My "audiophile" LM3886 approach

Another picture, closer in.

Notice I am not using input caps. I have a jumper there instead as I know I have no DC from my source.

Hey Max, the stickers only mildly effect sound, but I would get them as soon as possible. ;)

Cheers!
Russ
 

Attachments

  • revc1 003.jpg
    revc1 003.jpg
    81.5 KB · Views: 1,255
moreismore said:
Russ, there is no ground on the IEC block. How is anything to find ground?


You do not need a connection to "earth" for you circuit to have a "0V" point which unfortunately is also labeled "GND". The circuit GND is simply the midpoint between the rails. The chasis is wood, so there is no need to connect "earth" to the chasis. Mains ground is there for protection, not electrical function.

Mains ground and circuit ground should not meet as if they do you can introduce all sort of ground loop issues with other equipment which may be attached to your amp. This is a huge source of frustration for people, but it is so simple. Isolate circuit ground completely from your chasis (your jacks/binding posts are important here), and never connect mains GND to circuit GND.

Now some people try some techniques for joining the two GNDs, but there is abosultely NO valid reason for doing it. Circuit GND is virtual, it is a moving target, it is not provided by an external source like the wall socket. :)

Cheers!
Russ
 
OK... I'm a hack compared to you guys...

But, some connection between chassis and the star is better than connecting components through the GND of the interconnects (using SE). I usually effect this connection with a small value RC. For example, 10R bypassed with 1uF. The results are very good I think... QUIET.

I'll hang up and listen for your answer.
 
Ok, here are some closeups of the RevC 1.1 I built last night.

Notice that I have recently made some kit changes.

Namely, I have swapped the input cap (1uF) to an Epcos 305V MKP because I can get them (stock of Wimas through Mouser is sporatic) and they sound great, so...

Also, after Russ did some experimenting, I swapped the feedback cap (C9 220uF) to a Panasonic FM from FC (Ultra low-ESR).

Two more things... I used the MUR820 diodes for this build, and the LM3886 is not mounted in these pics, as I usually mount that to the heatsink first, then solder it to the board so the hight matches up.

Overall top shot:
An externally hosted image should be here but it was not working when we last tested it.


Close up top shot:
An externally hosted image should be here but it was not working when we last tested it.


3886 side close up:
An externally hosted image should be here but it was not working when we last tested it.


Relay side closeup:
An externally hosted image should be here but it was not working when we last tested it.


Diode side closeup:
An externally hosted image should be here but it was not working when we last tested it.


Zener side closeup:
An externally hosted image should be here but it was not working when we last tested it.


Shameless self promotion shot - Darwin* and Joshua Tree** along with RevC:
An externally hosted image should be here but it was not working when we last tested it.


* Doesn't show the rotary switches

** These are made from the boards with errors. New boards should be here any day...
 
chipco3434 said:
I'll hang up and listen for your answer.


Short answer is do what works for you, but I have never connected circuit GND to the chasis, and all my amps (including this proto amp and my mains amp which is in a par-metal case) are dead quiet. All of the gnd points only ever meet at their respective amplifier PCB which is where the star is. If you place a star anywhere else you have multiple stars, which would not be better. To maintain the itegretity of the star GND you need to make sure nothing can bypass the star. The star GND is built into the design.

Cheers!
Russ
 
I built the Twisted Pear Rev C just as Mauro instructed and Russ and Brian have promoted - with the chassis and circuit grounds separate. The result is impressive quietness. The only connections I made to the chassis ground were the mains earth wire and the chassis signal interconnect and chassis mains cable shields, (I posted photos page 205). It is important to ensure that the chassis RCAs and speaker binding posts are insulated from the chassis.

This really is a great amp. It is very transparent and will reveal flaws in the signal fed to it.
I've been mainly listening to it in my secondary front end which comprises a Garrard 301 on a Slate Audio marble plinth and power supply. The arm is the Dynavector DV507 and modified Sound Reference valve preamp.
The Grado Platinum Reference cartridge I first used began to sound very good after breaking in (50hours), but still lacking in some areas.
I'm now using a benz Micro MC3 cartridge that Benz Micro has fitted with a new boron cantilever and Fritz Geiger FG11 stylus.
The sound I am getting through the Rev C is almost unbelievable. Far beyond any expectations that I had. I almost think it must be some sort of fluke of component symetry. The MC3 has enough output to drive my preamp without a stepup, but only came to life when I used it with an AT650 stepup transformer on the 20 ohm setting.
The sound is as direct and immediate as can be imagined. It doesn't roll off the frequency extremes. It is clean,detailed,dynamic,with great separation and definition. Bass goes very deep with detail and impact.
This is a new experience for me. It really grabs your attention.
If anything it could be toned down a notch. I think this may come once I fine tune the sound with some pickup transformer resistive loading.

One thing I notice is that I do not feel the need to bi-amp my large floorstanders, as I do when using only one pair of my Quicksilver 8417 60W monos.

The bridge rectifier is a 25A one. I bought MUR1010 Schottky diodes for the latest RevC boards, but I cannot imagine them sounding better than what I am now using. Miught be some time before I build them though.


The point of my account is to show that the Rev C is capable of great sound if you feed it with the right signal.

Sorry for the long (boring?) post . There has been little posted about peoples impressions/experiences of the Rev-C, so I think it may be time that the silence is broken.

Audie.
 
Hi Nordic.

I've been enjoying the Rev C ever since I built it in late January!
It is just that I am enjoying it even more, now that I am using a different cartridge.
By adjusting the Stepup transformer resistive loading I expect to reduce the gain a little and fine tune the sound at the same time.
At present my preamp stepped volume control operates at a low 9PM position and the steps being a little coarse, the volume is just a little too high. That is what I meant by toning it down.

Thanks for welcoming me to the "Club".

Regards,

Audie.
 
Hi guys :)
It's good to hear you are enjoying your amps.
Mine are waiting for the trannies :bawling:

Nordic: in your view, what R are worth changing?
Yes I'm a bit lazy :angel:
The truth is I thought about swaping some R on signal path for Riken-Ohm's because I find they made worthy enhancements in listening pleasure when used on DACs.

Sorry Mauro, I am thinking of modding it even before I listened to it :(

Or better, I can mod a second My-ref with Rubicon ZL and Riken:D
and let the first one "stock" to compare.

Cheers
Mauricio
 
I think the two biggest mods you can make are the inout cap and C9 (feedback cap), which is why they have been changed in the kits.

Max, I think your kit has the new MKP input cap, but still has the FC for C9, which does sound great, but a lower ESR is always better. The FM I just substituted has an ESR of 0.030ohms, vs the FC's 0.090ohm. It was the best value I could find from the big suppliers, and actually a couple cents cheaper. I kept the FCs in other places for their ripple handling ability.