My take on the LM833 vs TL072

[888guitars]

I built a Klon boost/ line driver and was swapping ic's. I thought that the hi fi 833 might be the ticket but it was so bright? I thought it was not supposed to color my tone? It made the guitar sound sharp in a goofy way. The TL072 May color the tone but I feel it does so in a good way. If someone knows of a better layout for the 833 please share it with me as I have not abandoned it but it sure does not work in this circuit.

Just my 2cents

Nate

[Jazz P Bass]

What you might like to try is find an 'old' TL072.
Like from the 80's or early '90's.

The current production IC's are too good. (if that makes any sense)

I ran into this issue on a Tube Works TD752 amp.
The distortion IC took a hit when the tube went south & replacing it with a current production IC did not float the customers boat.
I ended up robbing the TL072 from the reverb circuit & placing a 'new' IC there.
Happy as a clam was the customer.

[doctor]

The LM833 is the bipolar, and TL072 has JFET input. They cannot be used one instead another.

[Mark Hammer]

The Klon was not designed based on theory. Rather it was arrived at via trial and error, and thoughtful translation of Bill Finnegan's ideas and goals into concrete specifics by a consulting engineer he worked with. As such, you can expect the sound to be fairly dependent on the components, and values, used at each stage of the way. That's not to say that other components would not produce a pleasing sound to somebody or for some contexts, but Bill used the components he did because they sounded the way he wanted.

I wish he could have sold more of them himself, and seen more of the profit (the only money he made from units that go for $500+ on e-bay were the day he sold it new to somebody), but I suppose if you're gonna clone one, you may as well build it right.

RE: 833 vs TL072

The LM833 is intended for lower impedance sources and lower impedance loads. The TL072 is not. Each can outperform the other when used appropriately, and underperform when not used appropriately.

[888guitars]

Mark, I understand he had a specific design for the Klon. I was hoping to use the 833. I have yet to find a buffer/ driver layout for guitar. So if you know of a better way to go please share! The reason I want to use the 833 is from what I have read it is colorless? The Tl072 does have some color to it (not really a bad thing) but, if I could build a good pass through to compare that is the goal.

[Mark Hammer]

There is no more, or less, "color" to an LM833 (or NE5532) than there is to a TL072 or any of a host of other dual op-amps. If it is used in contexts that are "inhospitable" to its characteristics, maybe. If it is pushed to the perimeter of its characteristics (e.g., open loop or very high frequency input), maybe. But that's a matter of what's around the chip, than the chip itself. There is enough overlap between what an 833 "likes" and what a TL072 likes, that there should not be a difference. I've had the pleasure of having two entirely ungooped Centaurs provided to me by Bill Finnegan, along with a schematic, and can assure you that there is nothing in its design that would command fussy use of op-amps. Go for low noise, and you're good.

Much of the hoopla about op-amps in overdrives emanates from the Tube Screamer, and its derivatives, where clipping diodes in the feedback path interact with the particulars of the op-amp. Lesser amounts of hoopla emanate from the Rat and the MXR Distortion+, both of which were designed around the inherent weaknesses of the chips they employed. The Klon does not use diodes in the feedback path, nor does it rely on any weaknesses in the op-amps for its sound.

We tend to make too much of the difference between op-amps. Yes, there are a great many varieties, and those varieties exist because the chips are different in various ways. But there is tremendous overlap between them, and the many varieties tend to exist to serve very specific niches that are required for very specific contexts. If I need something that has the sort of low current requirements appropriate for an unmanned spacecraft, or can handle super-low impedance loads, or run off super-low supply voltages, there will be an op-amp for that. But there will be a great many applications where a wide variety of op-amps can be subbed for each other with no difference in performance.

The 833 and 5532 are capable of providing lots of current to low impedance loads, which is why they are used for things like headphone amps, and are also useful for feeding long cables, where more current drive is useful for combatting cable capacitance. It is possible to provide the same current drive by paralleling multiple op-amps of other types ( AMZ Super Buffer for Guitars ). The 833 and 5532 just happen to do so with ease, and without the need to use lots of parallel stages. This particular virtue has no relevance to any op-amp stages "early" in the Klon circuit, or in any other overdrive circuit for that matter. I use them because they're pretty quiet, pretty cheap, and sitting in my parts bins.

[loudthud]

Things happen in opamp circuits that you won't notice unless you have an oscilloscope, Without a scope all you know is that they just sound bad. A long cable can cause an otherwise "better" opamp to oscillate at frequencies over 1 Megahertz. You need a resistor between the output and the cable, outside the feedback loop.

The other major gotcha is driving the input too hard. At some point beyond the input common mode range, the current source for the input diff pair will be cutoff and the output will slam the wrong rail. See attachment below. It's a TL072 with +/- 15V supplies. Top trace is the capacitor coupled input at the chip, bottom trace is the output.

The Tube Screamer adds the input signal and the diode clipped signal by virtue of the "+1" in the non-inverting gain equasion.

[888guitars]

I'm pretty sure I am not clipping the front with just a strat?

Mark, I am having a hard time that the fet Tl072 doesn't sound different than a 833. .?.?

[Enzo]

Loudthud, I love the way you can tell a story with scope pictures.

[Mark Hammer]

Why exactly should it, unless the various components at its input and output are "all wrong for it"?

Although maybe we're talking past each other here. My sense is that the 833 might have slightly different noise specs than the TL072 in the Klon circuit, but the bandwidth of the guitar signal is so restricted that neither chip should be at any disadvantage reproducing the guitar signal, given the higher supply voltage used.

I'm not saying it can't "sound different", depending on what you attach to it, but that the chip itself has no particular "sound".

And I'm also pretty sure you're not clipping the front with just a Strat...not unless you have an onboard preamp, or super-overwound pickups, or another pedal between guitar and Klon.

Just as an addendum, I know that the Klon Centaur uses a charge pump to provide a higher supply voltage to some stages,and I know that some charge pumps can be limited in how much current they can provide. I also seem to recall that the NE5532 and LM833 can eat more current than some other chips. Any possibility that the sound you are hearing stems from the 833 exceeding what the charge pump can deliver?

[flyingdutchman]

I also seem to recall that the NE5532 and LM833 can eat more current than some other chips.

According to the Philips datasheet for the 5532, the supply current is 8mA per package (4mA per opamp) typical, 3 times as much as the 072 at 1.4mA per opamp typical (according to Motorola datasheet).

Anyway, using low-noise bipolars in instrument preamps is a bit pointless. Consider the requirements of a typical input stage. The source impedance of a typical instrument is in the 50k ohm range. Since bipolars have higher bias currents, the input current noise in combination with the source impedance, plus the thermal noise from the source impedance itself, will render the excellent voltage noise specs of opamps like the 5532 rather moot.
Also, the distortion performance of bipolar inputs tends to degrade when driven from a high-ish source impedance.

The higher input bias currents of low-noise bipolars also tend to cause considerable offset voltages if you want your amp to have a reasonably high input impedance - which for an instrument preamp should be at least a couple of 100k ohm and preferably much higher. This is not a problem when using the 07x - one can often omit coupling caps in places where they'd be absolutely required if you'd use a 5532.

Amps such as the 5532 really come into their own in circuitry where the impedances are at least an order of magnitude lower than those commonly seen in instrument preamps. For those who want to know more on these matters, "Small Signal Audio Design" by Douglas Self is an excellent read.

[Mark Hammer]

Uh-huh. See the article starting on page 8 here: http://hammer.ampage.org/files/Device1-8.PDF

Keep in mind this is 34 years old, with whatever chip choices there were at that time in history.

[888guitars]

Whoa! I gotta read that again! Looks like I'm screwing up the input impedance for the 833. I'm not sure noise is of huge concern and may be part of some charm like carbon resistors?