Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Solid State Talk all about solid state amplification.

Ultra low noise amp with JFET's - How to reduce the THD ??
Ultra low noise amp with JFET's - How to reduce the THD ??
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
Thread Tools Search this Thread
Old 20th May 2010, 05:05 PM   #1
toogool is offline toogool  Germany
diyAudio Member
Join Date: Apr 2010
Default Ultra low noise amp with JFET's - How to reduce the THD ??

Hey guys!

I'm trying on building some nice simple ultra low noise amp with noise values below the 1 nV/rtHz level. Read many papers and threads about that topic and built up several little amps with low noise JFET's like 2SK170, 2SK369 and IF3601. I use them in source mode going into a non-inverting Opamp LT1028 as second stage. My noise values and bandwith are good.. reached like 0,3 nV/rtHz and bandwith up to 1 MHz. So that's good, but what's not so good is the THD ! Often I got like 3-6 harmonics resulting in THD's of 0,5 to 3% so that's way too high for my application. I need some value below 0,1%.
Biassettings of the 2SK170 for example are like Uds = 8V , Ugs = -0,12V and Id = 6 mA. So those values should work out for good linearity in the input and output curves. But they don't :-( Mostly I get strong second harmonic.
I put up my basic schematic and a picture of those bad harmonics I get, hoping that anyone knows how to effectively reduce that THD!! Or is it not possible in a source-circuit to drop the THD below 0,1% ? Do I need some complementary stage or feedback to get better?

Supply voltage is +-12V not 5V and signalsource is like 10kHz with 10mVpeak.

Thanks for your answers,

Attached Images
File Type: jpg forumsbilder.jpg (33.8 KB, 754 views)
File Type: jpg Id_Abgleich.jpg (27.1 KB, 717 views)
  Reply With Quote
Old 20th May 2010, 06:01 PM   #2
Gordy is offline Gordy  United Kingdom
Account disabled at member's request
Join Date: Nov 2006
Hi, with small input the 2SK170 can do 0.003 - 0.005%.
See Table 2 and 3 here: Audial DIY :: Module LAB :: Preamp

Also see interesting discussion here: Simple 60dB discrete low noise amplifier (lna)

Last edited by Gordy; 20th May 2010 at 06:04 PM.
  Reply With Quote
Old 20th May 2010, 06:10 PM   #3
jcx is offline jcx  United States
diyAudio Member
Join Date: Feb 2003
Location: ..
what are source impedance, signal level for the distortion # ? - nonlinear Cin is sometimes an important effect at higher Zin

but the no input stage feedback common source operation is likely the issue

and why 90 Hz corner frequency
  Reply With Quote
Old 20th May 2010, 06:17 PM   #4
knutn is offline knutn  Norway
diyAudio Member
Join Date: Mar 2006
Location: Oslo
Default The VGS matters

You must remember that the input level must be low, much lower than 0.12 V, to get some descent THD figures. Otherwise you will have to increase the source resistor (increasing the noise) to increase VGS and lower the drain current. I think the simplest solution would be to include the FETs in the OPAs feedback loop.
  Reply With Quote
Old 20th May 2010, 07:37 PM   #5
toogool is offline toogool  Germany
diyAudio Member
Join Date: Apr 2010
Hey guys, thanks for your answers!!

source impedance will be some coil for detection of magnetic field... somewhat with parameters like R = some Ohms (DC), L = some ÁH and selfwound. perhaps some coupling transformer will be added to that in the input stage. so thats why I chose JFET's with high input impedance.

but at this time I don't really know what signallevel will come to amplify. I think some mV. So let's assume 10 mVpeak. That should be small enough for Ugs = 0,12 V ?!
Would be still in the linear part of the transfer curves of 2SK170 with Idss like 10 mA (BL). I matched 4x 2SK170 with Idss = 10 mA.

The ULNA is not for audio (should I say that on this forum?!) but it's quite the same because my bandwidth starts at 1 kHz up to max. 1 MHz. (ok thats not so audio-like but it's very similar to those phono amp designs..) and it should be usable for audio, too. so that's why corner = 90 Hz... doesn't really matter - main signal frequency starts at some 10 kHz or so...

You got some example how to include the JFET's into the feedback loop of the opamp? I just found some quite complicated ones.. so the main thing is changing the opamp into inverting-type and give back the output over some coupling R or C to the Source of the JFET's?? How does that effect the bandwidth of the amp? Often there are additional Transistors set to Source and Drain contact of the Main JFET in order to control or balance the Drain-current?! Do those Transistors help with the THD? or what is the need for those?

I add some example schematic of those complicated feedback loop amps, but it's with discrete second stage, too.

So I'll try out some Amp with feedback, I'll simulate it first in Spice and see what I get there But more informations and comments are welcome !!
Attached Images
File Type: gif lnpreamp_sp.GIF (17.7 KB, 685 views)
  Reply With Quote
Old 20th May 2010, 07:45 PM   #6
Nico Ras is offline Nico Ras  South Africa
diyAudio Member
Nico Ras's Avatar
Join Date: Nov 2005
Location: East Coast of South Africa
Try doubling, even tripling the supply voltage, that raises both second and third order intercept.
  Reply With Quote
Old 20th May 2010, 08:09 PM   #7
iko is offline iko  Canada
diyAudio Moderator
iko's Avatar
Join Date: May 2008
Location: Toronto
Ultra low noise amp with JFET's - How to reduce the THD ??
Perhaps you should have a look at this other thread, we've been talking about something similar. Starting with post #126

Simple 60dB discrete low noise amplifier (lna)

we were discussing a transformer input amp that capable of very low noise.
  Reply With Quote
Old 20th May 2010, 10:50 PM   #8
Joachim Gerhard is offline Joachim Gerhard  Germany
diyAudio Member
Joachim Gerhard's Avatar
Join Date: Mar 2009
Try cascoding with a BJT.
  Reply With Quote
Old 21st May 2010, 07:13 AM   #9
analog_guy is offline analog_guy  United States
diyAudio Member
Join Date: Apr 2007
Default Reducing Distortion with JFET Front End

Another effective approach is to utilize a symmetric front end with both N and P channel devices. You will want to match at least Idss for devices of a given polarity. The distortion caused by differences between the N and P devices can be reduced by connecting the N and P devices' outputs together with a capacitor that yields an RC frequency in the 1-10 Hz range. This allows the amp to maintain the proper DC operating points but effectively parallels the upper and lower halves of the diffamp at AC frequencies. I have simulated such a design with up to 10 mVPP, and have consistently obtained distortion levels well below the noise floor of approx 0.3 nV/sqrt(Hz). Using 74/170 FETs gives an AC gain of approx 22 dB for the JFET front end.
  Reply With Quote
Old 21st May 2010, 09:54 AM   #10
Bonsai is offline Bonsai  Europe
Account disabled at member's request
Join Date: Jul 2003
Location: Europe
If you are feeding this amplifer with a source of a few uH and a few ohms, try a bipolar front end. Its quite easy (in simulation anyway) to get distortion figures with low level inputs (say up to 10 or 20mV) of under 1ppm. For some ideas, look up Syn08 on the forum. He did an ultra low noise, low THD front end amp for an RIAA amplifer. The front end amp has a flat response, so this would suit your requirements.
  Reply With Quote


Ultra low noise amp with JFET's - How to reduce the THD ??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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Ultra low noise measurement amp tschrama Solid State 20 22nd September 2008 04:28 PM
Published/DIY ultra-low THD designs? leadbelly Tubes / Valves 10 2nd April 2007 03:41 AM
Ultra low noise filters tiroth Solid State 14 21st May 2002 10:19 PM

New To Site? Need Help?

All times are GMT. The time now is 09:02 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio