Nelson Pass inspired headphone amp - Page 3 - diyAudio
Go Back   Home > Forums > Amplifiers > Headphone Systems

Headphone Systems Everything to do with Headphones

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 8th May 2011, 05:05 AM   #21
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Suggestion: for some good noise reduction cascode the LM317 current source with a DN2540N5 depletion mode mosfet. I'm always looking for ways to wedge in a d-mosfet!

See figure 13C, page 4, in Walt Jung's audio current source article here

http://waltjung.org/PDFs/Sources_101_P2.pdf

then compare the graph 13D for the cascode with just the LM317 current source alone in figure 10B. With the cascode your rejection stays below 110dB, on out to 200kHz, vs. just 60dB. That 110dB rejection is also lower than just the depletion mosfet current source alone, as he shows in figure 12B.
Attached Images
File Type: png dn2540N5 cascode LM317 isource.png (2.0 KB, 926 views)

Last edited by agdr; 8th May 2011 at 05:11 AM.
  Reply With Quote
Old 8th May 2011, 02:34 PM   #22
Fenris is offline Fenris  United States
diyAudio Member
 
Join Date: Jun 2004
Location: The last frontier
Funny you should mention that, I've got 30 PCBs for that CCS waiting to be populated on my workbench. The problem using that though is threefold.

First, the voltage drop across it is greater, limiting the voltage swing. With a 19v power supply, assume a 3v drop across the 317, 5 volts across the MOSFET, and leave a volt or so at each end for headroom. Now you've only got 9v of swing. It should be enough, but by DC coupling you're setting an operating point that may or may not be ideal. I set a pretty decent one, but didn't plan on the extra voltage drop from a depletion mode MOSFET. That would cut my voltage swing down from 9 to 5.5 volts making the set point unusable (11.2v - 4.5v - 6.5v = 0.2v swing). This means I would have to alter the setpoint by lowering the gain to about 1, which would eliminate the whole point of a gain stage. Alternately, I could AC couple the output stage but this would add complexity, cost, and another capacitor in the chain.

The second problem is increased cost, complexity, and space taken up. I was aiming for simple and low cost for the beginner, vice an ultimate cost no-object amp. Right now there's only 10 resistors, two pots, six three pin devices, and a half dozen caps. I literally built it from my (frighteningly expansive) junk box.

Third, 110 db is a really good number already. That's as good as the resolution of most good DACs. I don't think I could tell the difference between 110 and 130db.

There will be more noise coming from the front end than the source follower anyway. Let's do a quick analysis. I did design in some noise rejection, but it's not perfect. Let's use 1v of noise just for an easy to work with number. Starting from the left, the voltage divider for the virtual ground will give us 3.25/19 = 0.171v. Voltage gain is -1300/560 = -2.32, so there's approximately -.4v on the JFET drain. But there will also be current flowing through the JFET from the power supply, and at the drain it will be 11.2/19 = 0.59v. So the net will be +0.19/1 or 19% of the PSU noise will be injected. Not a great PSRR, but better than many circuits.

I'm less concerned about the PSRR - it's easy enough to regulate the power supply (although I thought I was already using a regulated power supply) or add filtering. My big problem is with static and noise. There's a constant hiss. Whenever I touch the gate of the JFET, or the RCA input jacks, or the source (I'm using a Zune) the hiss gets much louder.
__________________
Listen to the music through the stereo, not the stereo through the music.
  Reply With Quote
Old 10th May 2011, 09:48 PM   #23
Fenris is offline Fenris  United States
diyAudio Member
 
Join Date: Jun 2004
Location: The last frontier
Here's an even simpler 9v version. I really like the simplicity/synchronicity of this one. The LED provides power indication and JFET bias, and the output current can be selected via the resistor on the LM317. Quite elegant and almost usable as a portable amp (maybe with some Li-Po batteries in series?). Too many designs to try out!
Attached Images
File Type: gif headphone2.gif (15.3 KB, 897 views)
__________________
Listen to the music through the stereo, not the stereo through the music.
  Reply With Quote
Old 10th May 2011, 10:47 PM   #24
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Whoops - I forgot to circle back to the thread after I posted above.

All good points. I was thinking that the extra parts would probably go against the low parts count design goal. As for the voltage drops, they probably can be a lot less. At the low currents involved the LM317 data sheet shows a dropout around 1.3V, if I'm reading that right. From the DN2540 data sheet it looks like the part would also work at lower voltages, even though it is good for 400V. But good point about the (lack of) need for more rejection here. May be a moot issue.

I don't have any ideas on the hiss and static issues, but I do like your reduced-calorie 9V design!
  Reply With Quote
Old 11th May 2011, 06:02 PM   #25
Fenris is offline Fenris  United States
diyAudio Member
 
Join Date: Jun 2004
Location: The last frontier
I'm still searching for a device with a low Vgs and high Idss for the source follower role. Most MOSFETS have at least +2v and it goes up to about +4v at 100ma. JFETS have about a -0.1v Vgs that goes more negative with increasing Idss current, up to several volts. With only 9v to work with, a Vgs of -0.1 to -0.5v is ideal. Any JFET will work as a source follower, but five to ten 2sk170s in parallel are pretty ideal.
__________________
Listen to the music through the stereo, not the stereo through the music.
  Reply With Quote
Old 13th October 2011, 03:44 PM   #26
udailey is offline udailey  United States
diyAudio Member
 
Join Date: Mar 2008
Fenris,
In regards to your first schematic:
Assuming I have 5 pairs of Jfets. Each pair is made of 2 matched Jfets. Do the 5 pairs need to match each other? So A and B make a pair and C and D make a pair.. do A/B need to match C/D?
Uriah
__________________
purchase LDRs anytime Also try my Resistor Replacers or LDR based Input Selector Email me. diyldr@gmail.com
  Reply With Quote
Old 14th October 2011, 01:19 AM   #27
EUVL is offline EUVL  Europe
diyAudio Member
 
Join Date: Oct 2003
> Assuming I have 5 pairs of Jfets. Each pair is made of 2 matched Jfets. Do the 5 pairs need to match each other? So A and B make a pair and C and D make a pair.. do A/B need to match C/D?

No. But A & C should be in parallel (say the top device), and B // D (the bottom device).
Also A & B should be thermally coupled (by thermal glueing face to face, or use a DUal JFET heatsink) for best thermal tracking (DC stability).
Same applies to C & D.

Hope you don't mind me chipping in.
I have a similar set up in the output stage of my preamp which was published at the blowtorch thread about 5 years ago.

John Curl's Blowtorch preamplifier


Patrick

.
Attached Images
File Type: jpg 051121 preamp board.jpg (92.1 KB, 586 views)

Last edited by EUVL; 14th October 2011 at 01:39 AM.
  Reply With Quote
Old 14th October 2011, 01:26 AM   #28
udailey is offline udailey  United States
diyAudio Member
 
Join Date: Mar 2008
I dont mind at all. I appreciate it Patrick. Are you still selling those sinks? Think I saw a quad sink to.
__________________
purchase LDRs anytime Also try my Resistor Replacers or LDR based Input Selector Email me. diyldr@gmail.com
  Reply With Quote
Old 14th October 2011, 01:34 AM   #29
EUVL is offline EUVL  Europe
diyAudio Member
 
Join Date: Oct 2003
I was hoping Fenris wouldn't mind.
Yes, you can get those sink from us, but that wasn't why I posted.


Patrick

.

Last edited by EUVL; 14th October 2011 at 01:40 AM.
  Reply With Quote
Old 13th December 2011, 01:49 AM   #30
pidesd is offline pidesd  Canada
diyAudio Member
 
pidesd's Avatar
 
Join Date: Apr 2008
Location: montreal (cotes des neiges yo!)
hi all,

i ve took the chance and tried to build a source follower headphone amp using 10 2sk170 in parallel in a source follower configuration. just like the dcb1 but with no output resistor and 10 times the j-fets. now it sounds good but i hear a low level background noise all the time music is playing. i ve tried with and without a pot, different VDS voltages but nothing does. I was wondering if the circuit is flawed to start with. any idea what it might be? i might take it to the school lab for further analysis.

by the way i m driving grados and the fets are only idss matched and each fet has a 121 gate resistor
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Advice for Linkwitz Orion system with Nelson Pass-style pre-amp and amp GeorgeBoles Pass Labs 123 27th December 2012 04:54 AM
Power Amp Construction inspired by Curl, Slone and Pass circuitscc Solid State 11 21st May 2010 05:25 AM
Nelson Pass discrete op-amp. beppe61 Solid State 12 18th February 2005 02:06 PM
Checkout the Nelson Pass inspired GC hifimaker Chip Amps 0 25th January 2005 11:12 PM


New To Site? Need Help?

All times are GMT. The time now is 11:17 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2