Now you're talking 😀
I once prompted PRR to defend the Jensen 990 discrete opamp, when I suggested one could do "better". And then there was the discussion of how to make a drop-in for the 5534 that was also "better". I put the quotes there because the constellation of constraints is complex and will depend upon the envisioned applications.
I may have related the story of an acquaintance who was aware of my design/development of charge preamps in the mid-late 70s. He was working for aerospace and was continually showered with free samples. He wasn't really all that swift when it came to fundamentals. He impatiently assured me that an OP07 (iirc) would work better than anything I could come up with, as the reps had assured him it was the lowest possible noise. I explained that it was completely inappropriate for the application owing to the high parallel (or "current") noise. He didn't want to hear the explanation.
When considering discrete and hybrid discrete/integrated amplifier designs, I like to pick things that are difficult to do monolithically. Otherwise, what's the point, other than catering to some notions about how, somehow, the integrated designs are, in Hansen's words, "dirty sand", and thereby prima facie not as good 😛
How's that sound, a DIYAudio discrete op-amp. "An op-amp even John Curl can use", has a nice ring to it.
Scott, would you like to 'critique' the Jensen 990. Did you know Dean Jensen? I did. Have you ever used a version of the 990? I have.
In fact, the Taiwanese, for an early Parasound preamp, more than 20 years old, adapted the general Jensen topology, adding an 'improvement' of using dual jfets on the input, instead of bipolar transistors. It was the first preamp that I worked on for Parasound. I have one in a closet, somewhere.
In fact, the Taiwanese, for an early Parasound preamp, more than 20 years old, adapted the general Jensen topology, adding an 'improvement' of using dual jfets on the input, instead of bipolar transistors. It was the first preamp that I worked on for Parasound. I have one in a closet, somewhere.
Now, if we are back to discrete design, let's first answer the question, do we need it to be an opamp? If yes, why? Which qualities of the opamp are required, which are not? 😉
A group design of a discrete op-amp is a great idea. First, choose the basic topology.
This thread or new? -RNM
This thread or new? -RNM
Last edited:
The 990 claim to fame was in the patent -- using small inductance in series with the emitter resistor, Re. The example circuit used to demo the affect on noise was pretty funky even for the times back then. Of course, the principle could be used in anyone's differential input circuit.
Last edited:
I listen to a version of the Jensen 990 Op Amp everyday in our Boulder 250AE amps used vertical biamp. They sound good to me!!
John
>A group design of a discrete op-amp is a great idea.
> First, choose the basic topology.
> This thread or new? -RNM
New thread please, so that it does not have to be mixed with all other on-going discussions here.
Topology :
JFET input, 2 stages, possibly SMD parts ?
Patrick
> First, choose the basic topology.
> This thread or new? -RNM
New thread please, so that it does not have to be mixed with all other on-going discussions here.
Topology :
JFET input, 2 stages, possibly SMD parts ?
Patrick
Dick Burwen did this in 1966?, I think Walt's history has a schematic.
EDIT - http://www.analog.com/library/analogDialogue/archives/39-05/Web_ChH_final.pdf AD Model 121
IIRC - When the patent was obviously unenforcable Deane placed it into the public domain (moot now).
Last edited:
I have no problem but this technique is violently opposed to the mantra here in that it increases dramatically the low frequency Aol as opposed to the gain at high frequencies.
Right you are, Scott, however a jfet input does NOT need the inductor to get the same noise (or better), because the gm is lower in the jfet compared to the bipolar.
Of course, Scott! That is part of the 'critique' '-) As a drop in module, it is pretty good, certainly better than any IC I have used. Constellation does not use them, and they are about as expensive and exotic as Boulder.
If I design an audio opamp using several IC opamps with local feedbacks, will it be considered as discrete design? 😀
OK, new thread. I will dig out my simple suggestion, you guys will have to help on the discrete NPN/PNP choices. I was going to use LSK170/74's on the input.
One must ask what are the performance advantages of ICs? Among them, intrinsic matching of adjacent devices; reduction of some parasitics. Deane used what is in essence an IC in the input stage, the LM394 pair, which is an array of interconnected devices. Good parts, although the same or better performance can be achieved in other ways.
Jensen's white paper on the design was instructive, and included the way that the compensation was done to extend bandwidth where it was thought useful. And as remarked, he thought he was the first to conceive of the use of inductors in the emitters to limit gain at high frequencies, but it was, like so many things, prior art, which (like they do more times than not) the USPTO missed.
Jensen's white paper on the design was instructive, and included the way that the compensation was done to extend bandwidth where it was thought useful. And as remarked, he thought he was the first to conceive of the use of inductors in the emitters to limit gain at high frequencies, but it was, like so many things, prior art, which (like they do more times than not) the USPTO missed.
Here is My challenge for a descrete opamp --> using any topology and transistor types (except expensive and/rare/obsolete/exotic types): no more than 8 transistors; drive 30 ohms or lower with no more than -105db harmonics, 20-20Khz; S/N ref 1v -unweighted to be at least -110db (input shorted). ; no use of dc servo, allowed - dc offset at output less than 2mV over time and normal use temp range.
How much better than these requirments/specs can be made??
RNM
How much better than these requirments/specs can be made??
RNM
Today it is not needed. I would take one low noise dual opamp and combine it with one more single fast opamp that has better output stage, adding few resistors. The result will be one opamp that you can't beat by discrete designed opamp.
- Status
- Not open for further replies.