Op Amp Noise Ranking? (4558 vs 1458 vs 741)

[Steve Mavronis]

I want to clone a DOD 250 Overdrive and/or mod my YJM308 Overdrive pedal and one of the steps (besides changing  the input cap) would be to replace the stock 4558 Op Amp IC to get closer to the original pedal's sound. The 1970's pedal it was modeled after used a 741 chip. The newer 4558 currently onboard could be easily changed to a 1458, which is a dual 741 with the same specs. If I replace the 4558 with the original circuit's 741 it would require some PCB trace cuts and re-wiring because of the IC pinout differences. But to account for everything I'm curious to know how these 3 Op Amp's noise levels compare to each other, especially if the 1458 is less noisey than a 741 or not, so can someone give their opinion and rank these chips noise levels for me from highest to lowest? Thanks...

[PRR]

All three chips are very nearly the "same" input topology and current. The fundamental noise level will be too close to call.

All three chips have been in production for 30 years at dozens of factories while semiconductor fabrication has improved vastly. Back in the dawn, you did occasionally get a hisser, microscopic dirt on the wafer. Later wafers became essentially perfect. But these are low-end chips, the high-end chipmakers have moved to fancier products, new low-end chips are cranked out on older machinery with what is now considered lower-grade material and junior workers. The economics of chipmaking suggest that hissy parts are rare (excess hiss suggests dirt which always suggests excess failures which is bad for business); still you could get a less-good chip.

But aside from weeding-out the 1%-2% excess-hiss parts, I doubt there's any noise improvment possible via chip-change.

> would require some PCB trace cuts

Nah. Find an 8-DIP header. Get a 16-dip socket. Bend wires one to the other so that you put two 741s in the top 16-DIP socket and the bottom header fits a "4558" socket. But I really doubt it is worth it in this case.

A lower-noise product must consider the working impedances. If sources (both inputs) are under 5K, the 5532 is often a good bet. If they are over 50K, the TL072 will make less noise. But the vast majority of audio circuits are in the middle. And in this zone the 741 and its duals do OK. I've had impressively low hiss from them. If a clean 741 is not quiet enough, you probably want to change the plan somehow.

Of course 741, 4558, 5532, and TL072 all "distort" different. 741 won't slew fast. TL072 can go bonkers when input crosses the rail. Because they are not intended to be overdriven, the way they recover after overdrive is different. All these "small" differences may (or may not) matter in your use.

[Steve Mavronis]

The goal of this pedal mod is to get the 741 sound and the 4558 sounds too different. So I was wondering about the 1458 noise level compared to a single 741 since it "should" sound the closest (being composed of two 741's but only one half is used) without having to modify the traces of the YJM308 PCB. I know these chips are noisier than the other one you mentioned but wondering if the 1458 is any noisier than the 741 or about the same. Here is the PCB layout that I made with ExpressPCB for the YJM308 PCB. I got the mod instructions at the bottom off another forum but I don't really agree with replacing C1 because the grey 250 circuit I'm doing uses the .047uF cap and I want to stick to the spec even though it's a minor difference:

[anchovie]

http://www.analogman.com/dodmod.htm

Note the bit about the Texas Instruments 1458.

[Steve Mavronis]

Yeah I was reading that where it says:

There is a higher noise level. I test background noise by setting the studio noise gates to just barely cut sound when guitar volume is rolled to zero. When the gates are set to just go quiet with the 741 pedals, there is still a rush of HISS that comes out of the system with the 1458 pedals.

But that doesn't say how either compares to the stock 4558 that comes with the YJM308 pedal? But I guess I'll go with the full 741 mod (instead of the easier 1458 change) since it appears to be the lesser of the 2 evils noise-wise.

Also the mod instructions I got say if switching to the 741 that one of the steps is to cut the trace coming from IC pin 1. Pin 1 is an offset null in the 741 but is one of the outputs in the 4558. But I saw a photo of someone's YJM308/250 Analogman 'mojo' mod with a 741 chip that has a jumper wire going across from pin 1 to pin 6. I guess their mod doesn't cut the trace and the wire re-routes the op amp pin 6 output to the pin 1 pad. I guess it doesn't hurt anything if also touching the 741 offset null pin which is not used in the original circuit:

[anchovie]

It does say that the 4558 has better performance, which is why they included the 25pF cap to roll off the highs that were being inherently lost by the 741. The 4558 is also marketed as an "improved low noise dual replacement for the old 741", though as with the best marketing speak it doesn't specify whether that means "lower noise" or "some things have been improved, plus it's still low noise".

I suggest you put a 1458 in your YJM308 and find out if your ears really can tell the difference!

[Steve Mavronis]

It does say that the 4558 has better performance, which is why they included the 25pF cap to roll off the highs that were being inherently lost by the 741. The 4558 is also marketed as an "improved low noise dual replacement for the old 741", though as with the best marketing speak it doesn't specify whether that means "lower noise" or "some things have been improved, plus it's still low noise".

Well the noise is one aspect but the sound is more important. I thought the 25pF cap was an attempt to roll off the extra highs produced by the 4558. But they were supposed to be aming for the 741 sound, so how could you say highs were lost by using a 741 which was in the original 250 overdrive? Even Analogman removed the 25pF cap when using the 1458 or 741 chip in their mod.

[anchovie]

Sorry, I may not have phrased that quite right. If the 25pF cap isn't needed with a 741, then surely that means that those highs are already being rolled off when a 741 is used?

[Steve Mavronis]

Apparently, relative to the 741 being your benchmark starting point that is the goal to get the YJM308 and re-issue 250 to sound like again.

[Steve Mavronis]

I saw a photo of someone's YJM308/250 Analogman 'mojo' mod with a 741 chip that has a jumper wire going across from pin 1 to pin 6. I guess their mod doesn't cut the trace and the wire re-routes the op amp pin 6 output to the pin 1 pad. I guess it doesn't hurt anything if also touching the 741 offset null pin which is not used in the original circuit?

I couldn't modify my original post any longer so here is the interesting pic that I mentioned above (from a Japanese site) about the Analogman 250 mod using IC pin 6 to 1: