Replace low noise opamps with "ultra-low".

I'm looking for ultra-low noise opamps suitable to replace TL072 and NE5532 (I don't mean that TL072 and NE5532 are the same). I found something about low noise opamps, but anything about ultra-low.

It doesn't matter the package, I care only that they are not surface mount.

Has anyone experience with this?

Thanks

5532 are very low noise, tl072 are not

mav_kbd said:

I'm looking for ultra-low noise opamps suitable to replace TL072 and NE5532 (I don't mean that TL072 and NE5532 are the same). I found something about low noise opamps, but anything about ultra-low.

It doesn't matter the package, I care only that they are not surface mount.

Has anyone experience with this?

Thanks

Click to expand...

You will probably want to use a different opamp to replace the TL07x as you use to replace the 553x.

I have used many TL07x and 553x but preferentially depending on the circuit application. Replacing with new uber opamps will lead you to at least two different replacement parts maybe more.

553x were used for their lower input noise and stronger output drive capability but bipolar input devices required consideration for input bias current and such. BIFET TL07x were used for general purpose, higher impedance with modest drive demands.

JR

 

The amount of noise produced by any op amp depends on the individual circuit conditions. For example, many of the ultra low noise op amps can only meet their low noise spec with a very low source impedance, due to their relatively high input noise current. In a lot of situations a 5532 would actually be quieter.  It's easy enough to work out from the data sheets, but it isn't quite as simple as internet wisdom often suggests.

All of the above, plus the fact that most of the time, ULN opamps have a higher quiescent current (if only by the fact that the input pair works at 1-2mA instead of 100-200uA). As a result the current draw may exceed what's available, so a redesign of the PSU is in order.
In addition, noise improvements rely on a very robust audio ground; very often it is necessary to improve the parasitic resistance of ground connections.

Where are they going?

What is the actual source resistance? (Most modders don't know.)

Everything has resistance. All resistances have Noise Power. Assuming no crappy resistors, Noise Power is the same for all resistances; but the ratio of Noise Voltage to Noise Current changes with resistance.

Some opamps have "excess noise" (low power or low bias current or low cost may trump low noise for many commercial applications). All that we use in audio have "ultra low noise" FOR some specific range of source resistances.

For sources much over 10K, TL07x IS lower noise than the circuit it is in.

For sources 1K-10K, 5532 has lower noise than the circuit it is in.

A few recent chips stand out in the 100-1K range. (However several of them are low voltage, so your resulting S/N may be worse; and some have 1/f noise higher than we want for audio. )

Analog Devices has done well, but searching their site for "ultra low noise" brings up hits for regulators (traditionally not as low-noise as they could be) and an audio opamp which is a fine thing but not lower noise Voltage than a 5532.

What you really want is Signal To Noise Ratio. It is often easier to find 3dB more headroom than 3dB less hiss. (However going to extremes increases secondary noise sources, particularly IM which is more insidious than simple hiss.)

Or I can sell you a magic cream.....

JohnRoberts said:

I have used many TL07x and 553x but preferentially depending on the circuit application. Replacing with new uber opamps will lead you to at least two different replacement parts maybe more.

Click to expand...

Actually I found some schematics of many mixer that use TL07x and NE553x in their preamp and EQ stages.

My idea is to build some preamps using these schematics (intended for home recording) and I thought to replace these opamps with something better. But I realize that is not so simple...

I thought to find something that can replace them without too much trouble.
For example, I read that the TL05x series can replace the TL07x in any case with better results.... that's possible?

It isn't that simple.. It is logical to assume that there are better opamps available today than the TL07x and 553x that were hot stuff 40 years ago, but they didn't suck, so replacing them will only make a marginal difference at best. If you do a poor job laying out the circuit boards you could forfeit any benefit you pay for with hotter parts.

If you are seriously interested in understanding the design process, noise alone requires understanding the difference and interaction between noise current and noise voltage and how even the feedback resistors contribute to the total noise. Then noise is only one aspect of signal purity.

JR