My "audiophile" LM3886 approach

Just a thought...

You might want to include schema, and parts list and pictures on your site. The reference to the diyaudio thread clearly works but requires the customer to parse through the iterations and modifications of your project ~ 1000 posts.

Just my opinion...

I might be a "buyer"!
 
BrianDonegan said:
They are coming... time, time, time.

There is a parts list now, though I seem to have blown away the link to it (yikes!). Putting it back. We will be posting pics of the boards and kit contents as well.


I'm just wondering if this group buy is a one time thing or do you and Russ plan to keep a supply of boards and kits on hand for the foreseeable future? So many niffy projects, so much to learn, so little time and money. ;)

regards

Dave
 
My audiophile approach to (My "audiophile" LM3886 approach)

I have just completed by fourth Ref C and this one is a killer.

Well there have been speculation as to some of these mods that i am going to list here.

1- dual transformers with dual secondaries (220VA). each transformer supplies only one polarity rail one for negative one for possitive ( you need dual rectifiers to do this mod
2- i have a seperate power supply per op-amp. at 18V 7818/7918 with seperate transformers

gone is the noisy zeners. and the hot resistors.

sound
the tops are much cleaner and much more balanced.
mids (never liked the mids on this amp to much - compared to my digital amp) very clean and very detailed yet smooth - on a par with my digital amp (tripath eval board RB-tk2350 - also modified)
Bass cleaned up as well. more definition also beter balanced less of a muddyness (not that it to muddy to start of with)

Regards
Rudi
 

Attachments

  • 3886-318.jpg
    3886-318.jpg
    82.9 KB · Views: 1,339
Re: My audiophile approach to (My "audiophile" LM3886 approach)

rudi said:
I have just completed by fourth Ref C and this one is a killer.

Well there have been speculation as to some of these mods that i am going to list here.

1- dual transformers with dual secondaries (220VA). each transformer supplies only one polarity rail one for negative one for possitive ( you need dual rectifiers to do this mod
2- i have a seperate power supply per op-amp. at 18V 7818/7918 with seperate transformers

gone is the noisy zeners. and the hot resistors.

sound
the tops are much cleaner and much more balanced.
mids (never liked the mids on this amp to much - compared to my digital amp) very clean and very detailed yet smooth - on a par with my digital amp (tripath eval board RB-tk2350 - also modified)
Bass cleaned up as well. more definition also beter balanced less of a muddyness (not that it to muddy to start of with)

Regards
Rudi




Looks great, that is a lot of iron. One question, the 18 volts for the LM318, was that a random selection?

George
 
Hi Brian

the PS for the opamps are individual units (one per opamp). so i basic removed the resistors and zeners and wired the PS to the 100uf caps onboard. so you have 3300uf - 100nf - 78xx/79xx 100nf - wire to board -100uf - 100nf

regarding giving the opamp a higher voltage from the zener, i have not tried it. did not like the idea to start with. rather regulate it properly and without any option of x talk on the supplies

cheers
Rudi
 
rudi said:
Hi Brian

the PS for the opamps are individual units (one per opamp). so i basic removed the resistors and zeners and wired the PS to the 100uf caps onboard. so you have 3300uf - 100nf - 78xx/79xx 100nf - wire to board -100uf - 100nf

regarding giving the opamp a higher voltage from the zener, i have not tried it. did not like the idea to start with. rather regulate it properly and without any option of x talk on the supplies

cheers
Rudi


Hi Rudy, I too also made an amp almost exactly as you describe, but it is such a drastic change I did not feel comfortable calling it a "Mauro MyRef" anymore. It is now a Russ "MyREF". :) But I would tend to agree that all of the changes are good.

One thing to note though, I have noticed on the scope, that the 78xx/79xx are not really much better noise wise than the zeners. I wonder if the problem we both solved is more noise from your power resistors than the zener. The biggest problem with the zener seems to be the brute force inefficiency.

I ended up using noise compensated 317/337 pair for my design as noise was significantly less on the rails than with the 78xx/79xx. In the end though I think my design was overkill.

Intersting your comment on the mids, I have never heard any lack there even with my original REV A, so I am very surprised. That could all be about component selection, or some other factor.

Good work. :)

If there is enough interest I would be glad to show the "RussRef" but I think that is for another thread. :)

Cheers!
Russ
 
I'm just wondering if this group buy is a one time thing or do you and Russ plan to keep a supply of boards and kits on hand for the foreseeable future? So many niffy projects, so much to learn, so little time and money.

Hi-

The reason we started the web site is basically that group buys can be a real pain to manage, and people always get left out. We hope to keep these boards and kits available on a continuing basis (unless we run out of money ;) ). It's really just a lot easier to deal with this way: no Wiki to maintain, no flood of emails to keep organized, etc. So, take your time.
 
Hi Russ

I would like to see what you have done. ;)

I think that i pick up the difference in the mids is the fact that my mid driver is actually a ribbon and not a cone. cones are very slow and "hides" these changes.

but for the rest. I just wanted to build a all out amp to compare with my digital. (this one is on a par)

the one thing about my mod that does not work with your boards are the fact that i have dual rectifiers. and for that reason i can run a transformer per polarity rail. one transformer for possitive and one transformer for negative. this made big difference aswell compared to the one bridge per amp.

cheers
Rudi
 
Mods on Rudis Rev C

I have had an opportunity to listen to the Rev C Rudi made at various stages. In my mind and to my ears, the most recent changes have made a dramatic impact across the frequency spectrum.

The changes were:
1. 318 fed from 78xx/79xx rather than the noisy zeners. These were chosen above 318/337 since the latter tend to be wider band, allowing greater potential for noise contamination.
2. 318 operating voltage increased to 18V-Op-amps prefer to run at higher voltages in general.
3. Seperate transformers for the 318's to reduce power supply noise further.
4. Seperate 200VA transformers for the 3886, one feeding +, the other -

Vs the old, single transformer configuration, bass is now MUCH better controlled, especially the very low frequency stuff. Highs and mids have become much more transparent.
One of the charming characteristics of teh amp is it's mid presence, which has improved with these mods.I must also say that highs are not sharp/bright/metallic at all, although this has tended to be a characteristic of this amp from the start.

Vs Rudi's existing 2350 digital class T amp, I prefer the REVC, and that is a bold statement, since up until these mods, I was pretty convinced that the Class T was the way to go.

Do yourselves a favour and implement these changes, they are VERY worthwhile.
 
rudi said:
Hi Russ

I would like to see what you have done. ;)

the one thing about my mod that does not work with your boards are the fact that i have dual rectifiers. and for that reason i can run a transformer per polarity rail. one transformer for possitive and one transformer for negative. this made big difference aswell compared to the one bridge per amp.

cheers
Rudi

You are correct of course. :)

One thing to remember about Mauro's design (not so much a factor in mine) is that has certain practical design goals which manifest themselves in low component count.

On my design I actually left the PS as a seperate PCB simply because I was tweaking and changing it all the time. :)

Cheers!
Russ
 
Re: schematic...

UnixMan said:
Hi Rudi,

could you please post a drawing of the schematic of your complete setup?

That would made things a bit more clear and avoid possible misunderstandings, I think. Thanks a lot.


Hi Paulo

unfortuneately i do not have software that can do a nice schematic for you. but I did a quick hand drawing of how the main trofos get hooked up.

for the op amp, i used two standard power supplies and just removed the zener and resistor

the power supplies are then wired to the 100uf caps

cheers
Rudi
 

Attachments

  • trafos.gif
    trafos.gif
    12.4 KB · Views: 1,198