Op amps & signal loading?

[GibsonGM]

Any thoughts about this setup? Obviously caps, feedback resistors etc. are not shown.  I want to use the right-hand op amp as the main circuit, and the one to the left can be switched in to boost the input to the right-hand one.   I'm just not sure if the presence of the one on the left will load the right-hand one when it's the only one on (?).  I thought op amps had inpedances in the Mohms....but know nothing about how the OUTPUT would react.  I'd hate to have to use a 3PDT switch to cut the left one out...I want to use a DPDT and also have 2 LED's to indicate state.
Thanks

[R.G.]

Opamps with no other parts connected - that is, open loop, no input resistors - have input impedances from 100K or so up to many megohms. They have output impedances in the range of tens of ohms to hundreds of ohms.

When you close the loop with a feedback resistor, the (+) input is unaffected, so it's still high to very high impedance. The - input is held at what's called virtual ground by the feedback action, and will not change more than a few millivolts. That is one reason that the input resistance to an inverting input is just the value of the input resistor. The (-) input under feedback conditions looks like a very low impedance point.

With feedback, the output becomes a very low impedance point too. The opamp will spend all of its available gain and output current to make the output node be *exactly* what the inputs tell it to be. It looks like a fraction of an ohm in most cases.

Bottom line - yes, you have to switch the bottom output away from the first input if you do it that way.

However, there's the question - why do it that way? First, it changes from non-inverting to inverting when you flip the switch, and unless that's what you were trying to do, it could be a problem. There is a one-opamp polarity switcher show at GEO that's simpler. As to boosting the second opamp, generally opamps have huge gains already. You probably have enough gain in the main opamp to just open up the feedback loop and let it run a bit higher gain. But I don't know what it is you're trying to do, so that may not apply.

[GibsonGM]

Sorry about the confusing pic, R.G.; I do that sometimes!....the only important thing in there is the fact I want to switch in a 2nd opamp.  I want to drive the front end of a distortion circuit harder (think Dist+ with a boost on the front)...when I play, I like AC/DC type rhythm sounds, and GNR type leads - unattainable with 1 pedal, and I hate using them in series.

This seems the best way to be able to stomp in a brighter, higher gain sound.    Both inputs would be to the non-inverting terminals of the OA's.  The boost OA would have a gain pot for level.  I worry that the boost part of the ckt (left) would load the main distortion ckt (right). 
I'd also want to have some HP filtering on the boost output.  Or rather, should I do that on the input, so it's not operational when the boost OA is cut out??   Or just put the booster on the OUTPUT of the dist ckt??
Design considerations, LOLOL.   
Thanks for any input!

[slacker]

You could put the boost opamp in series with the main distortion on so it's in the circuit all the time. Have the boost opamp set up as a non inverting gain stage with your gain pot in the feedback loop. Then have a switch that shorts out the pot which will effectively make the boost opamp into a buffer. This shoul let you switch the boost in and out without any pops and the buffer shouldn't affect the main distortion stage.

[GibsonGM]

By shorting the feedback/gain pot, that will give me a unity boost, and therefore transparency? I'm also thinking I need to have a cap in there too - to cut some bass when I go to a lead sound.  That should work, I think - thanks for the input, slacker, I'll be looking into that 

[slacker]

Thats my understanding of how it works, shorting the pot gives a gain of 1 which is unity. If you adjust the cap/resistor combo between the - input and vref you can roll of some bass, like a tubesreamer does. This will only affect the sound when the pot is in the circuit, when the pot is shorted the response will be flat.

Hopefully RG or someone can confirm that

[GibsonGM]

Hey, thanks for that tidbit, slacker.  I'll try to plot out frequency response over the pot's rotation.  I modeled something like you suggest last night, and it looks good....simple OA booster with gain of about x5 seems to do what I want on the screen, I will breadboard it and try some things out.  Looking into moving towards treble with higher boost, and keeping it bassier at less boost, so the - input cap thing probably will work...

[GibsonGM]

Hi Slacker (or anyone good with op amps!)
What I'd like to do is use your idea of switching out the gain pot in the boosting opamp to set its gain to unity...but also to have it cut some lows (possibly something variable).  It seems if I try to do this, the filtering network needs to be somewhere on the input, and will always be involved in the circuit!  All the models for active high pass use the input cap/resistor to ground for this.

Do you (or anyone) have a schematic that shows this idea in action, or could point me to a link?
I'm looking for AC/DC thumping rhythm sounds with GNR screaming leads, LOL.

I think I need the filtering to be in the feedback loop, so it can also be shunted by a switch when not needed - make sense?
Thanks!

[slacker]

If you have a look at this image http://www.geofex.com/Article_Folders/TStech/tsxtech2.gif from the Geo technology of the TS page it shows how the Tube screamer cuts bass. In your case the input from the buffer would be your input and the out put would go to the distortion stage.
The 0.047uF cap and the 4k7 resistor set the point above which frequencies are boosted, this gives the bass cut sound because low frequencies are boosted less than high ones. If you short out the pot none of the frequencies get boosted so you should get a flat response.
Another idea to try is to replace the pot in that circuit with a fixed resistor and make the 4k7 resistor your gain pot. That way as the resistance on the pot goes down the gain goes up the roll off frequency goes up as well. So you get a bigger bass cut as the gain increases. If you then short the fixed resistor in the feedback loop you should get a flat response.

[GibsonGM]

Awesome, slacker..appreciate the heads up, that's where I need to be!  A little value tweaking should do it, I'd think. Looks like putting the pot where the 4K7 is will work great - 10nf instead of 47nf looks good on my simulation software, that's where I'll start switching caps.  You were right about shorting the feedback resistor for flat response