Some other Source Follower Configurations

[EUVL]

I have been thinking about some other configurations than the 3 classics published in Borbely's article JFETS : New Frontiers. I hope I did not re-invent the wheel by accident.

To avoid any trouble with copyright issues, here are the links to figures or circuits referred to in the article :


http://www.borbelyaudio.com/adobe/ae699bor.pdf

http://www.analogzone.com/col_01302004.pdf

http://www.diyaudio.com/forums/head...dphone-amplifier-post1130743.html#post1130743

http://www.diyaudio.com/forums/soli...rch-preamplifier-post1507926.html#post1507926

Taylor Source Follower

http://www.passdiy.com/pdf/zen-ver2.pdf


Enjoy. 😉


Patrick

 

[EUVL]

The article.


Patrick

 

[audiotux]

Hi Patrick ,

thanx for sharing this !
have questions about the Lovoltechs
can you please send an email ?

Greetings

Jürgen

 

[EUVL]

Someone asked about modifications to the DAO follower as built by Steen.

This is the version with the Taylor Current Source.
The 2SJ103 should have about 8mA Idss.

Depending on JFET Vgs variations, the resistor values might need adjusting. I should start with a 1k trimpot for R1 to set DC offset to zero first.

Hope you'll build this. Total dissipation about 15W.


Patrick

 

[h_a]

Hi Patrick,

that's most welcome reading! Thank you very much for your article!

By the way, thank you very much as well for updating your DAO! It's all my fault that another unit wasn't build yet as I waste a lot of time currently with toying with new amp ideas 😀

All the best, Hannes

 

[lineup]

Hi Patrick,
that's most welcome reading! Thank you very much for your article!
-----------------
All the best, Hannes

great article - interesting jfet amplifiers & stuff
I agree, Hannes

 

[EUVL]

This one for Steen.

🙂


Patrick

 

[lineup]

Patrick
Attachment: dao w taylor cs.zip
This has been downloaded 1 time(s).

hehe. Lineup was the first one to download.
I attach the layout in scale 1:1 (almost real size) .. in my 17" monitor

 

[EUVL]

But he already has the boards, working.

😉

 

[EUVL]

> I attach the layout in scale 1:1 (almost real size) .. in my 17" monitor

You have unfortunately posted the wrong layout. This is the old layout of the DAO follower without the Taylor CS. The one in the Power Point includes add-on components to turn the old layout into one with Taylor.

I hope you can delete this one to avoid confusion, or at least post the proper one with the 2SJ103 and R1 to R4. I tried to save the Power Point as JPEG but it gave very poor resolution, which was why I zipped instead.


Patrick

 

[Nelson Pass]

Thanks, Patrick

😎

 

[EUVL]

You're welcome.

I hope you don't mind me posting it here, and that you like the current sensing circuit, inspired by your Aleph CS patent.

😉


Patrick

 

[Nelson Pass]

Of course if you're dealing with high current, it becomes a virtual necessity.

When you're driving 10K, I've been able to tweak circuits with just two
2SK370's (one as a CCS) and no additional parts down to 0.0004% thd
+ noise at 2 volts. It doesn't leave me with a burning desire for improvement.

😎

 

[lineup]

You have unfortunately posted the wrong layout. This is the old layout of the DAO follower without the Taylor CS. The one in the Power Point includes add-on components to turn the old layout into one with Taylor.

just used you latest ZIP
if you have the wrong layout in the ZIPPED powerpoint
it is nothing i can do about it .. is there?

Regars

 

[EUVL]

> When you're driving 10K, I've been able to tweak circuits with just two 2SK370's (one as a CCS) and no additional parts down to 0.0004% thd
> It doesn't leave me with a burning desire for improvement.

Agreed. And I did mention the motivation for the alternative circuits in the article -- driving loads (e.g. passive filters) 1k or below.

> if you have the wrong layout in the ZIPPED powerpoint it is nothing i can do about it .. is there?

Just to be sure it wasn't my mistake, I downloaded the zip file from the forum just now, opened it, made it full screen, and printed that into a jpeg (sorry for the poor resolution).

And this is how it should look (see especially bottom left corner) !!!
Nothing wrong with the ZIPPED powerpoint as far as I can tell, probably something with your computer ? Or you are using some Windows Office equivalents ?


Cheers,
Patrick

 

[steenoe]

This one for Steen.

🙂


Patrick

Thank you very much, Patrick🙂 Cant be easier to modify the DAO with that drawing handy. ( BTW I didnt think that I would ever consider modifying the thing )
Your writing is hotstuff, thanks for sharing.

🙂

 

[EUVL]

Steen,

I sent you a PM yesterday concerning boards.
Did you receive ?

Patrick

 

[steenoe]

Thanks for the mail.
It had ended up in the spamfolder (In good company I might add; Paris Hilton, Britney Spears and CialisViagra and the likes )


🙂

 

[EUVL]

A brief update on the DAO with Taylor CS.

I have done some more optimisation using measured data from real devices, with the following adjustments to resistors values :

R1 750R (use 1k trimpot to adjust DC offset)
R2 4.7R 3W
R3 4.7R 3W
R4 41R (= 2x 82R //)


The distortion at 2.6V output into 60 ohm load (AKG 701) is compared to the original DAO Follower :

DAO vs w Taylor

Gain 0.926 0.959
2nd -89dB -67dB
3rd -87db -110dB
4th -99db -120dB
5th -121db -143dB
6th -151db -174dB
7th -188db -206dB

One can argue how accurate these values are, but IMHO a general conclusion can still be drawn.
If you want minimum total distortion, then the original DAO is already very very good. And the Taylor brings nothing extra. But if you believe in weighted spectrum. then look at the difference between the two from 3rd order distortion and onwards.

I know which one I would choose, but hope that Steen will give us a subjective comparison before long.
😉



Patrick

 

[audiotux]

Hi Patrick ,

is the Taylor CS possible with
sureaMBs version of the Dao?
(see the Dao thread)
he had used 4 Lovoltechs .

Greetings

Jürgen