Go Back   Home > Forums > >

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

Balanced Volume Controller / Line Stage
Balanced Volume Controller / Line Stage
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
Old 19th April 2019, 05:22 PM   #71
alexcp is offline alexcp  United States
diyAudio Member
 
alexcp's Avatar
 
Join Date: Jun 2009
Balanced Volume Controller / Line Stage
It is an interesting discussion that may lead away from volume controls.

My view is that we are talking about entertainment, and the ultimate criterion is "better" or "worse" entertainment. It is subjective, of course (Nelson Pass commented somewhere that the brightness of front panel LEDs on his FirstWatt amplifiers affected the perceived amplifier performance). I asked diyralf about expectations in an attempt to figure out what part of entertainment was missing .

Moving from subjective performance to engineering, the major factor seems to be distortion - although not only the THD+N @ 1kHz. For example, Vladimir Lamm of LAMM Industries, a successful maker of expensive hifi gear, suggests, based on his primary research, ideal distortion measurements (see pics below), which focus on distortion as a function of frequency and volume, as well as on distortion spectrum. The special PCB layout in my balanced volume controls is specifically designed to control HF distortion. (BTW, measuring THD+N @ 20kHz is useful, but if a soundcard is all you have, a two-tone test - say, 19kHz+20kHz - is a good proxy.)
Attached Images
File Type: gif lamm_thd_freq_ideal.gif (1.4 KB, 483 views)
File Type: gif lamm_thd_power_ideal.gif (1.6 KB, 474 views)
File Type: jpg lamm_harmonic_residue_ideal.jpg (12.2 KB, 470 views)
  Reply With Quote
Old 19th April 2019, 05:36 PM   #72
diyralf is offline diyralf
diyAudio Member
 
diyralf's Avatar
 
Join Date: Aug 2016
There are many criteria for a good level control. E.g. offness, THD at higher frequencies, what happens at 10Vrms input, L / R deviation at large and very small levels, etc.? Electronic solutions fail here on some points, other solutions on others. You have to know what's important to you.
  Reply With Quote
Old 19th April 2019, 08:16 PM   #73
ah693973 is offline ah693973  United States
diyAudio Member
 
Join Date: May 2007
Hi, Getting ready to start on this project. Do you have power requirements so I can put together a power supply? I know I need +- 12 to 15 for the 2310 but what is the amp draw? I will likely bypass the regulator on the board. Thanks.
  Reply With Quote
Old 19th April 2019, 08:41 PM   #74
alexcp is offline alexcp  United States
diyAudio Member
 
alexcp's Avatar
 
Join Date: Jun 2009
Balanced Volume Controller / Line Stage
With LM4562s and regulators onboard, each board draws about 70mA from each rail; negative rail takes slightly less.

One LM4562 draws about 10mA, see figures 102-104 on page 21 in its datasheet (NE5532s may take slightly less depending on the make); PGA2310 takes about 7-8mA per rail; and each regulator needs about 5-6mA of quiescent current.

With the onboard 12..15V regulators bypassed, I would expect it to be 5-6mA less on each rail.

This consumption doesn't include the controller (e.g. an Arduino) that can be powered with 5V from the 8-pin connector.
  Reply With Quote
Old 20th April 2019, 08:18 AM   #75
jaddie is offline jaddie
diyAudio Member
 
Join Date: Mar 2011
Quote:
Originally Posted by alexcp View Post
It is an interesting discussion that may lead away from volume controls.

My view is that we are talking about entertainment, and the ultimate criterion is "better" or "worse" entertainment. It is subjective, of course (Nelson Pass commented somewhere that the brightness of front panel LEDs on his FirstWatt amplifiers affected the perceived amplifier performance). I asked diyralf about expectations in an attempt to figure out what part of entertainment was missing .
Subjective data can be analyzed objectively if collected properly. For example, you can analyze the perceived audible impact of the brightness of the front panel led by making it the single variable in an ABX test with sufficient tests and testers.
Quote:
Originally Posted by alexcp View Post
Moving from subjective performance to engineering, the major factor seems to be distortion - although not only the THD+N @ 1kHz.
If you're speaking of the audibility of distortion, then it's still subjective. Engineering and subjectivity do not exist in isolation of each other.
Quote:
Originally Posted by alexcp View Post
For example, Vladimir Lamm of LAMM Industries, a successful maker of expensive hifi gear, suggests, based on his primary research, ideal distortion measurements (see pics below), which focus on distortion as a function of frequency and volume, as well as on distortion spectrum.
But that's not new, nor peculiar to him. Though not typically published, any audio developer will have done distortion vs frequency and level tests. It's quite easy to do with modern test methods. But those test results are not published because they comprise a rather large amount of data that really can only be charted on a 3D graph, which is hard for the typical consumer to assimilate, hence the published single-figure THD spec, easier to understand, but also meaningless.

The tests and measurements are objective, the audibility of the results are subjective. Particularly so if one considers that the audibility of harmonic distortion products depends on the specific spectrum of the distortion products AND how they may be masked by the musical signal. For example, there is ample evidence that suggests relatively strong amounts of even-order distortion products are necessary before they reach audibility, yet lower amounts of odd-order harmonics are objectionable, though even that is a gross simplification.
Quote:
Originally Posted by alexcp View Post
The special PCB layout in my balanced volume controls is specifically designed to control HF distortion.
Please elaborate on how a PCB layout affects HF distortion.
Quote:
Originally Posted by alexcp View Post
(BTW, measuring THD+N @ 20kHz is useful, but if a soundcard is all you have, a two-tone test - say, 19kHz+20kHz - is a good proxy.)
A good many soundcards can be set to sample at 192kHz these days, making 20kHz THD measurable to the 4th harmonic possible. Useful? Not too sure about that. Multi-tone signals at high frequencies may have more relevance to the audibility of nonlinearities because the distortion products fall easily mid-band, and in relative frequency isolation of the stimulus (the difference product of 19kHz and 20kHz is 1kHz, easy to measure and hear, whereas the third harmonic of 20kHz is 60kHz, which is impossible to hear). I don't think conventional THD measurements above 10kHz have a lot of meaning in terms of audibility of distortion because the distortion products fall outside the range of human hearing. However, whatever nonlinearity causes higher THD will also likely cause IMD, and IMD products can become quite audible even when the signals being distorted are just outside of the audible frequency range. There is validity to looking at multi-tone tests done above 20kHz, and especially between 10kHz and 20kHz. The paper I like to refer to on this is "Spectral Contamination Measurement" by Jensen and Sokolich, JAES, November 1988.
  Reply With Quote
Old 20th April 2019, 01:44 PM   #76
alexcp is offline alexcp  United States
diyAudio Member
 
alexcp's Avatar
 
Join Date: Jun 2009
Balanced Volume Controller / Line Stage
Thank you jaddie for elaborating on there topic. Didn't I say it is an interesting discussion that may lead away from volume controls?

Regardless of how you measure it, distortion rising with frequency has an audible sonic signature, recognizable as tight or controlled bass. In a feedback amplifier, two contributors to distortion rising with frequency are (1) lack of HF loop gain and (2) errors in the feedback loop.

Lack of HF loop gain is particularly conspicuous when gain-bandwidth product (GBW) is low and gain is high, as in a power amplifier. With GBW fixed, second- and higher-order feedback loops and composite amplifiers help improve HF loop gain. You can also throw away some LF gain ("reducing feedback") to make distortion level vs frequency more uniform and thus less offensive. The volume controls that are the subject here, however, use LM4562 with GBW=55MHz at unity gain, still providing about 70dB of loop gain at 20kHz, which makes lack of loop gain much less of an issue.

Errors in the feedback loop are construction dependent, and here is where the PCB layout and routing become crucial. Noise and non-linear currents from e.g. power supplies couple into signal and feedback loops, and coupling increases with frequency. To deal with this, I explicitly treat each signal as a pair of wires routed together and place all components in pairs, thus minimizing the area between the signal and its reference. This way, any interference affects both signal and the reference equally and becomes a common mode signal, which LM4562 with CMRR=120dB rejects.

BTW, I previously applied the same construction technique to LM3886 based power amplifier with excellent results.
Attached Images
File Type: png avc+s_8816 bottom copper.png (18.7 KB, 449 views)
File Type: png LM3886_TH_R1_bottom.png (63.6 KB, 446 views)
  Reply With Quote
Old 20th April 2019, 05:11 PM   #77
jaddie is offline jaddie
diyAudio Member
 
Join Date: Mar 2011
Quote:
Originally Posted by alexcp View Post
Thank you jaddie for elaborating on there topic. Didn't I say it is an interesting discussion that may lead away from volume controls?

Regardless of how you measure it, distortion rising with frequency has an audible sonic signature, recognizable as tight or controlled bass. In a feedback amplifier, two contributors to distortion rising with frequency are (1) lack of HF loop gain and (2) errors in the feedback loop.

Lack of HF loop gain is particularly conspicuous when gain-bandwidth product (GBW) is low and gain is high, as in a power amplifier. With GBW fixed, second- and higher-order feedback loops and composite amplifiers help improve HF loop gain. You can also throw away some LF gain ("reducing feedback") to make distortion level vs frequency more uniform and thus less offensive. The volume controls that are the subject here, however, use LM4562 with GBW=55MHz at unity gain, still providing about 70dB of loop gain at 20kHz, which makes lack of loop gain much less of an issue.

Errors in the feedback loop are construction dependent, and here is where the PCB layout and routing become crucial. Noise and non-linear currents from e.g. power supplies couple into signal and feedback loops, and coupling increases with frequency. To deal with this, I explicitly treat each signal as a pair of wires routed together and place all components in pairs, thus minimizing the area between the signal and its reference. This way, any interference affects both signal and the reference equally and becomes a common mode signal, which LM4562 with CMRR=120dB rejects.

BTW, I previously applied the same construction technique to LM3886 based power amplifier with excellent results.
While I do not question the existence of these mechanisms, I do strongly question their audibility. Please understand, I'm not saying your claims here are not true, but they do raise some questions.

How can rising distortion at high frequencies (I assume, by HF we are talking 10-20kHz and above?) be heard as affecting bass?

How have you evaluated the audibility of HF loop gain? I'm asking specifically about the evaluation/test method. And yes, the test method does matter.
  Reply With Quote
Old 20th April 2019, 06:34 PM   #78
alexcp is offline alexcp  United States
diyAudio Member
 
alexcp's Avatar
 
Join Date: Jun 2009
Balanced Volume Controller / Line Stage
Quote:
Originally Posted by jaddie View Post
I do strongly question their audibility.
The best way I know to address that is to build yourself several different boards, listen to them and tell us what you think you hear and what you think you don't hear. Never trust strangers' ears
  Reply With Quote
Old 20th April 2019, 08:19 PM   #79
jaddie is offline jaddie
diyAudio Member
 
Join Date: Mar 2011
Quote:
Originally Posted by alexcp View Post
The best way I know to address that is to build yourself several different boards, listen to them and tell us what you think you hear and what you think you don't hear. Never trust strangers' ears
Um...no, that would be the worst way. Costly in terms of parts and time, and totally subjective and under the influence of full expectation bias.

This is precisely what I'm getting at. That kind of testing yields erratic and non-duplicatable results. If that's the kind of testing used to verify the theories above, then we have no proof.

You can trust strangers ears, provided the tests are carefully controlled and you have enough strangers. In fact, you need strangers to validate the test!
  Reply With Quote
Old 21st April 2019, 12:23 AM   #80
alexcp is offline alexcp  United States
diyAudio Member
 
alexcp's Avatar
 
Join Date: Jun 2009
Balanced Volume Controller / Line Stage
What happens when a panel of strangers in a carefully controlled test (or John Atkinson of Stereophile - I am a subscriber!), prefers something that, subjectively, sounds like crap to you?
  Reply With Quote

Reply


Balanced Volume Controller / Line StageHide this!Advertise here!
Thread Tools

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
Balanced Audio Volume Controller with PGA2310 yangsmm Chip Amps 5 16th March 2009 08:37 AM
Balanced Audio Volume Controller with PGA2310 yangsmm Chip Amps 7 24th November 2008 12:54 AM
Balanced Zen v5 + balanced Zen line stage Adrie Pass Labs 0 30th May 2004 05:27 PM
Zen Balanced Line Stage Balanced vs Unbalanced macka Pass Labs 28 11th December 2002 06:18 AM


New To Site? Need Help?

All times are GMT. The time now is 10:56 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio
Wiki