Discrete Opamp Open Design

[scott wurcer]

Originally Posted by dadod

Better?

That's better, interesting. I think you are missing a couple of diodes on the right side of the input stage, this might saturate the cascode.

[dadod]

Loop Gain.

[dadod]

Originally Posted by scott wurcer

That's better, interesting. I think you are missing a couple of diodes on the right side of the input stage, this might saturate the cascode.

I used LED instead, that is 1.9 V, more then two diodes.

[EUVL]

Would you be kindly enough to post your LT Spice files ?
Then we can all play with it.


Many thanks in advance,
Patrick

[dadod]

Originally Posted by EUVL

Would you be kindly enough to post your LT Spice files ?
Then we can all play with it.


Many thanks in advance,
Patrick

If you meant me here it is.

[qusp]

Originally Posted by bcarso

I compliment those who use "complement".

ha, sorry for the laziness, autocomplete gives me that as I use it more often I guess and I havent trained myself to correct it. it also isnt set in my brain to use complement as a singular noun (by itself), or complementary as a noun at all… I know its correct, it just isnt habit. it seems to get used as a technical adjective, noun, or pronoun interchangeably.

[qusp]

pronoun? lol..no

anyway back on topic, I can see another model of custom jfet SMD heatsink coming on. I guess I better break out the matching rig and go through my j103/k246. would these compound mosfet networks be of any use here? i've been thinking about getting some to play with for a while.

fairly low voltage though

[kgrlee]

Originally Posted by scott wurcer

... Re: post #1380 circuit ..

Q15 just samples a voltage to drive the current sources for the output stage, this saves some devices and keeps all currents set by the input stage. It also provides a slew boost to the output devices, not much benefit here at band limited audio but it makes prettier square wave pictures.

Sorry about the bad JPEG conversion that dropped some 1 pixel wide lines, post #1387 has a schematic that caught this.

Don't the CCSs, J1 & J2 isolate Q6, 7, 8, 9, 11, 35 and hence Q15 from the signal path?

I can see a signal path through Q15 on some of the later 'symmetrical' versions but not on the simple () #1380 circuit.

Or are there some missing tracks on #1380

ricardo

[kgrlee]

Excuse, my ignorance but az i kunt reed en rite, I'm only on page 97 in my journey through these guru postings.

I get the impression this is a competition to be judged against Guru Wurcer's offering via LTSpice

From his #1380 post, the constraints are
- -100dB for 2nd, 3rd etc @ 20kHz with 20dB gain into 600R
- 2nV/rtHz noise
- zillion V/us into 10n
- 50R headphones too.

Please correct me if I'm wrong. I'd like to enter with my offering using NOS Mullard EF37As.

[fas42]

Many thanks to dadod for posting his version, I'm having problems with the output operating point, which is varying all over the show, up to nearly a volt offset. I need to play around with the models a bit more to get a handle on things ...

Frank