Gainclone with pseudo-balanced input?

[Mad_K]

Are there any advantages in power output or current capacity with the dual chips?

The opa549 is a power op, just like the lm3875. No real advantages, except you get to use your favourite balanced preamp . You can use any op you like with the schematic I have drawn, maybe even better results with the lm3875. I haven't tested it yet, but I guess I will at some point. (If I get my hands on them -I have no way of purchasing them, and National won't sample me any )

 

[protos]

AAAAAAaaahhhh!

After a week of burning in and listening to a lot of records I can tell you:

-IT IS SIMPLY MAGIC!

After I finished my gainclone, I have just been listening to music... (As opposed to trying to figure out what to improve, or which amp I should build next) Now, I am more interested in finding more music. Now it's time to watch "Animatrix", and hopefully some more jaw-dropping

Yeah! I sometimes think we spend too much time tinkering and not enough listening to music.
Yesterday I was thinking that the hobby is probably very close to drag racing.They might spend days and weeks fixing their cars so they can "let 'er rip" for 10 secs.
By the way if you want to listen to more music don't start tinkering with battery supplies because that can really get you hooked. It works well on my x-bosoz too!

 

[Mad_K]

I know what you mean Protos; I have battery power in my "El Cheapo" MM-amp... It really makes big a difference. I was thinking about trying it on my BGC, but then I have to learn myself a whole lot of new tricks and tecniques (time-consuming). I will have to try all kinds if different setups etc. Maybe sometime this winther, if I run out of records/get through Half-Life 2. Now I am just going to build a balanced cable driver for my SB Audigy Player, so I can enjoy noise-free PC-sound in my main system

 

[Mad_K]

What can I say...

This evening I got a little tweaky, so I buildt a new module with opa548. Some small changes in the ratios on the resistors. Different coupling cap (RIFA MMK -polyester 0,47uF). Most significant circuit change was that i dropped the 22K to gnd on the inv. inp.

 

[Mad_K]

From my earlier listening to the different opa-chips the opa548 was my favourite, but I this time the difference was much bigger! I know, -I have changed some other things, but I can tell that much of the improvement comes from the chip (I remember how it sounded like in the other setups). Faster, smoother, more oohmp, more detailed -MORE MAGIC. I suspect some of the treble qualities to have improved due to the loss of the 22K on -IN to gnd. The different impedances also plays a part; less noise, lower Z in etc.

 

[Mad_K]

I know this is not the last time I'm tweaking my BGC... That's some of the beauty of having a GC -you can make a completely new amp in an evening.. Initially I chose the opa549 because I thought I needed the power, but not so -> the chassis is stone cold at all my listening levels. The DC-offset also went down with this circuit (as planned) From -16/-26mV to -5/-10mV (this is cold, right after fire-up measurments).


/Mad_K -one happy listener

 

[Mad_K]

I forgot; new overview pic; I also changed power and out cable from stranded/tinned silicon coat to flat copper with teflon coat (2FLAT). At some later stage the gnd wiring will also be like this, but stripping teflon sucks.

 

[protos]

I implemented a test set -up of a DRV134 balanced line driver as a balanced pre for my gc.
First assessment comparing one channel only is that it may be better than my x-bosoz. Extension , dynamics , treble detail really excellent , tight bass as well - more up front than bosoz but not annoying.
Could this be the perfect mate - Bride of Gainclone - for those who want an actively driven balanced gainclone but with minimum hassle ?

 

[tbla]

drv134....

thats what i tried to tell you month ago....use the drv 134....!!!

 

[Matttcattt]

Re: drv134....

I implemented a test set -up of a DRV134 balanced line driver as a balanced pre for my gc.
First assessment comparing one channel only is that it may be better than my x-bosoz. Extension , dynamics , treble detail really excellent , tight bass as well - more up front than bosoz but not annoying.
Could this be the perfect mate - Bride of Gainclone - for those who want an actively driven balanced gainclone but with minimum hassle ?

thats what i tried to tell you month ago....use the drv 134....!!!

im using the drv134 for my bridged and parellelled gainclone, not quite finished yet.

 

[Panelhead]

Re: drv134....

thats what i tried to tell you month ago....use the drv 134....!!!

Hi,
I am serverely technically challenged. Can you share your circuit used to drive a Gainclone in a balanced mode with the DRV134? This is something I would like to try.
I tried a search and found a lot of circuits for bridging, but not for balanced input.

Thanks,
George

 

[carlmart]

Re: Re: drv134....

Can you share your circuit used to drive a Gainclone in a balanced mode with the DRV134? This is something I would like to try.
I tried a search and found a lot of circuits for bridging, but not for balanced input.

Maybe you didn't look well. These were the first circuits suggested on this thread:

http://www.diyaudio.com/forums/showthread.php?threadid=11133&pagenumber=1

Later on some others were added.


Carlos

 

[Variac]

Those look like paralelled, not balanced, just Like George said.
I don't blame him for being confused- I am too but no time to study it-gotta work, gotta work.

 

[protos]

Just follow the BB application notes.All the DRV134 needs is a good small PS and a couple of bypass caps.You can add the vol pot at the input.+ two output caps.
I must say that I had huge amounts of Dc offset (100mV !)even after the output caps which seemed to drift around over several minutes.This disappeared when I grounded the outputs with a couple of 10k resistors.Can anyone enlighten me?
Anyway I am sorry to say that further listening confirms that this actually blows my Pass Balanced Zen line stage away. In hi-end terms its like going to two classes of pre-amp higher with no compromises that I could detect after a few hours of break in.I even tried it with a battery PS with 4 9V batteries (+- 18V) and it was a killer!

 

[carlmart]

Those look like paralelled, not balanced, just Like George said.
I don't blame him for being confused- I am too but no time to study it-gotta work, gotta work.

Sorry, my mistake. Maybe because I didn't quite understand the need to have another active stage in-between just to balance it.

In any case you could try the same design on the Bridgeclone thread and substitute the bridged gainclones with a single GC chip.

Have a look here at how Marchand does it on their amp:

http://www.marchandelec.com/ftp/pm21man.pdf

Protos:

Which circuit are you using from BB's datasheet and which balanced gainclone schematic?


Carlos

 

[protos]

.

protos has attached this image:

As to the DRV I only saw two basic schematics , one just has some output caps added for DC blocking.works for me.

protos has attached this image:

 

[protos]

Can't seem to attach the image of the schematic , just look at the first page of this thread.

 

[MBK]

There is a circuit proposed by Siegfried Linkwitz, using LM3886, with a lot of smart little details such as balanced impedances, input low pass, RF shunt cap between (-)in and (+)in of op amp, and bandwidth limiting cap parallel to feedback resistor. It can be found here:

http://www.linkwitzlab.com/images/graphics/3886amp.gif

along with a description of PSU here:

http://www.linkwitzlab.com/images/graphics/3886pwr.gif

I use that circuit on my biamped dipoles (using active X-0) with an OPA548 and it sounds very good indeed. It just works. The topology though meant for single ended is already practically balanced - all you have to do is connect (-) bal out to the feedback-to-gnd resistor of the (-)input of the op amp.

I have also listened with headphones, very very sweet. No PS hum of any kind and very, very low noise (sounds sweeter than my dedicated headphone circuit using OPA2134 and BUF634 which I will soon scrap unless I can tame it). DC offset of 9-20 mV with no DC blocking caps. No branded components anywhere btw.

Incidentally I can't hear a difference between non-inverting, inverting, or pseudo-balanced (incoming gnd connected to feedback loop gnd, balanced impedance in incoming gnd) connection mode IF using this topology with these values. The topology lends itself to using any of the above. And no Zobel networks, output stabilizing R, or output inductor network seemed to make a difference, it sounded good right with the basic circuit.

MBK

 

[carlmart]

Can't seem to attach the image of the schematic , just look at the first page of this thread.

It's an unusual circuit for balancing the GC as an opamp. The 22K resistor to ground from inverting input may be unbalancing things.

The resistors you use should be in equal pairs, I think. So there should be another 10K resistor to ground at the input.

The Marchand solution, 100K to ground on both inputs before gain series resistors looks a bit better.

Sorry to insist: what resistors are you using at the DRV134 output? In any case the datasheet mentions an output offset.

Have you tried this same setup unbalanced, with a buffer chip?


Carlos

 

[protos]

It's an unusual circuit for balancing the GC as an opamp. The 22K resistor to ground from inverting input may be unbalancing things.

The resistors you use should be in equal pairs, I think. So there should be another 10K resistor to ground at the input.

The Marchand solution, 100K to ground on both inputs before gain series resistors looks a bit better.

Sorry to insist: what resistors are you using at the DRV134 output? In any case the datasheet mentions an output offset.

Have you tried this same setup unbalanced, with a buffer chip?


Carlos

Well, I followed the instructions of the man himself Thorsten Loesch , alias Kuei Yang Wang to balance impedances from a posting on the original Just another gainclone thread.Originally I had followed the Marchand type set -up with equal resistors but he said that this was unequal in practice.
As I said my DRV has 10k output resistors.The schematics mention only a couple of mv possible offset which is supposed to be cured by the output caps but I was getting up to 100mv offset peak which was then drifting all over the place over time.
I havent tried other buffers because I wanted a balanced driver (in one package) since I believe (pseudo)balanced is the way to go .
Anyway I finished the second channel yesterday and the results are really very very good even though I am using a passive pre as a volume control on the input to drive the prototype which means one set extra interconnects.
I really never expected it to sound that much better than my BOSOZ.
Plus today I was listening to it battery pwered with battery powered GC. It rocks!