DarkRain wrote:--the original TS808 schematic with NO modifications added.--
--that was what i was always said: the diodes will clip way before the opamp will--
--there is way way more difference if you are going to use different diodes--
That is obviously a big part of it. I would say that it is particularly some of the other circuit elements in combination with the chosen diodes that will also mostly or completely mask what is actually attributed to the op amp.
The 1N4148-type diode will begin to conduct with a forward voltage that's a little over 600mV. While I'm too tired to get to the bottom of it, research seemed to indicate that 808s had MA150-types, which have a forward voltage of closer to 1000mV.
While some may beg to differ, my own experiments have lead me to believe that even 700mV is potentially too low for what might amount to "the best soft clipping sounds" when used in a negative feedback loop configuration. There is also the phenomenon of having a unity-gain unclipped signal mixed in, and diodes that conduct sooner end up creating harmonics that seem to be buzzier.
The Zendrive was mentioned - it uses a combination of diodes in series that will give you forward voltages that are probably around 1100mV and 1400mV. There have also been reports that different opamps in that pedal will produce results that are a little more audible.
It could be that part of the reason we usually don't determine differences that are relevant to our own uses is because diode choice or substitution seems to be a sort of random thing, with pedals like the Zendrive being more of the exception to the rule. Popular diode combinations will typically fall either below or above the resultant forward voltages in the Zendrive, especially with almost all LEDs starting at 1800mV, and moving up from there.
Most likely, the other predominant reason that the opamp clipping is masked in the 808 has to do with its highpass and lowpass filter corner frequencies. When you are aggressively removing everything above and below 720hz at 6db per octave slope, the opamp isn't taxed with reproducing any true bass frequencies, and much of the harmonics are rolled off before the signal even reaches the tone circuit.
Again, using the Zendrive simply as an example, the voice control provides a means to alter the highpass filter corner frequency. The opamp's clipping characteristics might not be so masked when that corner frequency drops a bit below 720hz, and it probably gets to be more significant the lower you go. Since the Zendrive doesn't have a fixed lowpass filter on its output, but instead has a simple variable first order lowpass type, you can set it to remove little to no high frequency content.
I think that the Tubescreamer in original form is probably a poor choice for purposes of opamp clipping as is relevant to what a guitarist might find to be either overtly noticeable, or potentially pleasing. The Zendrive probably represents a much better foundation, but I think some things could even be made a little more generic for test purposes. I think that the voice control could basically be hardwired with a lower highpass filter corner frequency than a Tubescreamer, and some combination of diodes that conduct at some range between 1000mV to maybe upwards of 1600mV would be better choices. The Zendrive also has a simple unity gain opamp stage after the tone circuit, which is actually good - the tone circuit can work properly, and opamp clipping will be limited to IC1a.
I don't think that any tests done with all clipping diodes removed, and any frequency filtering removed, would prove to be significant for our purposes. But the Tubescreamer/808 is kind of at the other end of the spectrum as far as not really providing any helpful results, either. There needs to be a targeted attempt to get at the bits of the spectrum somewhere in between. The good news is that a very simple overdrive pedal design will probably suffice quite well.
a box full of JRC will keep you a steady income for a verrry long time