What do you think about LT1357/1358 opamps?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

lagoausente

Well-known member
Joined
Oct 2, 2006
Messages
288
Location
Spain
I think they are made for video. 600V/µs Slew Rate (still not sure what really means).
A respectable guy that make mods and sale own preamps recomend them to replace of 5532.
That toke my attention, and I´m curious that a search on this forum gives only results on threads were I mention them. So seem a rare use in audio, at least here.
The guy claims "you´ll hear much more clear on the high end".
Maybe it´s worth a try?
 
It looks like an OK part. Slightly noisier than 5532, and 600V/usec is a little more slew rate than you need (by about a factor of 100x).

It's only money, have fun.

JR
 
My guess would be that was Black Lion Audio. He was really into replacing audio opamps with ones designed for video. I did a couple of his mods that replaced TL084s with LT1363 and LT1359. They certainly sound better, but not much is being asked of them in that application.

In what application do you think the 5532 needs replacing?
 
Here's a drastically oversimplified explanation of slew rate.

Amplifiers with multiple stages and a lot of feedback, like op amps but also like some other circuits, are often unstable -- they'll break into oscillation. To stop the oscillation, compensation capacitors are placed within the amplifier circuit to roll off the gain beginning at a particular frequency.

When the amplifier is hit with a signal that has a very steep rise or fall time (a square wave if you're testing, a fast transient if you're using), an active device has to charge or discharge that compensation capacitor. (I'll call that active device a transistor, although some tubed amplifiers behave in the same way.) To charge or discharge it quickly, the transistor has to source or sink a good deal of current. If the amount of current needed is greater than the transistor can provide (or, in the other direction, the transistor runs into zero) the transistor will go into current clipping. Instead of following the fast-changing signal, the amplifier's output will ramp upward or downward in a straight line whose slope is determined by the size of the capacitor and the current coming from the transistor.

That straight-line ramping upward or downward is called "slewing", and the rate at which it ramps up or down is the "slew rate". During slewing, the amplifier is essentially operating without feedback and produces gross distortion. (Some writers call the rise-time or fall-time of actual signals the slew rate, but this is a misnomer unless the amplifier is actually slewing. If it's just going up and down without that transistor-capacitor overload, it's not slewing, so the rise- and fall-times aren't slew rates.)

Ideally, an amplifier should be operated so that it never slews. One way is to bandlimit the signal before it hits the amplifier, so the amplifier never sees those fast-rise-time signal slopes. Another is to self-bandlimit the amplifier by putting a capacitor in parallel with the feedback resistor.

It should be noted that while the above mechanism is by far the most common source of slewing, it can also happen when the amplifier is asked to drive a capacitative load like a long cable or, in the case of a power amp, an electrostatic loudspeaker. In those cases the output devices reach their current limits while trying to charge the capacitance rather than an internal transistor, but the effect is the same, and bad.

Peace,
Paul
 
finally, after 40 years, i understand slew rate, and thanks.
better late than never.
 
Just to put a fine point on the discussion, the way to describe the rate of change capability of a properly bandlimited audio path is "rise-time". Slew rate is effectively clipping in the rate of change domain.

The 1357 uses an interesting internal topology, and the slightly higher noise is due to compound darlington input stage and emmiter degeneration. I suspect it's an OK opamp, but the speed is beyond what you would ever need for audio.

Kind of like putting 600 mph tires on your car...

JR
 
Kind of like putting 600 mph tires on your car...

What does mph tires mean? I´m spanish .

what speed you consider is the max, would be necesary or audible, y slew reate spec terms? For curious. 6V/uS?
(by about a factor of 100x).



My guess would be that was Black Lion Audio. He was really into replacing audio opamps with ones designed for video.

Is not. It is Jim Williams from Audioupgrades. The mod suggestion can be found in a search in Gearlutz. Not dificult.

In what application do you think the 5532 needs replacing?

Not replacing, just want to try any one more that the 5532 in my supergreen project.
Any suggestion of other opamps to try?
 
ok 600 kph.. habla?

Jim reccomends several models, just today over at slutz he reccomended to replace 5532 with National LME49710, BB OPA134, LT1357, BB OPA211, OPA827.

regarding slew rate, enough to cleanly pass the desired power bandwidth. 6V/usec is probably adequate for opamp circuits but common opamps have been delivering 10+ for decades. Slew rate IMO is not really a design concern with good modern parts.

JR
 
[quote author="JohnRoberts"]It looks like an OK part. Slightly noisier than 5532, and 600V/usec is a little more slew rate than you need (by about a factor of 100x).

[/quote]

So, a garbage will be 100x deeper buried under the rug (20 dB quieter)... :cool:
 
[quote author="Wavebourn"][quote author="JohnRoberts"]It looks like an OK part. Slightly noisier than 5532, and 600V/usec is a little more slew rate than you need (by about a factor of 100x).

[/quote]

So, a garbage will be 100x deeper buried under the rug (20 dB quieter)... :cool:[/quote]

Nah...
 
Walt Jung propounded a Rule of Thumb which stated that for conservative design and audio frequencies, the slew rate of a device (in volts per microsecond) should be greater than or equal to the peak output (in volts). A more relaxed rule states that the slew rate should be greater than or equal to half the peak output.

So if you wanted to put out 10V peak (20V peak-to-peak) then by Jung's conservative criterion your device should be at least 10V/microsecond. By his liberal criterion, it should be at least 5V/microsecond.

Peace,
Paul
 
[quote author="JohnRoberts"][quote author="Wavebourn"][quote author="JohnRoberts"]It looks like an OK part. Slightly noisier than 5532, and 600V/usec is a little more slew rate than you need (by about a factor of 100x).

[/quote]

So, a garbage will be 100x deeper buried under the rug (20 dB quieter)... :cool:[/quote]

Nah...[/quote]
Yeah, I mean, some kind of a garbage...

Everyone has in mind own pyramid of preferred garbages...
 
Slew rate IMO is not really a design concern with good modern parts.

It´s curious. Jim often refers to "high speed" preamps. His own preamps have that name, and I think sells also small cards for mixing desks, also "high speed". Maybe a bit of marketing, don´t know, but references to slew rate are a constant.
The specs of the Boutique preamp is: slew rate 4000V/uS.
So, like putting a 4000 mph to a car? It´s just curious.
I think it´s worth put the theory in test. So I think I´ll order one 1758, the dual version to try and compare in one of the channels of my, still not fisnished supergreen.
[quote author="JohnRoberts"]
Jim reccomends several models, just today over at slutz he reccomended to replace 5532 with National LME49710, BB OPA134, LT1357, BB OPA211, OPA827.


JR[/quote]

I would like to know your opinion about this issue. Leaving the specs apart, do you think any opamp makes a significantly improvement? And what ones are your favorites?
 
[quote author="lagoausente"]
I would like to know your opinion about this issue. Leaving the specs apart, do you think any opamp makes a significantly improvement? And what ones are your favorites?[/quote]

I don't currently have a favorite opamp but find the linearity and low noise of some of the new models remarkable. The data sheets are a starting point, I'd need to put parts on the bench and test how they act in specific applications.

Sorry about the non-answer but there are many good opamps, and even the much defamed 5532 is ok in many applications.

I don't believe there are huge audible differences to be gained by upgrades from already decent opamps, when properly designed into circuits. Weak designs may benefit more from newer parts. YMMV.

JR

PS regarding silly high speed, perhaps its marketing... Audio does not go that fast. You need some margin but some very high speed designs may compromise other characterisitcs that do matter to audio. If all things are equal more is usually better, but be alert for trade offs.
 
You need some margin but some very high speed designs may compromise other characterisitcs that do matter to audio

for example?



but find the linearity and low noise of some of the new models remarkable.

What do refer to? recent versions of the same opamps like hte 5532, or other new?
I have not intention on speding much money on this, but would like try 2 or 3 opamps for curious.
Can you tell me some of that models you refer? I´m really lost about the market.
 
As I already noted the topology of the subject opamp actually has slightly higher input noise than the 5532 due to more complicated input configuration. The topology does not look as bad as some other opamps designed for sample and hold applications where they actually plan to operate in slewed mode during sampling.

I find several new opamps inpteresting but won't make specific suggestions as I haven't bench tested them myself.

The short list provided by jim Williams is a good place to start.. I just don't think 600V/usec is very useful.

Why don't you compare the data sheets for those few opamps, compare prices, and choose.. The 1357 won't be terrible, I just don't know that it'll be an improvement, especially noise wise.

JR
 
Sorry if I sounded discouraging. If you want to try different opamps to see how they sound different, have fun.

A lot of these properties are explained and compared by Doug Self:

http://www.dself.dsl.pipex.com/ampins/webbop/opamp.htm
 
Audio does not go so fast, but own opamp's errors do, so the faster they are compensated by feedback loops, the better is the result (though I prefer a different kind of errors, for which opamps are not the best solutions...)
 
[quote author="lagoausente"]

(though I prefer a different kind of errors, for which opamps are not the best solutions...)

What do you prefer?[/quote]

Completely opposite type of errors.
 

Latest posts

Back
Top