My "audiophile" LM3886 approach

BrianDonegan said:
That's weird. I asked them by email a few weeks ago and got this reply:

Hello Brian,

These transformers have dual 115V/230V primaries.

So how would that work? The tranny is just a former with a set number of turns of wire on it that steps the voltage down right? If it's the same tranny that puts out 25V with 240V input, and you power it with half the voltage, wouldn't you see half the output voltage as well, about 12V ?

Michael
 
Russ White said:
Dual supply trafos are very common, they usually just use dual primary windings which are either connected in series or parallel.

Hmm. The website definitely shows dual secondaries, but says nothing about the primaries. Very helpful :)

I have some similar-spec trannies from Altronics, and they are definitely single primaries. They smell a bit, too, does that go away?

If the answer doesn't pop up soon, I'll drive in and have a look for you...

Michael
 
Ok, Just rang the Adelaide Jaycar shop, and got the sales guy to grab one and tell me what he sees :)

No package insert...
6 wires: Orange, Orange, White, Pink, Yellow, Red.
It says 240v/50hz on the shrinkwrap, 130 degree thermal cutout, lists the wire colors, attendant couldn't make sense of the meaning, but there is no mention of 110v anywhere on the product.

Michael
 
Going back

I installed the latest ideas in mine. 17 Volt zeners amd LM334 constant current sources.
It does not sound as good as before. It may be output related. This was only checked at low volumes. The amp seems colder and "smaller" than it did as a Rev C.
May be my implimentation of the LM334Z. The data sheet showed a formula to calculate Iset. The value is Iset = 67 mv/Rset. I used a 16 ohm resistor to get 4 ma. Is this close to the values other have used here?
If this is close to right I may pull the current source and just try the higher voltages on the LM318.

George
 
Re: Going back

Panelhead said:
I installed the latest ideas in mine. 17 Volt zeners amd LM334 constant current sources.
It does not sound as good as before. It may be output related. This was only checked at low volumes. The amp seems colder and "smaller" than it did as a Rev C.
May be my implimentation of the LM334Z. The data sheet showed a formula to calculate Iset. The value is Iset = 67 mv/Rset. I used a 16 ohm resistor to get 4 ma. Is this close to the values other have used here?
If this is close to right I may pull the current source and just try the higher voltages on the LM318.

George
HI George,

I actually agree with you, I am not sure this mod is for everyone. And I reported my results only after a very brief listening session on my bench amp which was classical music at medium levels. My vote is actually not to bias or tweak the zeners unless you feel like experimenting. I have not modded my mains system (Rev C monoblocs) at all, and it remains my favorite.

I am not 100% sure that the 318 is the best candidate for biasing...

For the record:

I used a 100R pot and my DMM to set the current, but measurements show it is close to 16R.

When you add the current source you definitely need to increase the available current to that part of the circuit.

I did that by running 2x 1.5K on each rail next to the zeners. This gets me about 18ma for the LM318 PS.

So my advice...

If it aint broke.... and it aint, don't dix it, but if you feel like it, play all you want! :D

Cheers
Russ
 
maf_au said:
Seeing as the Mauro thread has gone a bit quiet, I thought I'd spruce it up with a bit of nearly off-topic goodness :)

Top to bottom: Russ's FreeBird Pre, Mauro Rev A, New toy - Eastsound E3 CDP.

Now I need to finish the RevC...

Michael

Man thats a nice rig!! :) Saweeet CDP!!! And not a bad collection of Discs either!

Always nice to hear from you Michael!

Cheers!
Russ
 
Hi all,
I not suggestion to modify the biasing points of LM318 , in particular in revC.
The equilibriums of a " nesting feedback " circuit are important, and is not been able modify without a very profound knowledge of all the situations.
The motive for which I have used the 12V and zener is written much turned in this thread.
The active load ( current source ) it is not adeguate. The load with LM318 it is constant and greater of 10Kohm, and LM318 works already on class A to this impedances. Increasing the bias current it serves only to increase the THD.
I don't want do the "teacher", but is not has to ignore the theoretical complexity of the circuit. If not understands perfectly as works the net among pin 6 and 8 of LM318 (revC) , not it may modify the circuit...

ciao

Mauro
 
Changed mine back

I pulled the current sources out today. Sounds much better. Left the 17 volt zeners in.
Cannot tell if this made a big change from the 12 volt ones. It was Monday the last time I listened to it with the 12 volt zeners. If anything, it does not sound worse. May be a subtle thing.
Might have a little better dynamics, hard to compare from a week ago.
Like reported and recommended, this may not be better only different.
Thanks for your input Mauro. I could tell on power up. The RevC was easily better than the RevA in my system from the first note. The current sources and higher voltage were the same backwards step. Never got a chance, five minutes and it was out of the system.

George
 
Production boards here...

Hi All,

Happy weekend!

Hey all, just a note, my monobloc REV-C PCBs are here:drink:. This is what they look like...
 

Attachments

  • pic1.jpg
    pic1.jpg
    85.3 KB · Views: 1,384
Wow, that is one mean amplifier. I wish I had the tools at hand to do that nice metal fabrication. For fun let's show off "the ugliest mauro amp yet" that I made ;)

Here's a picture of the inside from mid-production. Only one set of inputs is wired at this point just to test that the amp works. Since this picture was taken, I have wired all inputs, tidied up the wires and any other fabrication that still needed to be completed. Lots of room left in this chassis for "expandability" :)
An externally hosted image should be here but it was not working when we last tested it.

Here's a picture of the front panel. It's actually has a lot more shine to it than what you see in the picture.
An externally hosted image should be here but it was not working when we last tested it.