MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Living in the past: op-amp substitution
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author Living in the past: op-amp substitution
ricko
I made this little chart to help organize my mind on op-amps, characteristics, substitutes/upgrades, etc. Maybe it could grow into a decision tree, but it is pretty rough. Anyway, perhaps someone else may find it useful.

So basically is a timeline showing the decade when a notable op-amp (in groups of three, single, dual, quad) arrived and some arrows to indicate some characteristic or usage relevant to synths. (Only op-amps with PDIP packages.)

I think many modern designs still use the op-amps that were available 40 or even 50 years ago: there is an inertia caused by old designs, and people using semi-obsolete internet material (or comment material from some irrelevant application, such as headphone amplifiers).

But there are so many op-amp types that without some organization, it is too hard to figure out why some op-amp type was used somewhere rather than something else: was it because it is really good or just because it was the best available at that time?

I guess the main thing that hit me was how bad the slew rate is of most early 70s op-amps. Using the slew rate calculators on the internet, at the kind of 5V or 10V pp signal levels in hot synths, they don't cope with the top octave of audio range at that level: at line levels (closer to 2V pp) they generally cope. (I think a slew rate of 2.5 V/us is the minimum for any audio jellybean op-amp used in synths, so that they can cope with 20Vpp (i.e. +/-10V swings) without drama or the designer having to think about levels too much.)
TOPLEL
Nince post!
What does the font and arrow colors mean?
ricko
An op-amp name is red if it can be run with a single supply: typically as required by effect pedals. (The grey are op-amps that people suggested on forums but which others said have are not so great. So I didn't want to emphasize them, nor omit them.)

The red section symbol alerts that it has a slow slew rate, which is a thing that seemed to be important.

What are the arrows? Left to right is time. In the 1960s, the 741 was the big success jellybean; along came the CA3140 for faster applications or ones that have high current out; then in the early 1970s along come the 412, 441, 358, 351 which have different sweet spot applications (speed, offset, impedance, single-supply). And these narrow what the 741 needs to do, so along come improved 741 (1458) and then TL071s with better general characteristics. But not optimal for audio outputs, so along comes the 5534.

And that point at around 1979, when the CEM/SSM chips had dominance, meant the effective freezing of the component sets used in analog synth circuits, because few people were making new circuits. Jump forward to 2019, and we have this wealth of circuits which completely ignore the progress in op-amps of the last 30 years: everyone uses old circuits or is more familiar with the old ones or has them in their drawers or let people save $1 on their $100 module. And, there is also diminishing returns: why cut down noise in the audio path when you have a VCA or LFG to clamp it off? If you need 20 op-amps for a phaser, you don't want to spend $20 per op-amp!

So there is almost no info on what can succeed the old stalwarts: so the arrows on the RHS are what seem plausible replacements in some situations. It would be great to firm these up to some small number of jellybeans.
BananaPlug
What? No Norton amps? A different branch of the tree I suppose but Serge used them for just about everything.
555x555
Hmmm...I like where this post is going, but I’m not quite sure that chart has the info I’d need to pick a go-to opamp. Like, I appreciate that you’re following the standard-usage model, this one is for audio, that one for summing cvs, etc. but I think this standard usage is maybe part of the issue. The TL07x gets used so often in part bc although it’s not the ideal choice for everything, it will work for pretty much everything (except rail-to-rail). This is in part also I think why the NE5532 which is supposedly amazing “for audio” sees so little use in synth design: it has a low input impedance and so...well, sometimes that matters and sometimes it doesn’t so one has to be able to understand the circuit requirements. Or one can not understand them and use a TL07x.

TI’s site is very searchable once you know what characteristics you’re actually looking for. But you may find that, if you want a through-hole with a particular set of characteristics you end up with one of your “70s/80s” opamps. Idk. I’m somewhat less experienced than others but, well, is inertia really the reason for the predominance of some of these “old” chips, or are they actually the right choice given all the factors?
ricko
@banana Wot no Nortons?

Norton amps are a good example of the issue. The 3900N from 1972 has circuits from that golden age and awareness. The LM359 comes along in 1982 and it is too late. Who uses that? And is it really because of technical differences, or the price gap, or just because Serge et al used the 3900 so no-one has heard of the lm359? But should we be switching our Serge circuits to use it?
ricko
@555x555 For audio signals, I think the old chops are rarely the right choice, because "right" is set by our expectations and our fantasies that if we have the same guitar as Hendrix we will ge as creative as him. (...or Zawinul, or Subotnic, or Carlos, etc.)

I remember whrn I replaced the mixer 301As in my eti4600 (with tl071s), how astoundingly better the sound was in every way. And, again, upgrading some crossovers to use lm49710s not the tl072s I put in for testing purposes. That changed my expectations of what more synths should sound like.

So, no, I dont see why we should accept that using the cheapest antique jellybean opamp for audio in every position in every synth module is good enough. Nor should we just accepf thst because 32(!) years ago Doug Self exposed how good the 3354 was, that is the end of progress and improvement, for thst matter. (And, yes I know that Self coupled the old and new in his 2012 elektor preamp with 5534 plus lm49710. And yes I know thst some people say that you dont need to have better components in your audio chain than the worst downstream one, which makes no sense to me because noise and distortion are cumulative.)
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group