Borbely ALL-FET DAC I/V stage

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm new to this forum so the answer to this query may be in its archives. Apologies in advance for any redundancy in this message.

In these articles (<http://www.borbelyaudio.com/adobe/borb502.pdf>, <http://www.borbelyaudio.com/pics/305borbely-new.pdf>), Borbely describes an ALL-JFET line stage, provides schematics and part values.

I'm interested in building the ALL-FET Dac audio board <http://www.borbelyaudio.com/allfetdac.asp>. Some schematics are provided on this page, but no part values are given. Can anyone hypothesize what the values may be? (Even the Starter Kit version of Borbely I/V may work for me: <http://www.borbelyaudio.com/eb604417.asp>, but again, no component values are given.)

In any case, is/are there an equivalent or better I/V design(s) in the diyaudio.com archives that may work like (or better than) the ones at borbelyaudio.com?

Please note my criteria for DAC I/V:

- No vacuum-tube-based projects, as I prefer smaller, more-portable solutions.
- Should work for a variety of DAC chips (e.g. TDA1541, TDA1543, TDA1545; and PCM1704, PCM1794)

Thx in advance for any help you can provide.
 
This is a balance version Jfet IV modified from NP D1.

R9/10 adjust output stage gate/source at 10V.

R11/12 adjust input offset at 0V or desire value.

Adjust R3/4 for different DAC out current. 1.5k is good for PCM63 and TDA1541.

Use half will be Single ended verrsion.

I have good result with that.

Spencer
 

Attachments

  • jfet iv 2.pdf
    16.4 KB · Views: 2,565
spencer said:
This is a balance version Jfet IV modified from NP D1.

R9/10 adjust output stage gate/source at 10V.

R11/12 adjust input offset at 0V or desire value.

Adjust R3/4 for different DAC out current. 1.5k is good for PCM63 and TDA1541.

Use half will be Single ended verrsion.

I have good result with that.


Thx for posting that.

Do you have a standard (unbalanced) schematic?
 
Hi hollowman,

I'd like to build a similar I/V stage for my Philips cd723 (tda1545a).
I've found an interesting article on Borbely's site, look:

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

Regarding this page:

http://www.borbelyaudio.com/eb604417.asp

the input discrete opamp is the same that he describes here (604/410)

http://www.borbelyaudio.com/pics/305borbely-new.pdf

I don't know what kind of fet and values he used in filter buffer, sorry.
I'm not referring to the filter values but to the cascade buffer.
Probably 2sk/2sj series are a really good choice.

I hope someone knows and share it here ;-)

Mark
 
This is my estimation:
Q9 = 2SK246BL
Q10=2SK170BL
Q11=2SJ74BL
Q12=2SJ103BL

It must be the jfet that he is selling in his website!

Also take not on the below stated by Borbeley:
The 3rd order Bessel filter is built around a complementary JFET follower Q10-Q11. These are cascoded with Q9-Q10 to reduce input capacitance and distortion caused by capacitance modulation at high source impedance. Q10-Q11 is selected for Idss=8-10mA. Idss of Q9-Q12 has to be higher than that of Q10/Q11 to avoid forward biasing.
 
spencer said:
This is a balance version Jfet IV modified from NP D1.

Was the orig. Nelson Pass D1 a single-ended class-A preamp (not sure if your schematic is SE class-A...sorry, I'm a novice, here)? How feasable (accurate-sounding) are solid-state single-ended class-A I/V's for DAC's? I can't recall having encountered any in the past, so I'm not sure if the lack-of-popularity is based on over-technicality-of-design, availability of needed components, etc.
 
hollowman said:


Was the orig. Nelson Pass D1 a single-ended class-A preamp (not sure if your schematic is SE class-A...sorry, I'm a novice, here)? How feasable (accurate-sounding) are solid-state single-ended class-A I/V's for DAC's? I can't recall having encountered any in the past, so I'm not sure if the lack-of-popularity is based on over-technicality-of-design, availability of needed components, etc.

You can't throw a stone in this place without hitting a ss/se i/v of some sort from the 2 mosfet original Pass D1 through Jocko's simple i/v and its variants, rbroer's i/v, Elso Kwak's i/v to Pedja's AD844 based i/v , an OP660 based i/v, the Hawksford i/v and the Wadsworth/Wadia/PA630 i/v and on and on.......
 
spencer said:

Thx. It looks like a very do'able project but rfbrw mentions a bunch of other "ss/se i/v's" so I'm tossed as to which one to build first.

rfbrw said:


You can't throw a stone in this place without hitting a ss/se i/v of some sort from the 2 mosfet original Pass D1 through Jocko's simple i/v and its variants, rbroer's i/v, Elso Kwak's i/v to Pedja's AD844 based i/v , an OP660 based i/v, the Hawksford i/v and the Wadsworth/Wadia/PA630 i/v and on and on.......

As I said in the last thread: "sorry, I'm a novice, here".

IAC, I appreciate the keywords you provided above. They will help me use diyaudio.com's search engine to find topical threads.

BTW, is there a "consensus" amongst DIYers as to which is the better design to build -- SE or Diff (both SS) -- if $/effort are equal but sound quality was the ultimate deciding factor?

Thx again!
 
Originally posted byK:\Temporary_G-drive_storage\crsupport_us\images
The Pass D1 stage and its variants is probably the most common and given its simplicity offers the most for the least.
The D1 i/v
The Klaus JFet remix and associated threads 1, 2 and 3
Another remix The acaudio All Mosfet remix and its thread.

A little light D1 reading.

Thx again for sharing this info -- redundant as it may be to seasoned modders.

Which D1 I/V stage is your favorite?

Shifting gears to a few related questions...

(first, a little background…)

I'm not happy with my current DAC -- it's the "Monica 2"
NOS (TDA1545A based Non-Oversampling DAC with Asynchronous Reclocking). Not sure why it sounds subpar… maybe it's because I'm not using a dedicated I/V (the designer of the Monica2 uses his tubed I/V called Grounded Grid Tube Gain Stage -- but I'm not a huge tube fan nor do I want such a big/bulky/non-portable "solution") and/or maybe these NOS DACs are not all they're cracked up to be. What I do know is that my mod'ed Toshiba SD-3990 sounds way better than the Monica2 NOS DAC.

Questions…

Is it the NOS approach itself -- or the way it's implemented in the Monica2 -- that's the limiting factor? (Some DIYers swear by a stacked TDA154x NOS approach, which uses several DAC chips soldered in parallel).

Is there, overall, more support among DIYers for oversampling DACs, like PCM1704?

Thx for info anyone can provide on this topic.
 
So far I tried DAC CS4398, PCM63, PCM64 (my old Denon player), TDA1541, TDA1543 etc.. but PCM63 and TDA1541 give the best result on sound quality. I guess AD1862/5, PCM1702 and PCM1704 should also give good sound. I am quite disappointed on CS4398 delta sigma DAC as the sound is too techno type.

I did not diy the digital part and I only modify the IV and buffer stage in DAC and the result is very good to me on PCM63 and TDA1541. I tried NP D1 IV before but the drive to pre-amp / power amp is lack of punch and sound too flat in my system.

Here is one photo how I put the Jfet IV and Borbeley low noise Jfet regulator in my DAC. I also added a Lundahl LL7902 audio transformer at the output to get Balance output to match my pre/power amps.

Spencer
 

Attachments

  • dsc_2337s.jpg
    dsc_2337s.jpg
    72.2 KB · Views: 3,395
spencer said:
So far I tried DAC CS4398, PCM63, PCM64 (my old Denon player), TDA1541, TDA1543 etc.. but PCM63 and TDA1541 give the best result on sound quality. I guess AD1862/5, PCM1702 and PCM1704 should also give good sound. I am quite disappointed on CS4398 delta sigma DAC as the sound is too techno type.

I did not diy the digital part and I only modify the IV and buffer stage in DAC and the result is very good to me on PCM63 and TDA1541. I tried NP D1 IV before but the drive to pre-amp / power amp is lack of punch and sound too flat in my system.

Here is one photo how I put the Jfet IV and Borbeley low noise Jfet regulator in my DAC. I also added a Lundahl LL7902 audio transformer at the output to get Balance output to match my pre/power amps.

Thx for replying and for sharing the pix. The layout looks quite clean but I'll be sure to study them more thoroughly before attempting the project.

Earlier you noted "[t]his is a balance[d] version Jfet IV modified from NP D1". And above you note "I tried NP D1 IV before but the drive to pre-amp...".

This may sound like a dumb/novice question, but would you consider your JFET I/V a hybrid of sorts? That is, single-ended, like NP's (pre)amp designs, but with some of the virtures Borbely's JFET designs (minus the diff. stage) -- perhaps, a best-of-both-worlds approach?
 
spencer said:
This is a new version of Jfet IV with regulator in one single board in TDA1541 DAC

From the other photos you posted, I noticed you were modding, TTBOMK, a GR Research DAC-72.

I know GR also makes an 8-TDA1543s-in-parallel DAC called "DAC-AH" but I don't think they make one that uses TDA1541s. So, which DAC, if you're using a commercial one, will this new version of Jfet IV with regulator be used with?
 
spencer said:
This is a balance version Jfet IV modified from NP D1.

R9/10 adjust output stage gate/source at 10V.

R11/12 adjust input offset at 0V or desire value.

Adjust R3/4 for different DAC out current. 1.5k is good for PCM63 and TDA1541.

Use half will be Single ended verrsion.

I have good result with that.

Spencer

Hi spencer ,
The DAC I/V stage can be use for AD1955 ?
Need any modify?
Thanks!


-------------------------
linhsichi
 
spencer said:
This is a balance version Jfet IV modified from NP D1.

R9/10 adjust output stage gate/source at 10V.

R11/12 adjust input offset at 0V or desire value.

Adjust R3/4 for different DAC out current. 1.5k is good for PCM63 and TDA1541.

Use half will be Single ended verrsion.

I have good result with that.

Spencer


Hi Spensor,

Just curious what values you are using for C9 and C10.

-David
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.