My "audiophile" LM3886 approach

needlz said:
Would I really need to use a headsink that big or can I use a 1"x1"x5" piece of solid copper?



A chunk of copper that size should be fine. The overall heat generated is fairly low. The chip will heat up with peaks, that copper should absorb a lot. The limited surface area will limit the average heat handling.
If you cut slots in one side it should quadruple the surface area. This would make it more like a normal heat sink.


George
 
I have my first "official" ref amp built. :)

Here it is.. some pics first...

The case is temporary until my parmetal case gets here. But it is quite functional, just ugly, well to look at anyway. It sounds as good as I hoped. :)
 

Attachments

  • m1.jpg
    m1.jpg
    86.1 KB · Views: 4,548
Re: I have my first "official" ref amp built. :)

Russ White said:
Here it is.. some pics first...

The case is temporary until my parmetal case gets here. But it is quite functional, just ugly, well to look at anyway. It sounds as good as I hoped. :)


Excellent job! Congratulations!

Does it sound better than the first version?

Some questions on your assembly:

1) You seem to have soldered the LM318 directly on the pcb, with no sockets? No intention on trying different chips there?

2) What resistors did you use? They look like Holco.

3) How do you hold the pcb on the chip side? No stand-offs on that side?


Carlos
 
Re: Re: I have my first "official" ref amp built. :)

carlmart said:



Excellent job! Congratulations!

Does it sound better than the first version?

Some questions on your assembly:

1) You seem to have soldered the LM318 directly on the pcb, with no sockets? No intention on trying different chips there?

2) What resistors did you use? They look like Holco.

3) How do you hold the pcb on the chip side? No stand-offs on that side?


Carlos


Hi Carlos,

This is a "Revision A" amp. I say it is my first official one because the other one I built did not use all the smae values as Mauro did, and It was built on my own PCB (which was not great quality). This one is much better in every regard including sound.

I will now answer your questions:

1) I did directly solder the op amp as I prefer this to adding the inductance of the socket legs (especially as it seems to adversely effect the byapass caps). If I want to change it I will simply desolder. :) Mauro has warned me against other opamps for this circuit as it is designed specifically to attributes which seem to be unique to the LM318. But DIY is all about risk taking right? ;) So who knows, maybe soon. :)

2) Resistors are KOA Spear, IRC, and Vishay-Dale, as well as Yageo.

3 The PCB is held by two standoffs in front, one towrad the center, and then it is fastened to the heatink. No other support. None other needed it is quite solid.


I am still forming my impressions of the amp, but I can tell you it has been a very pleasant evening so far. :):cool:
 
Hi.

Decent sort of heatsink there Russ :)

I'm making a case out of 4mm Aluminium, somewhat along the lines of Vikash's case but with different dimensions. Here's a photo of his case to save you searching for his site:

[IMGDEAD]http://www.vikash.info/audio/gainclone02/images/internals01.jpg[/IMGDEAD]

I'm thinking to put the tranny in the front, and attach the chips directly to the rear panel, or even to insert some copper like Vikash did between the chips and the case (Vikash used the base of the case and turned his BrianGT clones on their side):

[IMGDEAD]http://www.vikash.info/audio/gainclone02/images/internals03.jpg[/IMGDEAD]

Do you think I should add an external heatsink as well, or will the aluminium case be enough? I'm planning on using a 2x24v 400VA tranny, and my speakers are mostly 4-8ohm

Michael
 
OK, one small oncern I have. The amp runs hotter than any of the chipamps I own(which are legion) but, I can't hear any sign of ocillation(ringing and such). I don't have acces to a scope for a couple of weeks. :( So I am left guessing if it is really oscillating or not.

It is not terriblly hot, but I wonder if may be oscillating anyway,as it does not cool down even when I leave it idle for 20 minutes or so. The heatsink stays a pretty constant temp (about 56C).

I wonder if adding a zoble on the output terminals might help...
 
It may be oscillatin. Mine is only getting mildly warm. Maybe you shoudl put some holes under your heatsink, or better yet, cut out a small rectangle under and over it to allow the air to flow.

Mine did run warmer when I first started listening to it then got cooler (only 1-2 hours of use). It was never hot though. My BrianGT 4780s get quiet warm, however, even with a Zobel. Fairly sure they are oscillation. I need to buy a scope.
 
BrianDonegan said:
It may be oscillatin. Mine is only getting mildly warm. Maybe you shoudl put some holes under your heatsink, or better yet, cut out a small rectangle under and over it to allow the air to flow.

Mine did run warmer when I first started listening to it then got cooler (only 1-2 hours of use). It was never hot though. My BrianGT 4780s get quiet warm, however, even with a Zobel. Fairly sure they are oscillation. I need to buy a scope.


Yep, already have vent slot over and under the sink. Maybe I should not call it "hot" just warmer than I am used to.

I have both parallel(Brians) and stereo(my own) 4780s and they run pretty cool, cooler than this.
 
Ok, I think I may be onto the problem(well I now I think I may not have one) I turned off the amp 15 minutes ago. And the heatsink is still pretty warm. I think the difference is the type of case I built.

I usually (on my previous chipamps) put the heatsink with the entire length of the fins exposed to the outside air. With this amp only convection makes the air move inside the case, so much less fresh air gets through the case.

I any case I have been listening for 8+ hours and its still singing "mighty pretty." So I don't think its a big deal. I hope :xfingers::xeye:
 
Lookin' Good!

I have only got as far as matching up some of the caps and resistors for my first one. I plan on building the C version. I will leave the leads long on the backside to solder the added parts to.
What is the power supply that outputs AUX for? I see it, but do not see where it connects.
Waiting on Mouser to deliver the BC546 and BC639 transistors also. Going out this week to get a relay local.
I am matching up two sets of components also. Going to build another when the latest boards bwcome available.
My first one with be a two box unit. Want to keep the transformer at a distance. Also the box is a big cast aluminum that barely hold the board. I will mount the transistors to the box.
If I can find a nice box large enough, the mono boards will be used in a single chassis.


George
 
Re: Lookin' Good!

Panelhead said:
I have only got as far as matching up some of the caps and resistors for my first one. I plan on building the C version. I will leave the leads long on the backside to solder the added parts to.
What is the power supply that outputs AUX for? I see it, but do not see where it connects.
Waiting on Mouser to deliver the BC546 and BC639 transistors also. Going out this week to get a relay local.
I am matching up two sets of components also. Going to build another when the latest boards bwcome available.
My first one with be a two box unit. Want to keep the transformer at a distance. Also the box is a big cast aluminum that barely hold the board. I will mount the transistors to the box.
If I can find a nice box large enough, the mono boards will be used in a single chassis.


George
Hi George,

Thanks. :D:cool:

I am not sure what Mauro intended the aux power output for, but I did not use it. It will get you about 13.5VDC. I suppose you could use it for a fan or something.

I put a pin header on it(just in case), but I am not using it.

Good luck!

I am also building my own dual monobloc "rev C" amp based on my own PCB. I will share my results later. :D

Cheers!
Russ
 
My listening impressions... so far... :)

Ok, I have put this thing through about 10 hours listening. The audience was my wife, my brother in law, and myself. :) I have included highlights of the listening session as follows.

I will provide the notes I took as a narrative.

1) Song: "Lovin' Touchin' Squeezin'" by Journey.

Wow, the bass has real authority and grit, and Steve Perry's vocals stand on edge. You can really feel each note of the lower registers of the piano. I love it. The highs leave nothing to be desired. Excellent high hat and you can actually tell that is an inverted symbal. Wow.

2) Song: "Tryin' to throw your arms around the world" by U2.

This is admitedly not the greatest recording, but it has some great left right stereo details which I use to judge channel seperation. I like the detail of the quiet guitar details on the right channel, and the ringing metalicness of the keyboard parts is distinct on the left. As close to a perfect 10 on this song as I have ever heard. Bono's vocals are badly compressed, but thats not the amps fault, still you can easily hear the quiet lyrics in the part right after the last bridge. Excellent!

3) Song "Money for Nothing" by Dire Straights.

This is an all time classic for testing amps(And a little worn around the edges). :) I use it all the time. The opening drums to the first guitar riff simply shine. That guitar riff with the bass counterpoint leading into the verse is sweet! And the the seperation and soundstage is clear and distinct.

4) Song "Low Self Opinion" by Rollins Band.

One of my favorite grunge songs. :) You have to dig Rollins vocals and they come right to the front. The recording sounds very "live" and this comes through with excellent authenticity. That bass riff just throws it down. Awesome, all this with no subwoofer! :D
Favorite Line "if you could see the you that I see when I see you you'd see yourself much differently, believe me". :) Great lyric.

5) Song "Love Me Like a Man" Dianna Krall.

Wow, this is a great female vocalist, this recording is SACD and I love it. Dig the piano, its like it was a few feet away, and the silky soft way she delivers each note. You can hear each breath she takes, and the nuance of the scolding tone of voice she takes at times. I find myself playing it twice. Excellent.

6) "Through Her Eyes" Dream Theater.

I love this group, So I am particularly picky about how it is reproduced. James Labrie's vocals get the attention they deserve, you can here the emotion in this great ballad, the quiver in his voice as he sings, distinct and correct. John Petrucci's genius on the guitar is evident throughout and it is as if you were on in orchestra seats on the Gaylord Entertainment center as the performed this song. Brings a tear to my eye. Excellent.

7) "Symphony No. 2" Mourton Gould, performed by the Albany Symphony Orchestra, David Allen Miller conducting.

This is a great american composer, and one of the modern era's early masters. Listening to his symphony is somewhat like viewing a picasso, the genious is evident, bu the message requires thought. The timpani, and the xylophone add texture to a complex, and never boring ride through the thought process of a early modern impressionist composer. The presentaion is spot on, with just the right amount of in your face "up frontness". You can feel the smack of the kettle drum. A+.

Thats all I have time for for now. I can't wait to listen to more.

Thanks for this fabulous design Mauro. You have made this man very happy. :up: :cool:

Cheers!
Russ
 
Hi Russ,
I am happy that the first impact with my "electronic tastes" you pleasure. I have built and listened diverge amplifiers, but this prototype is native exclusively on the basis of my listening impressions, privileging the "presence", the 3D feeling and realism of the sound stage.
A lot of amps has a good tonal equilibrium or a good dynamics, but a few are able to create a " presence feeling " of the musicists. I love this type of amps, and I hope of is me approached with my_ref.
Technique:
doesn't use zobel net on this circuit. The particular structure of this amp is more unstable with the zobel net that without.
To the your first descriptions of the temperature it is worried me, but only because have not the possibility of use an oscilloscope.
The circuit would not have to create problems of spontaneous oscillation ( if all the components are of the correct value ), to the maximum would be able become more unstable on load capacitive (pure capacitive!) or to the "hard" clipping.
My circuit " squeezes " much LM3886 ( during the work ), and the powers res. are "hot". If close the box after a few time the temperature it settles on a medium value.
Obviously it is a lot of important to perform some preliminary test to establish with safety the absentia of oscillations...

Ciao

Mauro