Dual opamp equivalent of 741: LM1458, LM358, or something else?

[hannibal827]

Some background:

I'm working on a Maestro FSH-1.  Again.  I've already built one of these, but I'm not totally satisfied with it, so...
Anyway, to save space on the board, I'm going with Tonepad's strategy [project file here: http://www.tonepad.com/getFile.asp?id=138 ] of replacing the single opamps with duals and replacing the two CA3080s with one LM13600.

There's a thread over at "the other site" about the LFO opamp [IC3 in the Tonepad schem.] and how the TL0x2 opamps are more likely to produce ticking and, with certain NPN "noise" transistors in the Q1 slot, a high-pitched whine as well.  I'm currently using a TL022 for IC3, and while it doesn't tick any more than this circuit normally does, I do get the whining with most NPNs, and most NPNs that don't whine aren't sufficient to generate the randomly fluctuating CV that the sample + hold function requires.

The original circuit called for 741 opamps, and this thread I'm referring to suggested that LM1458 works well as IC3 in this circuit.  But it got me wondering, which dual opamp is the *actual* equivalent of the 741?  Is it the LM1458?  Or the LM358?  Or some other device I'm overlooking?  I did some [probably lazy] internet searching but can't seem to find a definitive answer.  In more than one place I read that the LM324 is the quad equivalent of the 741, but there doesn't seem to be consensus about the dual opamp equivalent.  Hence my question.

Does anyone have any ideas/thoughts/suggested readings about this?

thanks in advance.

[zombiwoof]

According to Analogman, the 1458 is the dual equivalent of the 741.  It's the drop-in replacement he uses in his DOD OD-250 reissue mod (which uses half of a 4558 in it) to make it sound more vintage.

Al

[R.G.]

This is a good place for a discussion of what a type number really means.

Manufacturers publish datasheets, which amount to a guarantee that whatever you buy from them with that number stamped on it will ** meet or exceed ** the performance characteristics listed unless both a minimum and maximum are listed. And in the manufacturer's view, that's all they will guarantee. Further, any part they make that satisfies that "meet or exceed" is justifiably considered worthy of the number being stamped on it and sent to a customer.

Note that this *completely* ignores anything not listed on the datasheet. EEs with little experience get into trouble by ordering parts, finding that the parts they received worked fine in some application, and then ordering zillions of same-numbered parts and finding out they don't work. When this is audited, it's usually found that the design depended on some quirk of the sample parts that was not listed on the datasheet, and between sample and volume order, the part maker improved the part or changed its manufacturing somehow, inadvertently invalidating the design.  Lawyers are very happy when this happens, and it happens over and over and over and over and...

The part numbering scheme for electronic parts turns Shakespeare's " a rose is a rose is a rose" upside down. It's like having a datasheet for a rose, saying "floral excresence, one of many colors, solid or shaded, containing the reproductive organs of the plant and having leaves and thorns on the stem." Under this specification, a blackberry bloom qualifies as a "rose", and the florist is justified in sending you one. It's even more fun when more than one manufacturer is involved, since both makers have only a set of numbers on a datasheet, and may well make anything inside completely differently. If it meets the numbers, it's a rose. Or 741.

Dual 741? Whatever meets the numbers.

And what we're interested in for pedals is usually the side effects that are not listed on the datasheet. Just to make it interesting.   

[hannibal827]

Al, thanks for the Analogman lead; I visited his page and found some good info. there.

R.G., thanks for your insight--

it's usually found that the design depended on some quirk of the sample parts that was not listed on the datasheet, and between sample and volume order, the part maker improved the part or changed its manufacturing somehow, inadvertently invalidating the design. 

--I'm wondering if this is what happened to the Maestro FSH / Oberheim VCF.  The S + H function seems dependent on a careful selection of the NPN and the JFETs [I'm working with 2N4303s in this circuit, as per the original schem, and they make a huge difference] and the opamp.  Could it be that part improvements and manufacturing changes just made this circuit, with all its variables, too difficult to manufacture in an efficient way?  I'm thinking, for example, about 2N3904s and how, seemingly, no one who tries to build a FSH-1 nowadays regards that as an optimal [or workable] noise transistor, but it was the specified part in the original, and they obviously made a large number of units 40 years ago that *did* work.  It seems like every relevant part has been "upgraded," but the success factor in building the circuit depends on the confluence of all these parts as they were in the early 70s.

Anyway, interesting and helpful replies, so thanks again.

[PRR]

The old-old dual '741 is the '747. I have not seen one in decades.

The '458 and '1558 are very much like '741 internally except duals, and "improved" (faster yet cheaper). I normally consider them "same", but they are not: one is popular for certain distortion effects, so there must be some small difference.

The LM358/LM324 is very much NOT a dual/quad '741, though in certain circuits it will work just fine. (Doing audio in low-Z loads it has nasty crossover distortion.)

To some people, the Toyota Camry and the Honda Accord are "the same". To me they are not. The Accord has a pocket in the trunk just where I've always kept my flashlight (through Willys, Falcons, and T-Bird); Camry doesn't. Yes, that is a fussy distinction. But much music-modification is about those very small/fussy details that chip-dweeps don't notice.

If you have a ticking whirring problem, you probably have to experiment.

[Keppy]

it's usually found that the design depended on some quirk of the sample parts that was not listed on the datasheet, and between sample and volume order, the part maker improved the part or changed its manufacturing somehow, inadvertently invalidating the design. 

--I'm wondering if this is what happened to the Maestro FSH / Oberheim VCF.  The S + H function seems dependent on a careful selection of the NPN and the JFETs [I'm working with 2N4303s in this circuit, as per the original schem, and they make a huge difference] and the opamp.  Could it be that part improvements and manufacturing changes just made this circuit, with all its variables, too difficult to manufacture in an efficient way?  I'm thinking, for example, about 2N3904s and how, seemingly, no one who tries to build a FSH-1 nowadays regards that as an optimal [or workable] noise transistor, but it was the specified part in the original, and they obviously made a large number of units 40 years ago that *did* work.  It seems like every relevant part has been "upgraded," but the success factor in building the circuit depends on the confluence of all these parts as they were in the early 70s.

In my FSH-1 build the transistor I chose was an MPSA18 that I removed from a production Dynacomp. Depending on which one I grabbed, I might have removed it to mod a unit, put it back in, and then removed it again, resulting in up to 4 separate solders, two of which were desoldering before I had my desoldering iron or anything approaching good technique.

I figured for a "noise" transistor, that kind of abuse was a plus. I'm not sure if I was right or if it was just coincidence that that transistor worked best.

I forget what my point was.

[hannibal827]

I just want to be clear that I'm asking about the opamp, not the noise transistor.  There are already plenty of threads about the noise transistor in the FSH-1.  I already have several different NPNs that "work" in my circuit to varying degrees. 

I'm interested to know which dual opamp is most similar to the single opamp that was originally used in the circuit. 

It could be that the circuit was really meant to have 741s in it--I'm probably going to try it that way if I don't like what the 1458 does.

[hannibal827]

The old-old dual '741 is the '747. I have not seen one in decades.

There are still some around: http://www.ebay.com/sch/?_nkw=747%20opamp&clk_rvr_id=377912790714
Pricey, though, of course.

[davent]

The old-old dual '741 is the '747. I have not seen one in decades.

There are still some around: http://www.ebay.com/sch/?_nkw=747%20opamp&clk_rvr_id=377912790714
Pricey, though, of course.

Bits and pieces here. http://rcfreelance.com/searchresults.php

Used Octopart search engine to get a supplier list. http://octopart.com/partsearch#search/requestData&q=lm747&start=0