Having fun with op-amps

[vanhansen]

Since I've been doing this DIY thing for 1 year, I figured it was time to jump in and try to do something with op-amps.  I have some dual OPA2134PA's left over from when I modded my SD-1 and changed the chip, I also have the JRC4558 from the SD-1 and a few single op-amps, mainly LM741's.  Aside from the Simple IC Buffer from GGG, I haven't done anything else with them.

Well, here I am, attempting to do a boost and/or mild overdrive using the OPA2134's.  I think I read through 10 pages of search results, GEO pages, AMZ articles, studied the Shaka Boost schematic, other schematics with IC's, everything I can find.  Funny, but there's posts I read that say there's loads of IC boosts out there but I've only found 2, Aron's Shaka Boost and the MXR Micro Amp.  Where's the others?

Anyway, I have the voltage divider to get Vref figured out, no problem there.  Hehe, that's probably the easiest part of this. :D  Last night I whiped up something going in to the non-inverting input (pin 3), and then pin 2 to resistor to Vref and to pin 1.  Then I decided, hey, why don't I try to use the second half too.  This is when it got bad.  Something is missing but I can't figure out what, and on top of that, the first half sounds bad and I can hear it shut off when I cut off any signal going in to it.  There's a 1M resistor to Vref on the input, through a 0.1uf cap.  Output also has a 0.1uf cap. 

I'm uncertain on what values to use for the other Vref points (pin 2) and should the 1M input resistor go to ground and use another resistor for Vref like on the Shaka?  Is a cap necessary between pin 1 and pin 5 or 6?  When using the non-inverting input (+) first, should the signal go to the inverting (-) input on the second half?  I see different circuits out there doing different things.  Some use both non-inverting, some use 1 inverting and 1 non-inverting.

Man, can this get confusing.  If I try to read any more without getting some good solid info from it I think my head's going to explode. Of course, it didn't help that I was watching the hockey game last night while breadboarding. (Melanhead knows what I'm talkin about....lol).

[Paul Marossy]

Something that may help your understanding of opamps is the "Electronic Music Circuit Design" series written by Craig Anderton in Device Issues 7 thru 11. It's a series of articles on opamp design that's written in easy to understand terms. You can get them here: http://hammer.ampage.org - from page 10 on.

[johngreene]

Here's a good source for ideas:

http://www.national.com/appinfo/amps/0,2175,967,00.html

particularly this one:

http://www.national.com/an/AN/AN-20.pdf

--john

[Fp-www.Tonepad.com]

I second what Johngreene said. Those app notes are very educational and give you starting points and equations.

Fp

[vanhansen]

Wow....why didn't I find this stuff before?  Thanks guys.  This is great reading.  I've read the Craig Anderton articles, at least parts 1, 2 and 3.  It sure makes a lot more sense.  I love how he shows how to use the gain formula.  I know darn well that's my problem and I think I have some values that will make my circuit work.

I've been on a booster craze lately, making them with minimal parts, yet good sound.  I've done one with a JFET, there's alread they LPB-1 for transistors, I tried a tube based one, now it's on to op-amps. I'll get the first half of the op-amp working first and then proceed on to the second half.

I knew there was a reason I loved this place. Where else can you get help like this? 

[Paul Marossy]

I've read the Craig Anderton articles, at least parts 1, 2 and 3.  It sure makes a lot more sense.  I love how he shows how to use the gain formula.

Yeah, they're helpful articles.

I knew there was a reason I loved this place. Where else can you get help like this?

Hmm... nowhere else that I can think of. 

[vanhansen]

Here's an update on this.  After reading the Device articles about op-amps, I FINALLY got half of the op-amp working.  I'm using a Burr Brown OPA2134PA.  The gain is set to 19 with a 180k resister between pins 1 and 2, 10k / 10uf cap to ground off pin 2.  It's pretty much just like Aron's Shaka Boost.  I just didn't use a gain pot and a resistor after the output cap.  I like this.  Very smooth sounding.

What I'm thinking of doing is a mild boost with the first half, then more boost with the second half utilizing the second half of the op-amp in the same manner, but I'm just not sure what to put between pin 1 and pin 5.  Do I keep the 0.1uf input cap and then add the Vref resister and impedence resistor (1M)?  When I did that, it didn't work, even if the gain on the first stage was set to 2.

At least I've made some progress though. 

EDIT: Just made some nice changes.  Not as hot and a little more even across the spectrum.  150k gain resistor and 0.022uf input/output caps.  Sweet.

[Paul Marossy]

After reading the Device articles about op-amps, I FINALLY got half of the op-amp working.

Cool! 

[vanhansen]

In my layouts gallery (link below in my sig) I put up a schematic of what I did based off reading those articles.  It's simply named, OA Boost.  Please, critique my work. 

[Fp-www.Tonepad.com]

You have a LOT of gain in that one.

It's rarely seen that an opamp stage has fixed gain that high (15x). I'd add variable gain. Just move the volume pot to the negative feedback loop of the opamp (wired as a variable resistor) in series with a 47k resistor (or a 10k to make it go from unity to 11x). Note, the pot taper here will be important.

I ran a quick simulation, and that's 24dB gain stage. Any signal above 180mV makes it clip. (TL071). So pretty much any guitar plugged in will make it clip. Or add it in series with the 10k resistor already there, a la mxr microamp.

The choice for the output cap seems a little small and does affect the low freq response at the output. I'd change it to 1uF.

Also, a small cap (22pF) could be placed in parallel with the 150k resistor to lower the ultra high freq response. (hiss)

HOWEVER: if it sounds good, then who's to judge.

By the time you're  done with these suggestions, you will have transformed your OA booster into an mxr microamp (with different part values)!

Fp

[MartyMart]

Erik, There's some great info in those "national" papers
You do have an awful lot of gain going on there !
Have a peek at my "Mr Drive" schematic in the gallery.
That will give you some ideas, it's based on lots of the op-amp
OD's we all know, and it sounds fAB !
Tubescreamer/SD-1 etc etc are all good to look over, have you
read "the technology of the tubescreamer" ?
Lots of useful stuff in that article

EDIT: I think it was RG when talking about "gain stages" said that
the trick is controlling the maximum gain of them ( you'll soon have two ) and
the eq before/after them also.
So think of in/out cap values, how much bass gets "driven" ie: too
much will just produce "mud" !
How much "high end" is let through, so use of roll off caps and high pass
/low pass filtering, is all VERY crucial
As a general rule, I've found I/O caps of 47n to 100n about right for
gtr and 100n down to 470n good for bass use.
Just some thoughts.... babble ...... babble !

Marty.

[vanhansen]

Thanks for the input guys.  I knew it was a lot of gain.  What's cool with that much is setting the OD channel of my practice amp to have half the overdrive and then kicking this circuit on.  It pushes it over the edge.  I think you're right though, too much.  I've got lots of different values laying around to try out there, from 10k all the way up to 180k so I was just swapping things around.  I did like 68k in there though.  That gave it a gain of about 8 and was really nice sounding.  It might help if I run the output through a volume pot too. :D  I was going jack to jack with no pots anywhere. 

I don't have a scope or any simulators to see graphically what the values will do.  Any suggestions on where to get that kind of software?  I use Windows 2000.

I'm going to play with it more this weekend and post updates then.

Marty, yeah, I've read the "Technology of..." several times.  I think it's time for a refresher though.

Thanks for the help as always.

[vanhansen]

Ok, I'm confused again. 

Been studying schematics with notes all over my screen (no, not sticky notes for you smart alecs).  The ones I've been looking at are the Marshall Blues Breaker, Ibanez Tube Screamer and Boss SD-1.  Yeah, it's scary how they TS and SD-1 are practically the same thing... wow.  But anyway, something I noticed between the two op-amp stages of them:

Marshall:  in to non-inverting input, out to inverting input of other half.
TS and SD -1: in to non-inverting input, out to non-inverting input of other half.

Now, I know from reading the Device articles that the inverting input "inverts" the signal 180 degrees out of phase, but why would you throw the signal out of phase from the input like that?  Obviously it works, but why?  What's the point?   

[davebungo]

Phase is a relative thing - if you invert your guitar signal you won't be able to hear any difference because there isn't any reference to compare it with.  On the other hand if you reverse the wires on one speaker of your stereo hi-fi, it will sound awful (like your head is about implode if loud enough).

Absolute phase isn't generally important.
Relative phase can be crucial.

[Paul Marossy]

It is my understaning that phase becomes more of an issue with the effects that follow it. It's not so much a concern when it is functioning as a standalone unit.

[Johan]

Now, I know from reading the Device articles that the inverting input "inverts" the signal 180 degrees out of phase, but why would you throw the signal out of phase from the input like that?  Obviously it works, but why?  What's the point?   

..in a high gain pedal, making the output out of phase from the input can greatly reduce the chance/risk of squiling/feedback, and on a monosignal, like that coming from a guitar, it doesnt make a difference in sound...( the classic Marshall 1987 and 1959 plays out of phace and nobody worries about that..)
johan

[vanhansen]

Thanks for the info.  I understand the phase thing now.  More experimenting this week some time. Right now I have a 68k gain resistor on the first half.  It may change to 10k though once I get the second half in place.