Discrete Opamp Open Design - Page 219 - diyAudio
Go Back   Home > Forums > Source & Line > Analog Line Level

Analog Line Level Preamplifiers , Passive Pre-amps, Crossovers, etc.

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 24th November 2012, 07:16 PM   #2181
diyAudio Member
 
Wavebourn's Avatar
 
Join Date: Sep 2006
Location: Pleasant Hill, CA
Send a message via Skype™ to Wavebourn
Quote:
Originally Posted by kgrlee View Post
linux, what current are you running in the output stages?
I suppose, electrical.
__________________
The Devil is not so terrible as his math model is!
  Reply With Quote
Old 24th November 2012, 07:40 PM   #2182
diyAudio Member
 
grhughes's Avatar
 
Join Date: Mar 2005
Location: Cubicle Sweet Cubicle, Springfield, Missouri, USA
I feel it should use the latest discreet semi conductors and latest topology tricks. I'm sure the new ones have fewer impurities even if they are damn small in SOT. I believe topologies sound different. Has anyone suggested an all N output stage as the OPA2604? Ray
__________________
" "The fool doth think he is wise, but the wise man knows himself to be nothing but a fool." Will Shakespeare
  Reply With Quote
Old 24th November 2012, 09:16 PM   #2183
diyAudio Member
 
Join Date: Nov 2011
Location: Cooktown, Oz
Default Voltage Followers into 15R @ 3.162Vp

Here's what old topologies can do with some dirty tricks. THD plots for Voltage Followers @ a headbanging 3.162Vp 20kHz into 15R.

#2060 is running 7mA through the output devices. SW-OPA 10mA. I've used Cordell's 4401/3 models cos they are worst case for both distortion & stability in these sims.

You can see the characteristic rising spectrum of crossover distortion with SW-OPA but remember these products are 90 - 100dB down.

More Iq and medium power devices will reduce crossover but #2060 will always show less crossover at the same current.

The best ADs/DAs today are juu..ust 20b so 117dB[*] is about the measurement limit today for distortion residuals. See Scott for his experience of a Panasonic AD with true 24b performance that was declared un-manufacturable.
____________________

Anyone know how to make LTspice produce a distortion residual waveform?

I like to look at this with old fashion THD instruments. You can usually see if THD is due to crossover, 'gm doubling', poor PSR etc.

[*] You can go a bit better with sophisticated 'averaging' and DSP but only a bit. The 'noise floor' on FFT measurements is NOT the same as what distortion products you can measure. That depends on the AD.
Attached Images
File Type: png #206015Rx1_3V.png (54.1 KB, 221 views)
File Type: png SW-OPA15Rx1_3V.png (62.0 KB, 211 views)

Last edited by kgrlee; 24th November 2012 at 09:37 PM.
  Reply With Quote
Old 24th November 2012, 09:35 PM   #2184
bcarso is offline bcarso  United States
diyAudio Member
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by kgrlee View Post
____________________

Anyone know how to make LTspice can produce a distortion residual waveform?

I liked to look at this with old fashion THD instruments. You can usually see if THD is due to crossover, 'gm doubling', poor PSR etc.
Many times I use banks of notch filters in sim, and besides showing the residuals in the time domain I also get a cross-check on the veracity of the DFT. Of course the notch filters must be given plenty of time to settle.

In a recent study of state-variable-filter-based oscillators, I found the DFT refusing to give anything close to the right answer, very much overestimating the distortion. I knew something was screwy when running two matched oscillators in opposing polarity had no effect on second-order when the outputs were differenced!
  Reply With Quote
Old 24th November 2012, 09:46 PM   #2185
diyAudio Member
 
Join Date: Nov 2011
Location: Cooktown, Oz
Quote:
Originally Posted by bcarso View Post
In a recent study of state-variable-filter-based oscillators, I found the DFT refusing to give anything close to the right answer, very much overestimating the distortion. I knew something was screwy when running two matched oscillators in opposing polarity had no effect on second-order when the outputs were differenced!
This wasn't by any chance Great Guru Baxandall's swept oscillator using semiconductor junctions?

John Vanderkooy recently did an analysis and update in homage for AES.

I had the dubious honour of getting this to work for production in Jurassic times. Though I worship GGB on my knees 5x a day, his designs often required extensive tweaking by expensive virgins in production.
  Reply With Quote
Old 24th November 2012, 10:14 PM   #2186
bcarso is offline bcarso  United States
diyAudio Member
 
Join Date: May 2005
Location: Canoga Park, California
Quote:
Originally Posted by kgrlee View Post
This wasn't by any chance Great Guru Baxandall's swept oscillator using semiconductor junctions?

John Vanderkooy recently did an analysis and update in homage for AES.

I had the dubious honour of getting this to work for production in Jurassic times. Though I worship GGB on my knees 5x a day, his designs often required extensive tweaking by expensive virgins in production.
No semi junctions involved for tuning, but I had remarked to Chris Paul that I'd played around with low distortion oscillators many years ago, having conceived of a stabilization scheme involving the trig identity, taking advantage of the presence of quadrature outputs for the SVF. I later noticed this approach published in the classic opamp book of Tobey et al.

And the gotcha for the trig identity approach is making really good four-quadrant multipliers to square the quadrature outputs.

The main trick with oscillators with a lot of internal loss (as opposed to very-high-Q resonators like crystals) is to make the required amplitude stabilization intervention as small as possible, and avoid the feedthrough of garbage from the detector of amplitude. Hofer and company did this in a clever way as to make frequency changes more rapid without degrading distortion too much (there is a patent) by feeding some of the control signal ahead to cancel.

The late Jim Williams got to explore these issues late in life (as it turned out, sadly) when all he wanted was a very low distortion oscillator, and characteristically figured he could do it on his own. IIRC it was a somewhat humbling experience.
  Reply With Quote
Old 24th November 2012, 10:15 PM   #2187
PHEONIX is offline PHEONIX  Australia
diyAudio Member
 
Join Date: Jul 2004
Location: Australia
jcx has a THD residual solution for ltspice

Spice simulation
  Reply With Quote
Old 25th November 2012, 01:31 AM   #2188
fas42 is offline fas42  Australia
diyAudio Member
 
fas42's Avatar
 
Join Date: Jun 2012
Location: NSW, Australia
Blog Entries: 11
Quote:
Originally Posted by kgrlee View Post
Anyone know how to make LTspice produce a distortion residual waveform?
Any output waveform can represent an expression of any level of complexity, so simply subtract the input from the output, with multiplier and phase adjusted appropriately, is a simple way to do the job ....

Frank

Last edited by fas42; 25th November 2012 at 01:38 AM.
  Reply With Quote
Old 25th November 2012, 01:33 AM   #2189
diyAudio Member
 
Join Date: Sep 2007
Send a message via Yahoo to linuxguru
Quote:
Originally Posted by kgrlee View Post
linux, what current are you running in the output stages?
LTSpice says:

For J113/J176 and 47R 47R: 8 mA
For J111/J175 and 100R 100R: 16.6 mA
For J111/J175 and 150R 150R: 13 mA
For 2sk170bl/2sj74bl, no source degeneration and 1x 1n4148 between gates: 2.9 mA

In all cases, the VAS current sims as 3.25 mA.

I'd go with selected J111/J175 with Idss around 20 mA. TO18 2n4391 runs about $1.50 and 2n5114 at least $5 on EBay.
  Reply With Quote
Old 25th November 2012, 02:07 AM   #2190
diyAudio Member
 
Join Date: Sep 2007
Send a message via Yahoo to linuxguru
Quote:
Originally Posted by bcarso View Post
It appears both parts are being made by Linear Integrated Systems, who however only gives an ambient rating of 500mW for the P part and evidently does not tie the gate to the case like the 4391. Older 5114s definitely had the gate connected to the case.
InterFET also has current-production TO18 2n5114s with the gate connected to the case, but rated at 500 mW Pd.

Is there a cheaper TO92 PN5114 or similar?
  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
Discrete OPAMP audio-gd Vendor's Bazaar 27 20th September 2012 05:02 PM
discrete opamp help blackpowderaudio Parts 0 16th December 2009 04:46 PM
THAT transistor headphone amp (250ma discrete opamp) design sanity check. Russ White Headphone Systems 19 13th December 2007 01:52 PM


New To Site? Need Help?

All times are GMT. The time now is 01:25 PM.


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