are these ICs equivalent

[Groover]

I know nothing about op-amps. I am looking at a couple of schematics and would like to combine parts of two different designs. One uses "4558" ICs, one uses "TLO72" ICs. Are these equivalent parts, or is there a relatively simple way to adapt circuits built around one to use with the other?

[Alex R]

They're pretty well interchangeable, certainly in most audio circuits. The TL0 series have higher input impedance, is all I know.

[Mark Hammer]

They are functionally pin-for-pin interchangeable. You can generally expect any circuit that uses the one to perform as intended if you replace it with the other.

Having said that, there are differences in current requirements that may be key to the functioning of any given circuit. Depending on the circumstances, one will outperform the other in terms of noise. The 072 is usually quieter, but that reduced noise requires certain conditions that a circuit intended for a 4558 might not fulfill. The difference will not be like standing in a funeral home vs listening to the shower head close up.

Once upon a time the one used to be much cheaper than the other, but now they are equivalently priced.

[J M Fahey]

On a high gain stage (say around 100 X), you'll find the TL072 *very slightly* brighter.
If fed from a low impedance source, as in a transformerless balanced mic preamp, the RC4558 is slightly quieter, and "smoother".
The 4558 was "designed for audio"; the 072 was not specifically made so, but it is so good, that it excels there too.
Just my 2 cents.

[doctor]

On a high gain stage (say around 100 X), you'll find the TL072 *very slightly* brighter.
If fed from a low impedance source, as in a transformerless balanced mic preamp, the RC4558 is slightly quieter, and "smoother".
The 4558 was "designed for audio"; the 072 was not specifically made so, but it is so good, that it excels there too.
Just my 2 cents.

072 has JFET input stage and provides low noise
if fed from high impedance source.
It is defined as “Low noise dual OpAmp”.

4558 has bipolar input stage and provides relatively
low noise if fed from low impedance source.
It is defined as “General purpose dual OpAmp”.

For this reason it’s not recommended to replace 4558 with
072 or vice versa at all, and particularly when opamp deals
with low amplitude signal (when signal to noise ratio is important).

[nickb]

The TLC series have FET inputs (much lower input current) than the bipolar version 4558 type. Thus you cannot use a 4558 in place of a TL072. It's less likely to be a problem going the other way.

[Ronsonic]

In audio circuits they are more or less interchangeable. More or less. One may work better than the other.

In some other circuits the differences add up to the point that they won't work as subs. The switching circuits in many more recent Fender amps come to mind. They really want that 4558.

[dougw]

Beyond the previously mentioned input impedances, the most significant difference is the slew rates. the TL-7x family is around 12 v/us while my old rc4558 data sheet shows it at around 1.7 v/us. This really relates to how fast the device can track input changes. Interesting, both devices have the same gain bandwidth of 3 Mhz (way more than you need for audio). I am thinking is this higher slew rate along with the higher input impedance, which results in less damping, is why the TL-072 can sound a lot brighter than the rc4558. Just my $.02 worth and possibly wrong.

Doug

[Robert M. Martinelli]

If I was to sub a 4558 I would go towards the NE5532, which is pin-to-pin compatible and offers superior performance, especially when low noise is a concern ( which means always, at least to me ).

FWIW, I have recently built a couple of TS clones using 5532s along with metal film resistors. I also switched to 5532s on a couple of amps ( a Vox Cambridge Twin Reverb 30 and a Vox Pathfinder 15 Reverb ) I modded. In both cases, I'm VERY pleased with the results.

JM2CW

Best regards

Bob

[Enzo]

5532 are fine low noise chips, but they do use twice the current of the 4558. If you replace one or two 4558s with 5532, I doubt there would be any consequence. But if you decide to "upgrade" your 24 channel mixer with all 5532 instead of 4558s you may have a power supply problem when you are done.

I think a lot of these concerns depend upon the context. If we look at potential high end response, it makes more difference to a PA channel than to a guitar amp which rolls off after 3-5kHz anyway. And noise is a lot larger issue to an inout stage than to some line out driver.

[Robert M. Martinelli]

Enzo,
thanks for pointing that out, as always you're absolutely right, I didn't highlight that myself because I thought that, since we were talking about effect and pedals, the number of ICs to replace is usually low, so the difference amounts to a few mA.

Cheers

Bob

[nickb]

Some Data to consdier

Beyond the previously mentioned input impedances, the most significant difference is the slew rates. the TL-7x family is around 12 v/us while my old rc4558 data sheet shows it at around 1.7 v/us. This really relates to how fast the device can track input changes. Interesting, both devices have the same gain bandwidth of 3 Mhz (way more than you need for audio). I am thinking is this higher slew rate along with the higher input impedance, which results in less damping, is why the TL-072 can sound a lot brighter than the rc4558. Just my $.02 worth and possibly wrong.

Doug

Hey Doug, the slew rate affects the large signal performance. Gain bandwidth is a small signal parameter.

Hopefully, for the benefit of everyone, I have some real data (spice simulation) I'd like to present. I'll leave the interpretation of the acoustic effect to others. The test circuit is slanted to using a TL072 in place of a RC4558 as I contend that due to the much lower currents and noise you are asking for trouble if you try to put a 4558 in place of a TL072 but not the other way around.

I've attached files (I hope!) for the frequecy response, transient response, output impedance, input impedance, noise, THD and DC offset. Surely that's enough for anyone ( I know it won't be) !

The attached test schematic is supposed to represent a 'typical', non-inverting x10 stage. I know typical is up for debate but I doubt that it would affect the conclusion significantly.


Input Z: TL072 vs RC4558 Zin.pdf
Frequency: TL072 vs RC4558 Freq.pdf
Transient: TL072 vs RC4558 Transient.pdf
THD, Noise, DC: TL072 VS RC4558 perf.pdf
Schematic: TL072 vs RC4558.pdf
Outoput Z: TL072 vs RC4558 Zo.pdf

You can see that for most parmeters this is little difference as the traces are overlaid.

The only significant figure is the DC offset - much higher in the RC4558 due to the bigger input current.

Hope this helps.

[Mark Hammer]

Beyond the previously mentioned input impedances, the most significant difference is the slew rates. the TL-7x family is around 12 v/us while my old rc4558 data sheet shows it at around 1.7 v/us. This really relates to how fast the device can track input changes. Interesting, both devices have the same gain bandwidth of 3 Mhz (way more than you need for audio). I am thinking is this higher slew rate along with the higher input impedance, which results in less damping, is why the TL-072 can sound a lot brighter than the rc4558. Just my $.02 worth and possibly wrong.

Doug

I used to think that way. And while slew rate can make an audible difference when we're talking about a mixer, or other device intended to process multiple-source signals, or something with very wide bandwidth, if it's somethng intended to simply modify an individual guitar or bass, the lowly 4558 is more than fast enough to handle the limited bandwidth demand of such a signal. For the overwhelming brunt of electric guitar purposes, slewing faster than 1v/usec adds little or nothing in most portions of most circuits.

So, I'm not pooh-pooh-ing your suggestion, but it tends to be true only for certain contexts, and this may not be one of them.

A lot of the nuggets we tend to pick up about audio lore can come from one context (where they may be VERY wise and empirically supported), but then we foolishly apply them to other contexts where they matter not a bit, and occasionally may even be counter-productive. Been there, done that, bought the t-shirt, and already worn it out and waxed the car with it.

I strongly encourage people to read the article on bandwidth and noise in op-amps in this old issue of DEVICE that I have posted: http://hammer.ampage.org/files/Device1-8.PDF Yes, it is over 30 years old now, and the field has expanded drastically in the way of available op-amps, but the basic premise still stands: when it comes to noise performance in op-amps, context is everything.