My "audiophile" LM3886 approach - Page 4 - diyAudio
Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

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 6th April 2005, 06:05 PM   #31
diyAudio Member
 
Join Date: Dec 2004
Location: gran sasso
Dear Mauro!

please don't misunderstand me - I am aware of the functionality of this config - I would quote myself:
Quote:
So as to be clear: the LM3886 is a voltage controlled bipolar current source in this configuration. It does not have any voltage gain; it has an input voltage/ output current transfer characteristic - which is, in this case, is 100.
What I wanted to point out, is that there are factors in it's operation which will make it distort, deviate from the ideal transconductance function. [to add to the previously mentioned, there is also the output impedance of the U1, non zero, and which "sees" a load current proportional to the output signal] Then this distortion is corrected by U1.

And it should be emphasised, that your design is visibly very well thought over, refined into the little particulars [like the bypass conf. for IC1; like, indeed, the input compensation / filter cap on U1]

So, really, I am far from attacking You - I was just provoking You, so while in defense, you would make it more clear your design decisions.

Also, I am not questioning the your choice for U1. And anyway, people can [and will..] roll their own.. Some other choices adequate in this position may be the LM6171; AD825; AD8065; AD8033..

May I notice one thing? As far as I understood, the "current dumping" method is useful in eliminating the low signal crossover distortions because of the contribution [included in the bridge] of a second, low power & class A output stage. Here that is missing.

Ciao, George
  Reply With Quote
Old 7th April 2005, 12:50 PM   #32
diyAudio Member
 
Join Date: May 2005
Location: TN
From your analysis on the circuit had not doubts.

My answer has been ample to grant even the request for other interlocutors.
The "ideal funtion transconductances" is "distorted" simply because that circuit is a compromise (as all the circuit).
For the precision, U1 introduces an impedance of exit ( open loop ) about to 50 ohms, does that maintains the balancing of the bridge within 1 %.
A perfectionist would be able insert a trimmer in series to R8 to compensate even this problem
(and uses resistances after 0, 1% or select).
the obvious solution to improve is: increasing the value of the resistor. But there is a problem.
Increase the delay of group of IC1 and the bridge shows traces un parasitic capacitance ( C31 eras been scheduled for this problems ).
Sincerely, given that my principal objective is "musical", after has tried practically all the variations,
I have stopped me on a compromise practise. I am sure that some is able find better "balancings".
(I have published this scheme one's own for this!).

U1 matter: I don't want do an extreme defence of LM318, but I have a pragmatic mentality ( the mile resulted to the minor costs ).
To leave the free field, I say at once that probably some of the OPamp that you have quoted better results in my circuit.
I rest "cools" to the "enthusiasms" on the chips because am convinced that the strength of the complex circuits is in their "equilibrium".
Is banal to show that a OPamp of new generation are able improve the things. Naturally, you am happy if succeed to improve the performanceses of this circuit.

Variation of " Current dumping"; This circuit is not "homologated" in accordance with the characteristics of the patent "dumping current"
(what is described as you have said).
I have defined it a "variation" for this, but am open me badly.
The good performances on the crossover THD is that verify on my circuit.
Evidently this circuit gets was characteristic of the " dumping current ", without use entirely his princes.

Is not a lot of difficult to shape this 2 OPamp in stasis, dumping current, feedback current, negative impedences, and with a modest work even something of more amusing, as Power feedforward. In practice, much of this topology is more elaborate of my circuit and don't great improve the performances.

Ciao

Mauro
  Reply With Quote
Old 19th April 2005, 02:55 PM   #33
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
Quote:
Originally posted by Joseph K
Mauro,

I'm highly intrigued by your design, and would like to understand a bit better.
All designs are a kind of compromise. So there are strong / weak points in all. While I was thinking of it, some weak points came to my mind. If you don't mind, I would list them, so you can tell the reasons behind your tradeoffs.[snip]It seems to me that this is quite against the tendencies presently followed - which is giving up some gain thus lowering the feedback & extending the open loop gain bandwidth, as much as possible.
[snip]Ciao, George

Hello George,

I like your analysis of this highly sophisticated circuit. What intriques me is the refernece to the "bridge" nature, meaning IIUC a current dumping aproach. But I see at the LM3886 a combination of pos and neg feedback, which brings memories of a Howland Current Pump, which is also a type of balanced bridge, to get a VCCS as you or somene else remarked.

In all these cases, the stability issue is very complex and depends on several zeros and poles in the circuit. The many compensation parts also point to this.

It may well be that the LM318 is one of the few that works in this particular circuit. It has a particular OL charateristic and for example has assymetrical slewrates, IIRC, which may also play a role here. IMHO, you cannot just replace opamps at will here like in a buffered gainclone. Here, the open loop opamp characteristic is an integral part of the circuit.

You are right that it goes against the trend by SOME to lower OL gain and wider OL bandtwidth, but as you know that in itself is a controversial issue and by no means universally accepted. For me, there are many ways to skin a cat and this circuit is an interesting one.

Jan Didden

PS I still have a few tighter spec'd LM218's, metal can, if anyone is interested.
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W. S. Maugham
Check out Linear Audio!
  Reply With Quote
Old 19th April 2005, 06:03 PM   #34
diyAudio Member
 
Join Date: May 2005
Location: TN
Good point of view and good description of the LM318 characteristics,janneman.
Is for this that have chosen this component. The stability of the circuit allows however to use some other type of chips. For what concerns the select philosophy, as I have already said this is the compromise that gave me improve general results, but on this circuit I have experimented nearly all the audio technical known ( driver current , dumping current ,stasis ecc...). The choice to working with all the negative feedback in this circuit has been taking simply because things rang better ( I have tried even to replace U1 with differential discrete circuits, with greater band and low OL gain, but without substantial improvement). Perhaps an excellent audio result would be able get using the current bridge in stasis configuration with a valve voltage driver, or high quality discrete solutions.

Mauro
  Reply With Quote
Old 19th April 2005, 06:49 PM   #35
diyAudio Member
 
Join Date: Dec 2004
Location: gran sasso
Dear Jan!

First of all, would like to emphasise again that also I 'am really happy that You are back! [and as sharp as ever].

Yes, this sophisticated circuit by Mauro IS a Howland Current Pump application, and his reference to a bridge was exactly in the way that you described.
And it is really interesting, and also his another, new thread. The best thing would be to be able to listen to his creations, it's a pity that, as far as I understood, we are not exactly neighbours.

In my "analysis" I was just thinking aloud about his possible reasons behind his choice of this more complicated circuit, instead of the simple basic application. As far as I understood, his point is an improvement on the input section of the LM3886. And it can be a valid point.
Also I wanted to show that it is a different approach, with respect to the "Gaincard" thinking, which emphasizes on the speed of the feedback loop.

And it is not that simple to substitute that opamp, I agree. This design of his is really a very well thought over, blanced one. I tried to list some opamps, which are known for good sound, and are having similar characteristics to the LM318.
What were You having in mind with the assymetrical slew rates? It is time to learn for me!

And If Mauro would be so kind to explain more extensively his point about the output Zobel causing stability problems, it would be very interesting, as well.

Ciao, George
  Reply With Quote
Old 19th April 2005, 07:01 PM   #36
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
Quote:
Originally posted by Joseph K
Dear Jan!
[snip]And it is not that simple to substitute that opamp, I agree. This design of his is really a very well thought over, blanced one. I tried to list some opamps, which are known for good sound, and are having similar characteristics to the LM318.
What were You having in mind with the assymetrical slew rates? [snip]Ciao, George

George,

What I wanted to say is that this circuit is really an integrated circuit in the real sense. In a buffered gain clone, you can play around all you like with the buffer from an engineering pov. In this case, changing the LM318 to something else will almost surely require other changes for stability. The assymmetrical slewrate (which we don't see anymore with more recent opamps) may be important in this circuit, that is all I wanted to say. Otherwise I fully agree with you.

Jan Didden
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W. S. Maugham
Check out Linear Audio!
  Reply With Quote
Old 20th April 2005, 06:09 PM   #37
diyAudio Member
 
Join Date: May 2005
Location: TN
Am working to a "Stasis" version of this circuit, to show the superiority of the connection to bridge of this chips.
The first results are excellent.
To lend...

Mauro
  Reply With Quote
Old 26th April 2005, 01:38 AM   #38
G33/33 is offline G33/33  Indonesia
diyAudio Member
 
G33/33's Avatar
 
Join Date: Dec 2004
Location: Djakarta
Hi Mauro,

Does your circuit will work with other type of chip amp such as TDA 7294 ? I have some of them doing nothing so I wonder if it will works with your circuit.

The same question I will ask to you for your Stasis circuit.

LM 3886 and LM 3875 is not easy to find around here, but plenty of TDA 7294.

Thanks,
Gede
  Reply With Quote
Old 9th May 2005, 10:13 AM   #39
diyAudio Member
 
Join Date: May 2005
Location: TN
I excuse me for the delay of the answer.
I have not make a will this chips, but if they have a similar open loop gain and phase margin are able use them. Eventually you have to retouch the R8 C1 net in the "stasis" version if see of the instabilities ( free or induced oscillations by the load ).

If it's of interest, am working to an (radical) improved version of this circuit, on the basis of the "stimuluses" of Jan and Joseph. the results are: improving of the THD and of the quality of the sound. If it is of interest to some, provided that me signal it

Ciao

Mauro
  Reply With Quote
Old 9th May 2005, 11:18 AM   #40
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
Quote:
Originally posted by maupenas
[snip]If it's of interest, am working to an (radical) improved version of this circuit, on the basis of the "stimuluses" of Jan and Joseph. the results are: improving of the THD and of the quality of the sound. If it is of interest to some, provided that me signal it

Ciao

Mauro

Mauro,

Your type of posts is the reason I come to this forum! Let's see what you have!

Jan Didden
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W. S. Maugham
Check out Linear Audio!
  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



New To Site? Need Help?

All times are GMT. The time now is 02:53 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