Sound Quality Vs. Measurements - Page 417 - diyAudio
Go Back   Home > Forums > Member Areas > The Lounge

The Lounge A place to talk about almost anything but politics and religion.

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 31st March 2012, 03:58 PM   #4161
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Hi,

Quote:
Originally Posted by KSTR View Post
Try referencing everthing to the neg rail instead of GND,
As I always say, GND isn't.

My gainclone does use a different reference, however it uses extra PSU filtering (CRC Pi-Filters), which also helps, plus my special local decoupling...

Quote:
Originally Posted by KSTR View Post
if you look at the PSRR and CMRR plots you'll instantly see why...
Yes, I understand this.

In my case I explicitly did not want to make a "super gainclone", but something for a "reality check".

Ciao T
  Reply With Quote
Old 31st March 2012, 04:59 PM   #4162
dvv is offline dvv  Serbia
diyAudio Member
 
Join Date: Sep 2004
Location: Belgrade, Serbia
Quote:
Originally Posted by ThorstenL View Post
Hi,



My Gainclone is 40W + 40W and it is a straight DC design. Incidentally, the PSU is a big torroidal transformer and two pairs of 3,300uF/50V Elna Silmic II caps in CRC.

It is a "straightline integrated", british style, input selection (relay), 50K Alps Blue Volume Control, direct to the output Amp, LM3875 with more Elna Silmic II, and baked/burned in/selected/mateched AB Carbon composition resistors.

I keep this one around as "reality-check" any amp that is not at least the equal of this humble little Amp in all areas is basically junk. The one thing it does not is unlimited power, drive super low impedances or stunt-bass.

But everything else it does on a very high level. The number of amplifiers that did a lot worse than this little "reality check" is surprisingly large.

Ciao T
Fine, but when? In 2005? That's merely 25 years after they stopped making that model of Maranz power amp.

Also, as you yourself pointed out, a powerhouse it isn't - whereas the Maranz could be said to be approaching one.

For a start, they were the usual damn liars of the time. Under the specified condition, it actually does better in literally all departments after refreshing. It drove my AR94 without a hitch, with never an indication that they were struggling.

And it (170 DC) did very well indeed with my friend's Apogees, a boast very few amps can make.

While this speaks well of Marantz' technology and know-how, much more imporatnt than the sheer power is the fact that it maintains its posture and overall composed sound right up to its top.

I should also add that this particular model, as well as some other models of theirs which are held in high regard and fetch the best prices today were made in the 1978-1980 period. Looking back, one will easily see that the other big names of that time were also producing what was probably their best in regular production (not some special models, special series). I imply such companies as Sansui, Kenwood/Trio, Pioneer, Onkyo, Yamaha, etc. In many ways, that could be called their golden period. It was just after 1980 that the video craze caught on, with the PC following soon after, when audio took first the back seat, and then stepped out of the public imagination's eye by 1982 or 1983, when enough IBM PCs were already around, and the first clones began to appear in significant numbers from Compaq, Olivetti, UK's Apricot, etc.

There's a season for everything. That was the pinnacle for Marantz, the last series designed in the U.S.A. and manufactured in Japan. After that, it progressively became more and more a purely Japenese company, until it became just one more kid on the block, losing what made it special. Their very clever circuitry soon gave way to simplistic circuits which cost less, very alike to other Japanese products of the day, and with a similar sound. That just wasn't the real thing any more.
  Reply With Quote
Old 31st March 2012, 05:35 PM   #4163
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Hi,

Quote:
Originally Posted by dvv View Post
Fine, but when? In 2005?
I build stuff based around the east german copies of the TDA2030 and 2040 in the 80's. While not explicitly made as DC Amp I used them that way.

Making a DC coupled Amp is trivial and has been for many decades.

Quote:
Originally Posted by dvv View Post
Also, as you yourself pointed out, a powerhouse it isn't - whereas the Maranz could be said to be approaching one.
Not sure if that "could be said". Mine was not designed as powerhouse, but if you want to drive very low loads you can use external transistors (the precise circuit John Curl showed for the Headphone Amp in his Blowtorch Thread "MK II").

Quote:
Originally Posted by dvv View Post
much more imporatnt than the sheer power is the fact that it maintains its posture and overall composed sound right up to its top.
To me it matters more that sounds good "down to the bottom", though I like stuff that goes to 11. If it goes to 11 but is bad at 1-10, I could not care less.

And I'd rather trade a good 7 for a bad 11...

Quote:
Originally Posted by dvv View Post
I should also add that this particular model, as well as some other models of theirs which are held in high regard and fetch the best prices today were made in the 1978-1980 period. Looking back, one will easily see that the other big names of that time were also producing what was probably their best in regular production (not some special models, special series). I imply such companies as Sansui, Kenwood/Trio, Pioneer, Onkyo, Yamaha, etc. In many ways, that could be called their golden period.
This we agree upon.

The top of the line stuff by Japanese Majors from the late 70's to around the Early 80's stands up to almost anything out there, given decent condition and loving maintainance. It was downhill from there, though we seem to be seeing a bit of a Renaissance (or just the dead cat bounce).

Ciao T
  Reply With Quote
Old 31st March 2012, 06:09 PM   #4164
KSTR is offline KSTR  Germany
diyAudio Member
 
KSTR's Avatar
 
Join Date: Jul 2007
Location: Central Berlin, Germany
FWIW, see pic for my take on that neg-rail referenced super-CG, showing the concept. There are gotcha's, as usual.
Attached Images
File Type: png super-GC.png (10.3 KB, 180 views)
  Reply With Quote
Old 31st March 2012, 06:33 PM   #4165
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
Thanks for that, Klaus. I haven't studied it yet but what are the main "gotchas" and the main advantages?
  Reply With Quote
Old 31st March 2012, 08:10 PM   #4166
KSTR is offline KSTR  Germany
diyAudio Member
 
KSTR's Avatar
 
Join Date: Jul 2007
Location: Central Berlin, Germany
Tom,

Pro's :
  • Circuit input and output AC-referenced to the same point than the chip's internals, the negative rail.
  • The remaining little ripple voltage on the opamp input vs reference is fully linear correlated to load current vs. a chopped copy + PSU ripple. And no PSU noise here.
  • Supply and load ripple runs pretty much into the void on the positive rail, which has much better PSRR.
  • DC-coupled, no output capacitor (as it INSIDE the full feedback).
  • Also works with true single rail, with only little compromise (R1 mutates into half-supply resistor divider then).

Con's :
  • Can't sustain long term DC or extremely LF signals or large chip voltage offsets, resulting in more than a few mA, otherwise the midpoint shifts (a voltage drop along R1) until balanced is restored -- well that's a pro as well, DC protection.
  • Can't be used with speaker which don't have a DC path. Output saturates to a rail without a load connected.
  • Need some protection, especially for power down, for both the speaker and chip. One diode from GND to chip PWR+ does the most important thing.
  • Needs higher voltage caps, or series connected ones (with bleeders/stabilizers).

In fact I'm using something like a 24V/3W lamp for R1, that catches a couple of things well. Output snubber also goes to the neg rail. It can be used for multiple channels without effort.

The circuit is useful for insight on what decoupling really is about.

Basically it is a single supply amp, but the output cap is inside the feedback and is not charged to mid-supply through the load. The mid-supply reference is derived from and used at the back end of the amp, not the front end.

I'll try to post a full-blown example amp (P2P style) in the chip amps forum some time.

Last edited by KSTR; 31st March 2012 at 08:23 PM. Reason: Typos and real errors
  Reply With Quote
Old 31st March 2012, 09:43 PM   #4167
tvrgeek is offline tvrgeek  United States
diyAudio Member
 
Join Date: Dec 2009
Location: Md
I get by with 50-ish solid state because I have a 4th order electronic crossover at 80 Hz. I don't play things terribly loud, which maybe why we both still have our hearing.

The cheap amp was to get a chassis to play with. It is total doo-doo, but served quite well to understand tube amps. Funny how we demand power. My dad had the giant powerhouse of the day, a 30W Allied to drive his AR-2. Unheard of at the time. I guess the Dynas broke that barrier.
  Reply With Quote
Old 1st April 2012, 10:23 AM   #4168
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Hi,

Quote:
Originally Posted by KSTR View Post
FWIW, see pic for my take on that neg-rail referenced super-CG, showing the concept. There are gotcha's, as usual.
I basically use this concept for the capacitors local to the individual LM3875 in my Gainclone, with 220uF Silmic Cap's and various Film Cap's as bypass/snubber.

The main PSU section is traditional symmetrical, rectifier -> 2 * 3,300uF -> tripple 0.235 Ohm resistors -> 3,300uF forming a 200Hz lowpass, incidentally. Rails are around +/- 32V without load, 320VA torroidal transformer.

Incidentally, power levels are comparable to my Tube Amp, which is not DC coupled and does not use looped feedback, but subjectively sounds notably better than the Gainclone.

My "big" Vincent Solid State Amp with 180WPC comes in a distance third on sound quality, except raw power, which is why it will have heck modified out of it, so hopefully it will no longer sound like the amp from Krell, ooops, Hell...

Ciao T
  Reply With Quote
Old 1st April 2012, 08:38 PM   #4169
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Folks,

Quote:
Originally Posted by ThorstenL View Post
FWIW, here my take on what to implement in a nice case, with a pair of 0.3K/W heatsinks and a 1200VA mains transformer...

...

PS, the attachment has not been build or tested (yet), but is intended to be implemented to some existing hardware, using existing PCB's and as many existing parts as possible, while getting something more to my liking.
So I posted and found the design simmed and criticised by others... I made some changes and got fed up with working in the "pen, paper, slipstick" domain.

I recently found the latest update of Tina-Ti included more devices and stuff, so rather then spending time in LTSpice, I assigned some moderately close devices found readily in Tina-Ti, to get a basic idea. Tina-Ti has various areas where results are either seriously funky, or their display is funky. It is necessary to discount/adjust for this. For example it's Fourier shows a "constant height grass" growing at higher harmonics if general HD is low, this seems a process artefact, as other simulators and reality beg to differ.

There are considerable differences between the modelled devices (which all seem to be 1960's/1970's devices - going as far as forcing me to use TIP33/34 as Outputs and IRF610/9610 as Drivers). So some devices ended up worse, others possibly better than what is on the page.

For example, the drivers in the Sim have way more Crss than the actual devices, but they have less than halve the output impedance too, the J-Fets have a touch more capacitance and transconductance.

So the simulation results should be taken "cum grano salis magnitudinem montis". I attach the current schematic, including the final DC nulling circuit (which combines very low gain servo and trick circuitry to get low offset and low sonic footprint).

Click the image to open in full size.

Anyway, simming cascodes convinced me that even heavily degenerated VAS benefit from cascoding. The added complexity of cascoded VAS vs. Parallel VAS is modest. So Hawksford cascoded, J-Fet buffered push-pull VAS it is. That said, I'd not loose excessive sleep without cascode.

In fact, the circuit uses many features that are also shown to reduce circuit thermal memory, but as feedback is low, memory distortion is minimised anyway.

I did most of my investigation with 50KHz sines and squares, adding 10Khz and 1KHz tests for completeness.

I also simmed the same circuit as 16dB Gain line stage (see the High End Tone Control Thread to see what circuit that will drop in) as well. This means a joined shunt regulated +/-60V supply and no BJT outputs, just a pair of Fets.

Overall I am reasonably happy with the results.

The Linestage sims at < 0.001% THD (H2 dominant ~ 10dB+ above the rest) for around 12V / 50KHz into 10K//1nF load. I guess the sim is overly enthusiastic, given only 36dB NFB, or maybe not... I guess time to drop the open loop gain...

Other measurements are way past good or bad.

The Poweramp sims as 0.1% THD (H3 dominant) for 26V / 50KHz into 8Ohm with 14dB global NFB. For reference, the 1KHz result is around 0.03% (equal H2/H3) for 26V/8Ohm.

Anyway, these two observations illustrate that most of the distortion in the Poweramp derives from the output stage.

Additionally, with the filter on the input and the antique (1968) TIP33/34 in the output stage (with 3MHz Ft vs the more modern transistors with 30MHz Ft) the 50KHz square-wave suggests > 75V/uS slew rate, I suspect with much lower capacitance from the lateral driver Fet's and much higher FT for the output transistors the real HF distortion and slew rate will be much better.

Given the parts models used and the real parts that will be used I am quite happy with the simmed results.

Ciao T
Attached Images
File Type: gif Sir Pottymouth Cascoded.gif (34.6 KB, 224 views)

Last edited by ThorstenL; 1st April 2012 at 08:55 PM.
  Reply With Quote
Old 1st April 2012, 09:26 PM   #4170
a.wayne is offline a.wayne  United States
Sin Bin
 
Join Date: May 2008
Location: Front Row Center
Quote:
Originally Posted by ThorstenL View Post
Hi,



I basically use this concept for the capacitors local to the individual LM3875 in my Gainclone, with 220uF Silmic Cap's and various Film Cap's as bypass/snubber.

The main PSU section is traditional symmetrical, rectifier -> 2 * 3,300uF -> tripple 0.235 Ohm resistors -> 3,300uF forming a 200Hz lowpass, incidentally. Rails are around +/- 32V without load, 320VA torroidal transformer.

Incidentally, power levels are comparable to my Tube Amp, which is not DC coupled and does not use looped feedback, but subjectively sounds notably better than the Gainclone.

My "big" Vincent Solid State Amp with 180WPC comes in a distance third on sound quality, except raw power, which is why it will have heck modified out of it, so hopefully it will no longer sound like the amp from Krell, ooops, Hell...

Ciao T
Vincent's are no Krell , there maybe hell to pay thou , trying for good sound ..
  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
Quality Control differences = variations in sound quality? KT Class D 3 4th June 2014 12:02 AM
Sound Card for Measurements Marik Solid State 2 2nd January 2012 08:59 PM
Sound Card Recommendations (For Audio Measurements) dchisholm Equipment & Tools 5 16th July 2011 09:40 AM
How to protect sound card during amp measurements? okapi Everything Else 13 2nd September 2008 03:06 PM
Sound cards - test and measurements jackinnj Everything Else 2 5th July 2003 03:02 PM


New To Site? Need Help?

All times are GMT. The time now is 09:32 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