Slew Rates And Gain Characteristics - IC Help Please

[frankie5fingers]

Please excuse my electronics learning mode.  I built the Serendipity.  It has lots of great tones but I'm having a little trouble dialing in the tone I want every time I stomp the switch;  one that I'll call the defining tone.  I read some posts on it prior and got a JRC4558, an OPA2604 and a 2134.  I didn't get the AD712 as the build calls for but the build doc allows for subs so that's what I ended up with.  I get that ultimately, it will boil down to what's most pleasing to my ears, but maybe some of the more experienced can help for next time.  I guess it comes down to two things, right - slew rate - the ability of the device to limit output voltage... and gain?  For the ones I have, the typical gains are: 2604 - 100 gain, slew of 25, the opa2134 has gain of 120 and slew rate of 20.  The 4558 shows a slew rate of 1(?) and gain of 100.  Should I be treating these like I do transistors; the comparing the gain figures as I would the hFe?  The Serendipity, although supposedly a low to medium gain OD, has tons of gain available with the OPA2604.  The AD712, if I read correctly, has a typical gain of 400.  So what's driving the breakup more, slew rate or gain - even though the 712 has 4 times the gain, it's slew rate is much lower, and the 4558 is almost nonexistant compared to the others.  So, where's the beef with these things?    Thanks

[oldhousescott]

For any of the opamps we use for audio purposes, the specs will be more than adequate for our projects. Even the 4558 with a slew rate of only 1v/us has an open loop bandwidth of 55kHz at 9vdc operation. Open loop DC gain is more like 100,000. The circuit configurations and, in this case, the clipping diodes used, will, I believe, play more of a roll in the character of the sound than the particular chip used. Not saying you won't hear a difference in opamps, but it will be the last 10% (or less) of the total sound (IMO).

[frankie5fingers]

Hmmm.  So Scott, you're saying tonally speaking, the IC plays a much smaller role than the diodes... interesting.  Then it follows that as I was asking, the "typical" gain figure and slew rates play a small part in the tone, right?  What spec should I be looking for when comparing the ICs then?
Then, since you mentioned it, what about diodes.  Once again, what spec should I be comparing when considering diodes?
I ask because at the moment, I'm building other's designs, tweaking them to my liking etc.  On the advice of a couple of Forum members, I'm reading up on electronics as often as I can, but while there's a wealth of information regarding performance in a given application, comparatively little pertains to audio - beyond the "we like it in this application" type of report.  At some point I'd like to design something myself, but while there will certainly be much trial and error, I'd like to use the posted spec but have a better feel for how that spec translates to audio characteristics.
Thanks again, for the info. 

[oldhousescott]

I believe where you'll hear the biggest differences are in applications where the output of the chip approaches the power supply rails, or even uses the chip itself for distortion. The internal topology of the chip will come into play there. For applications with diodes or transistors in the feedback loop, the output swing will be clamped by the forward conductance voltage of the diode(s)/transistor(s) added to the input voltage, usually well within the PS rails, and thus will exert the greater influence on the tone.

For non-distortion circuits, the input voltage and current noise figures of the chips, as well as their output distortion figures, will make the most difference, but degenerative feedback in those designs will tend to reduce the differences.

Ultimately, it's easiest to try several models and find which one you prefer in a particular circuit. Maybe not the most scientific method, but it works.

[stecykmi]

typically, we treat opamps as having an infinite amount of gain (at least open loop). This approximation (which actually works out to be very close) is valid because we use feedback loops to limit the gain to a value we want.

check out this thread from last summer, there was a good discussion about the characteristics of opamps: http://www.madbeanpedals.com/forum/index.php?topic=2468.0

[frankie5fingers]

That's a great thread, thanks Misha.  Scott I see what you're saying regarding the output/supply rails and essentially controlling them through the feedback loop.  That's what I was getting at, but everything isn't going to be OD or a distortion based effect.  That's where I was going and you put it well - thanks for that.
With transistors there are a few things like hFe that help us narrow down which would be more suited for a purpose, the ICs presented a bit more data, but I wasn't sure how to interpret it.  Thanks again guys, that's very helpful.
On to the diodes next!!

Thanks again, Frank

[culturejam]

Yeah, I wouldn't pay much attention to the rated gain of an op amp. You're just going to use feedback to set the gain anyway.