What is the cleanest Discreet Opamp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

hejsan

Well-known member
Joined
Oct 3, 2005
Messages
406
Location
Iceland
Ok, this info is probably spread around here somewhere,
I'm going to make a few discreet opamps for an active fader+pan stage that needs to be as clear as possible, and be unity stable.

I have not designed the stage yet, just kind of thinking of building blocks.
+-24 volts are prefered, due to other stages.

What are your thoughts?

Is the 990 in the public domain? I think I read somewhere that it is extremely transparent.

How about the melcor?

Thanks a bunch,
Hejsan
 
Never heard the 990, but you probably can't go wrong with it. Awful cheap, too.. Just buy a few from John Hardy. :cool: Or you could go for Forssell 993's. They're great but cost a bit more..
 
[quote author="hejsan"]that needs to be as clear as possible, and be unity stable.

Is the 990 in the public domain? I think I read somewhere that it is extremely transparent.[/quote]

I wouldn't assume that the cleanest op amp is a discrete one. But the 990 could very well be it.

The 990 is public domain. Deane Jensen published the design and made it clear that anybody could use it, commercial or otherwise, without restriction. Having built some discrete op amps myself, I tend to think that the John Hardy 990 is a fantastic bargain at about $45. You would be hard pressed to build a decent quality amp anywhere near that quality for anywhere near that price unless you want to buld a whole lot of them. Furthermore, he's a swell fella with a great attitude about building things and about sharing information. You can download everything you've ever wanted to know about the 990 and its applications right from the JH website. There's a great blurb in there somewhere about ceramic capacitors that every one of you should go read.
 
So what about the ones that we design and build ourselves....the no name ones. I would never buy an discrete opamp because I can roll one up in no time at all.

analag
 
I am very biased on this topic so think what you want of it. Due to the nature of design the JLM99v, Forssell 99X and the FetBloak/BigBloak are the cleanest ones I know. Speed is essential to transparent sound. Current feedbak opamps outpace, often by a factor of ten, any of the more traditional voltage feedback amps in audio applications. An exception, the Forssell opamps are amazingly quick in spite of being voltage feedback, but they have other issues. I think speed is why the Forssell opamps sound great and not because they are made with magic FETs although those parts help achieve that speed it can be done different ways.
There are many other parameters. One is distortion, but current feedback has the upper hand there too. It gives less distortion with the same number of gain stages as VF opamps.
The Melcor is the slowest and dirtiest of beasts so don't look for cleanliness there.
 
:)

Now think you can build one of those FETbloaks for less than $5. It is clean and wont clean your pockets :D

:guinness:
Fabio
 
[quote author="CJ"]If you use transformers with the discretes, there goes your "clean", so why sweat it?[/quote]

Nobody is forced to use thransformers unless one has transformer myopia. However, low turns transformers like the CMMI-2C and CMMI-3.5C are pleasantly clean. Even more suprising is that it makes a big difference even when using them with a 1:10 turn mic input transformer.
 
I've gone 'Bloak Crazy, myself!(oops! - did I really say that?)


They are SOOOO sweet, easy to build, and cheap, even with Darlinton outputs(Sound better than BD139/140 output to me.)


Just try a handful, 'Bloaks, 2520's, Melcor, 990, Forsell, JLM, whatever. The fact is they ALL sound great to me, but are better at specific aplications. I thought that the 990 sounded clean 'til I tried a BigBloak! MAtch the flavour to the specific Application. I have 4 different types of Output transformer(2503, Audix, Gardner's and Sowter, and I have now built them in seperate small boxes with XLR in/out, so I can mix and match for different applications. I also have several different "versions" of a API-312-type Mic-pre, with different types of Input Iron(Marinair, Belclere, OEP, GArdner's, Sowter, Beyer and DOA's - not so easy to swap about! This makes a really flexible range of sound - AWESOME!


I think that YOU should decide which is best by trying them, not asking for our opinions, no-matter how valid they are! Build with whatever you can get hold of, and try some others when they come along. I was seriously biased against the humble Melcor 'til I actually tried it! I like them just as much as the 'Bloaks, yet they are two ends of the spectrum effectively.

"Clean" is a subjective matter if you ask me. Some tube gear sounds "clean" yet runs at an order of magnitude more distortion than Modern chips, which can sometimes sound quite muddy. Some tube gear sounds muddy too! Even MAssenburg gear has a "signature", and the specs there for distortion, Slew-rate etc, are pretty much second to none.

As to transformers being dirty, well of coarse they can be, but I agree with wot TKHalmi sez. Especially the lower step-up part. My pre's with Marinair wired 1:2 with BigBloak's and Audix output trannies rank amongst the cleanest I have ever used, especially at the top end. Bottom is still nice and warm.Best of both worlds! (Definately a bit more grit at 1:4 tho'). I feel that I have finally found the holy grail! I will never be able to afford the Massenburg pre's which would probably be my ultimate comercially available choice for many different uses. Not sure I would WANT to now, anyway! Unless I needed 70dB of gain . . . .


Enjoy!



ANdyP
 
Hi, thanks a lot everyone for the info.

This fader/pan stage will be the last one before a 0 ohm summer,

The opamps will have to be unity gain stable. Are all those mentioned above stable there?

Also: How much would it cost to build a Deane Jensen opamp? and are all the parts commonly available? I could never buy opamps for 45 a piece. I need to many.

PS: I listened to the fetboy pre, and I found that way more colored than the 9k's or green's. Does that use the fetbloak, or is it just a coincidence with the names?

Do the fetbloak and 990 run on +-24V?

Thanks,
Hejsan
 
[quote author="hejsan"]I listened to the fetboy pre, and I found that way more colored than the 9k's or green's. Does that use the fetbloak, or is it just a coincidence with the names?[/quote]

They have nothing to do with each other, AFAIK. The only thing they have in common is the fact that they both use FETs, hence their similar names ("FET" being an acronym for Field Effect Transistor).

Peace,
Al.
 
One is distortion, but current feedback has the upper hand there too. It gives less distortion with the same number of gain stages as VF opamps.
Just a quick note on this one: assuming a unity-gain stable compensation for a given VF and CF the above statement is likely to be true for non-inverting medium and high closed-loop gains (say above 20 dB) and for all inverting configurations. It might not be true for non-inverting low closed-loop gains (e.g. a unity-gain follower) as common-mode distortion is usually much higher with CF. By the addition of a single JFET it is easy to improve the CMRR of the GainBloke by several orders of magnitude, if intended.

The opamps will have to be unity gain stable. Are all those mentioned above stable there?
With proper care (i.e. suitable external circuitry) yes. However it might be wise to arrange some gain (6 dB is sufficient) for several reasons (I have mentioned one of them above) by adding a voltage divider in front of the amp. For more detailed suggestions you'd need to post a schematic.

Do the FET-Bloke and 990 run on +-24V?
The 990 does. The FET-Bloak has IIRC 45 V transistors, but you can easily change them to other types.

The fun about discrete opamps is that you can tailor it to a specific application. If you have a detailed specification, I'd be happy to suggest a cheap-to-build design if interested.

Samuel
 
Thank you and a :sam: for all you kind people.

I've been reading through the gainbloak thread, and I'm becoming convinced of it's greatness, plus the coolfactor of using something designed by forum members :thumb:

[quote author="Samuel Groner"]... If you have a detailed specification, I'd be happy to suggest a cheap-to-build design if interested..[/quote]
Thanks, do you mean a cheap to build opamp or fading-panning circuit? I haven't found a fading-panning circuit to use yet, I think I'll start a new thread here at the drawing board to show my very naive schematic and hopefully the more experienced people here will point out why my circuit is hopeless :green:

The stage will be feeding a 0-ohm summer (LAWO 975, roughly the same as the much dicussed neumann V475, trafo in/out) and the summing resistors are 4k75 IIRC.
The stage will be fed by any output of my preamps, or soundcard, so the signal will be at line level allready.

cheers,
Hejsan
 

Latest posts

Back
Top