Opamps v Discrete kit

opamps v discrete

Do such devices exist and can they engineered to give high bandwidth audio circuits without any feedback whatsoever? I would certainly be grateful for a few names to check out if that is the case. My understanding is that all current devices are way off linearity but it nots really my field of expertise - I am just a dabbler. Obviously I am trying to minimise global feedback through the use of the composite approach or multiple loop discrete designs.

It sounds as if you are very familiar with the lithos - is it in fact a modified jung SR? The effect on the iso was profound - but then it was a phono stage, it might not have the same degree of effect on a preamp. I am looking forward to trying out a few options myself.

Thanks for the info!

 

Hi,

Well, if we accept that "feedback" really applies to "looped inverse feedback" (as opoosed to looped feedback which is often called positive feedback) which normally includes two (or more) active devices in series and a loop carying back the output signal to a point ahead of the output stage device, including it's input (I know, many qualifications), then yes.

The bandwidth in an audio circuit is simply down to circuit impedances and parasitics. If you want low circuit impedance you usually need a lot of current, which past a certain point is self defeating, as parasitics also increase as devices get larger chips for a higher current capacity.

As a rule, despite their many other failings BJT's are preferrable for wide bandwidth.

A combination of a fairly high current Emitter Follower and a Mosfet or J-Fet can be used, the emitter follower effectively buffers the input (miller) capacitance of the fet amplifying device, however the current must be high enough that at the higest frequency the circuit allows no slew limiting sets in.

LC Audio, Erno Borbeley, Nelson Pass, Jean Hiraga, Werner Walthers (Le Maison de l'Audiophile) and Pierre Johannet come to mind for learning excercises. They all have interesting open loop solid state designs in their portfolio.

The stuff done by Ayre and John Curl carries some interrest too.

The ones I have reverse engineered where in an Eikos CDP and they where not really Jung derivatives, rather more wierd....

If you have the time, take a nice open loop stage and compare the folowing:

1) LM317/337 followed by heavy duty RC filtering using optimised bypassing.

2) C4S current sources with TL431 Shunt Regulators passing at least the quiescent current of the circuit, preferably the expected peak load current PLUS the quescent current of the circuit, TL431 needs a fast (Os-Con or BG) bypass capacitor and application specific added bypassing by ceramic or stacked film (low inductance) metalised film mylars (metalised mylar because of the resultant higher minimu impedance tahn other types and thus better damping of resonances).

3) Anyones Hypedup Reg (R) according to their desired implementation.

I find depending on application 1 & 2 to be best to my ears.

Ciao T

 

Thanks - a lot for me to digest. Some way beyond my expertise I am afraid - being a humble theoretician and then not in electronics. Will check some of these things out... which would you recommend for daughterboard mounting - does the zap module offer best performance? What about the super-buffer circuit?

I assume lithos is still using opamps for correction - is it a better solution than jung and have you or anyone else posted further info online anywhere?

Thanks!

Nick.

 

Hi,

I have not tried all possibly solutions exhaustively.

In my DVD player I originally had SMD dual Op-Amp's at a time when good dual SMD's where as rare as Hen's teeth. Normally I replace Op-Amp's with much better Op-Amp's and leave it at that, or I rip them out completely and use Valves.

So I build the discrete J-Fet thing as it was closer to my philosophy. When I was given a Zap Filter by a friend (thanks Jus) I knocked up a "Q&D" supply similar to the original an swapped the module out. The Zap Filter was better. That's all I can say.

I would have to go back to be sure, but IIRC the IC on it was a Voltage reference (then again, that one has an internal Op-Amp...), rest discrete and rather simple, it was not your normal "Jung" or related at all.

Ciao T

 

hi 3d

your friend jus wouldnt be justin ( thermionic ) a very knowlegable man in pro circles would he ?

 

Hi,

It is Justin, but a different one, he posts on AA as jusbe....

Ciao T