less headroom (as anticipated). Note: My bigger protoboards are of poor quality and can measure as much as 100 ohms of resistance across various holes! - which may affect my final products.

Granted, my observations of sound quality are merely subjective, I wish I had an oscilloscope to objectively confirm my schematics and the circuits, and I’ve only played the actual EUNA once. But I am confident that the schematics I drew up are mostly correct.

Below I’ve reposted the schematics (Skeptical and, once again, my best guess at the Eunuch) as well as my respective breadboards that are quite literally mounted on breadboards.
 

Attachments

  • IMG_7993.jpeg
    IMG_7993.jpeg
    619.5 KB · Views: 192
  • IMG_7994.jpeg
    IMG_7994.jpeg
    609.2 KB · Views: 227
  • A5FF38FA-FB84-40A2-85C5-D65B7EF0A927.png
    A5FF38FA-FB84-40A2-85C5-D65B7EF0A927.png
    60.5 KB · Views: 376
  • EunuchBuffer Schematic.png
    EunuchBuffer Schematic.png
    195.2 KB · Views: 365
Last edited:
Man the normal Pot and Kettle is so easy to understand - boost, clip, smooth. The tight switch is not bad either, as long as you know it only works with boost enabled.

Why complicate things with "Circuit mode switch" or "Drive mode switch" I don't understand. One enables the boost, one controls the clipping. Very simple.

I would also object to it having enough volume to "break my neighbors windows", although throwing the thing would probably do the trick.
 
Can anyone verify the Euna schematic as correct?

I shared the schematic on Reddit and on my social media, and a few other users confirmed that the schematic is correct and matches the (Skeptical Buffer) PCB.

No but it certainly LOOKS incorrect I'm afraid. What sort of frequency response are you getting from the simulations?

I think the confusion arises because people are negating that the op amp is being supplied +17V/-17V, so they believe the op amp isn’t being biased correctly.

Admittedly I too thought the schematic looked wonky especially the power supply but the voltages check out:

IMG_8016.jpeg IMG_8015.jpeg

Plus, the supply looks similar to that in the DODFX10.

I have breadboarded the schematic and soldered up the board and felt they sounded the same. But I don’t own an oscilloscope to confirm that they do in fact “sound” the same. Although, I think my SPDT diagrams may be backward, and ground should go to pin 3 (rather than 1).

I’m kind of curious why no one has attempted to trace the Skeptical Buffer PCB itself given it has so few components. When someone else double-checked my schematic against the Skeptical Buffer PCB, it only took them 30 minutes with a multimeter. (For example, I had C6, R8, and R9 mixed up.)

I am completely amendable to the notion that my schematic probably contains errors but I do think it provides a framework to find those errors and implement those corrections - or at least takes away the guesswork.

Edit. Here is the circuit in a simulator.
 
Last edited:
Is the skeptical buffer schematic ready to post yet so we can double check with the schematic posted here the parts around c6/r8/r9 seem strange to me as it forms an single pole low pass filter with reactive capacitance
 
Could we please get the skeptical buffer schematic so I can verify the build before starting to solder up I like to breadboard the circuits first to understand how they work
 
That's completely understandable.

Just wait for the documentation to become available before ordering the board, there's no pressure or rush.
I have been waiting for over 3 months. But I don't have much time, I'm not young anymore. as some. )))
I would like good sounds at a funeral.
 
Last edited:
Back
Top