Discrete Opamp Open Design

[lineup]

Originally Posted by kf_tam

Hi Lineup,
Mr. Wurcer's latest is at #1568, with a slight mod to prevent latch up at #1776. The circuit discription is at #1675. In case you are wondering for a fully complementary frontend, that frontend is at #454, (though the symbols for P channel JFETs were wrong.)
Besides, he also published a FET input version of the JE990 discrete OPAMP at #1510
Thanks kf_tam,
your post is a blessing. Well done.
Now, I will go have a look

Regards,
Edward Tam

Thanks kf_tam,
your post is a blessing. Well done.
Now, I will go have a look

[lineup]

THis was found in Post #1568

[kf_tam]

Originally Posted by lineup

Thanks kf_tam,
your post is a blessing. Well done.
Now, I will go have a look

Hi Lineup,
Thanks. I have been waiting to see what you will add the party
Actually, I missed some preliminary sim results from Mr. Wucer but couldn't edit the post.

So here they are: N-ch JFET without cascode#1380 & effect of cascode on PSRR #1542,
N-ch JFET with Rush cascode #1387,
full complementary JFET #1397

The N-ch JFET version was selected, and cascode was added formally #1558, then additional compensation added #1568, JCX's idea of compensation was tested #1725.

Currently Mr. Wucer is testing the output section (diamond buffer) in real life #22032 (as oppose to stimulation only ) possibly out of 2N4401/4403.

BTW, a LTSPICE pack #1577 is also available for anyone to try.

Regards,
Edward Tam

[qusp]

its Wurcer, not Wucer =) and I think he prefers Guru rather than Mr.

[scott wurcer]

In the above Q3, 4, 8, and 7 can be ordinary 1N914/1N4148. I think your list of posts included the single 1N914 diode needed to catch a power supply sequencing issue (with it there is no need to worry at all). Depending on particular application not all the compensation components are needed and an output Zobel network should be considered for difficult loads.

[scott wurcer]

Originally Posted by qusp

its Wurcer, not Wucer =) and I think he prefers Guru rather than Mr.

Actually I prefer to drop the Guru . The name was originally German with an umlaut and a z. Such things were frowned upon by immigration.

[kf_tam]

Originally Posted by qusp

its Wurcer, not Wucer =) and I think he prefers Guru rather than Mr.

Sorry for this!

Being a former scientist, I prefer to remain completely neutral when reporting the development history

[lineup]

Originally Posted by kf_tam

Hi Lineup,
Thanks. I have been waiting to see what you will add the party
Actually, I missed some preliminary sim results from Mr. Wucer but couldn't edit the post.
------
Regards,
Edward Tam

What I can add is not very advanced, Edward.
But I have ideas ..... thats a start!

I have one solution regarding OPamp.
It is the link between the input differential and the VAS.
Warning!
I have not seen anyone use this approach before.
Not even John Curl.

This solution can we see simplified in figure #1.
And see it in a working OPamp in figure #2.

There are both benefits and drawbacks.
The benefit it elegantly takes the gain of input stage
and combine it in second VAS into one point.

The drawback is it takes some volts out of Vas stage.
But I like the idea very much.

Let's call it Lineup's JFETspecial.

[Chris Hornbeck]

See post #1884. One big issue with the series differential is the matching of devices and/or selection of the common source resistor. But a lot of potential for homebrew.

Thanks,
Chris

[aparatusonitus]

Originally Posted by lineup

I have not seen anyone use this approach before.

I have
Discrete Opamp Open Design
Discrete Opamp Open Design

And I have another one for you