Opamps: IC version vs. discrete version

[Duckman]

I find this page http://www.bursonaudio.com/index.htm and I feel very intrested.

[Duckman]

OOps ! little mistake!

I find this page http://www.bursonaudio.com/index.htm and I feel very intrested. Can this kind of device make any difference in a stompbox level? I see opamps internal schematics in all datasheets, but there's no values for the components and I like to build some and make my own experiments. It's possible to find this values?

I'm curious!

Tkx for any information and/or comment

[briggs]

I've used a discrete opamp in my hummingbird overdrive. You can find one in the Boss Blues Driver schematic, they are great to play with

[Duckman]

Tanks, briggs! I take a look soon. So, how about to make my own 4558 "discrete" version? How can I do that? Any idea?

[UZILSD]

It's damn interesting indeed! But I think that it's more a thing for hi-fi audio folks...
I would try it in those pedals in which changing the opamp changes the sound audibly, all the tubesceamers for example.

Here there's a PDF with a complete schematics, but it's quite complicated and requires inductors:
http://www.johnhardyco.com/990OpAmpDetails.html

Here a much simplier version to begin with:
http://sound.westhost.com/project07.htm

The venerable 741:
http://www.play-hookey.com/analog/inside_741.html

Several projects:
http://www.sg-acoustics.ch/analogue_audio/discrete_opamps/index.html

In diyAudio forum there are several topics with schematics, such as this:
http://www.diyaudio.com/forums/showthread.php?postid=370353
But you got to register to download them.

Bye!

[Duckman]

Tanks, UZILSD!! There's a lot of info!! You're right about that is a kind of hi-fi matter, but seems interesting to me, anyway. I would like to find the way to make discrete versions of classic opamps and put in a stompbox to ear the difference... maybe it's a useless thing to try and this kind of devices don't make any audible difference in a stompbox, but... here we go!

Tanks again!!!

[Jarno]

Well, if a different brand of opamp will make a audible difference, changing from DIL to discrete will certainly be audible. I still have a bunch of small boards for discrete opamps, bought them in a groupbuy at DIYaudio.com for a hifi crossover. This might be interesting to test them, but these use 2SK389, now obsolete/ even when not obsolete hard to get/ expensive, double fets.

[Silent Fly]

I try to be Devil’s advocate here but why build an op-amp using discrete components when there are so many IC versions that are already built with matched components, temperature compensation, frequency compensation and everything they needs to work. This without considering that the IC versions are probably cheaper, smaller, more reliable and easier to replace/debug if something goes wrong.

I understand in the old days where the best the industry could offer was the 741. It can make some sense today if there is a high current requirement but for low-power audio devices, what’s the point?

Moreover, there are so may types/brands/models of op-amp that I can’t think to an application or requirement that it is not covered. The problem is that probably there is too much choice.

[UZILSD]

I try to be Devil’s advocate here but why build an op-amp using discrete components when there are so many IC versions that are already built with matched components, temperature compensation, frequency compensation and everything they needs to work. This without considering that the IC versions are probably cheaper, smaller, more reliable and easier to replace/debug if something goes wrong.

I understand in the old days where the best the industry could offer was the 741. It can make some sense today if there is a high current requirement but for low-power audio devices, what’s the point?

Moreover, there are so may types/brands/models of op-amp that I can’t think to an application or requirement that it is not covered. The problem is that probably there is too much choice.

I agree.
But I think that the interesting thing is the capability of building a "worst" opamp, from an hifi viewpoint. I'm thinking of putting germanium transistors in it for example, ot tweaking things in a non canonical way.

Hifi world and guitar gears world are antithetical. In a guitar amp even a clean sound is unacceptable distorted for hifi standards. What made and makes the electric guitar sound is imperfection.

Anyway, experimenting a little bit with discrete opamps is higly instructive I think, it would make one more aware of what there is in those little black boxes.

[Duckman]

Thanks, guys. Every word count!
The point to me is if any classical opamp, say the long venerated 4558, could be improved by a discrete version.

why build an op-amp using discrete components when there are so many IC versions that are already built with matched components, temperature compensation, frequency compensation and everything they needs to work. This without considering that the IC versions are probably cheaper, smaller, more reliable and easier to replace/debug if something goes wrong.

As I said before, gentlemans, it's all about curiosity...founded in a big "why not?" That's why I asked in the forum. There's a big theoretical world, but maybe someone got previous/precious practical experience to share. Silent Fly is right about market possibilities, but the point here is to get an idea and let it flow...

that the interesting thing is the capability of building a "worst" opamp, from an hifi viewpoint. I'm thinking of putting germanium transistors in it for example, ot tweaking things in a non canonical way.

...and tweak it, of course!

but these use 2SK389, now obsolete/ even when not obsolete hard to get/ expensive, double fets.

any replace?

Big Thanks to everybody!

[Jarno]

2SK389 are actually two 2SK170, and I think those are available as LSK170, not sure if LSK389 is also available.
Anyways, even on DIYaudio it was acknowledged that the primary aim of using discrete opamps was "nicer" sound, not ease of use, replaceability, cost. The same goes for stompboxes, I'd say. Try them, and toss them if it doesn't sound right.

[briggs]

I try to be Devil’s advocate here but why build an op-amp using discrete components when there are so many IC versions that are already built with matched components, temperature compensation, frequency compensation and everything they needs to work. This without considering that the IC versions are probably cheaper, smaller, more reliable and easier to replace/debug if something goes wrong.

I understand in the old days where the best the industry could offer was the 741. It can make some sense today if there is a high current requirement but for low-power audio devices, what’s the point?

Moreover, there are so may types/brands/models of op-amp that I can’t think to an application or requirement that it is not covered. The problem is that probably there is too much choice.

I agree.
But I think that the interesting thing is the capability of building a "worst" opamp, from an hifi viewpoint. I'm thinking of putting germanium transistors in it for example, ot tweaking things in a non canonical way.

Hifi world and guitar gears world are antithetical. In a guitar amp even a clean sound is unacceptable distorted for hifi standards. What made and makes the electric guitar sound is imperfection.

Anyway, experimenting a little bit with discrete opamps is higly instructive I think, it would make one more aware of what there is in those little black boxes.

I put a Ge tranny in the boss discrete opamp. Sounds great.

[Emanuele]

It has been days that I think of doing that!

I put a Ge tranny in the boss discrete opamp. Sounds great.

Did you substitute all those trannies in that boss discrete OpAmp or only one?
I recall there are two fets connected togheter and then a sort of buffer (still made of a fet)

[blue_lu]

so did this ever happen? what did you guys use to sub the ic ? did you build a discrete op amp yourself or used any premade pcbs (I didn't find any on the web, only prebuilt and very expensive).

sorry for the thread necro, btw.