My "audiophile" LM3886 approach

interesting powering scheme

Rudi,

What is the type of thinking behind the scheme. What is the advantage expected from crossing between the two trannies?

Did I understand it correctly that you did a dual supply scheme for the two channels? I would expect that you introduce crossoning effects between the two channels again?

Thanks for sharing your thoughts

Regards
earsandeyes
 
Rudi,

I appreciate very much Your findings, and I have noted from the beginning Your true dedication to this nice design. Also the fact that you are producing your own boards for it, but in accordance with the original design. Bravo! And what you find is interesting without doubt.

But, please, don't repeat any more "those noisy zeners"... Please, go and look up carefully the findings of Christer in this regard, and You will see that actually the lowest possible noise reference [yes, lower even than green / red leds] is a 12V zener, measured & confirmed many times by Christer.. The noise of an 7818 with 100uF output cap is several order higher. Also, the dynamic resistance of the 7805 is lower, so can not even filter it out afterwards.

* If you can get the desired voltage drop with one single
diode or LED (excluding blue LEDs) then it seems usually
best to use that single device. Special requirements for
the current may affect the choice, though. For higher voltages
up to or just below 12V, string together a suitable number
of red, yellow or green LEDs. From 12V upwards, use a
zener diode (or several if required).

Read also the pdf here

So, You can well have better results by changing something in such a complex configuration, but surely for other reasons than the noise, if you have better sound out with much worse noise on the supply.

The other thing which is not logical for me, that why do you not separate totally the power supply with two transformers? Also here, what you did is including one more element, the diodes, in the common path for two channels, which was separated before..?
And also the transformers remained in the common path?

I would understand this, if it would be a necessity, because you have two transformers with half AC voltage. But then it's not a tweak, it's a local patch?
Or did you have special reason for doing that?
Otherways, a much interesting report, thanks!

Ciao, George
 
hi earsandeyes

well first of all you are seperating everything. you do not have X talk between positive and negative.

you do not have interactive forces within the transformers seeing that they only supply one specific rail each

the dual bridges seperates it even more.

but i can be completely wrong, i am just speculating and there is alot of different opinions out there. I am only sharing my experiences and what i found to sound best.

cheers
Rudi
 
Joseph K said:
Rudi,

Pardon me, it was not up yet your scheme when I started to reply.
So what I said about diodes is not valid. Do you have any idea why it can be better this way than just two transformers?

Ciao, George

Hi George

other than the fact that you have a no interaction between the two transformers, i have no idea. if you have a trafo per rail the sound is less open.

it is the same thing that NAIM had with there power supplies. they run the one secondary out of phase with the other and they get better bass this way. i have tried it and it works. but nobody can explain why


cheers
Rudi
 
LM318 Power Options

I had a brainstorm this weekend. Read a review of the 27,000.00 ASR Emitter II. It also uses an opamp input.
The opamp is powered by batteries. This I can do, still have a bunch of 12v SLA from my experiments with Tripath amps and NOS dacs. Since they did not pan out, the batteries are sitting unused.
The performance may not improve. But in my system the zeners and dropping resistors generate more heat than the LM3886. After being on for two hours the amp is the same temp idling as it is playing. My output power demand is minimal.
Doubt if I will ever get to hear the ASR, but look forward to trying the My_Ref with a pair of 13.2 volt batteries. They have to be as low noise as the bypassed zeners. I have a hunk burnt out of a knife blade testifing to the current delivery. Vaporized the stainless steel.

George
 
doggy said:
power supply question:

Rudi: do you think that 2 center tapped trafos could be used with the same effect as trafos with dual secondaries?I am still learning. how could the center taps be wired?

cheers;
doggy:snoopy:


Short answer, no not really, not unless you did some trafo mods(which would not be too hard) and made your center tapped trao a dual secondary trafo. :)

Honest answer though, try it as it is first, I doubt you feel the need for the mod, but if you do, at least you will have a reference, but to be honest unless you can do a direct A/B test it will be tough to guage the change.


Remeber that the trafo quality itself can color a test quite a lot. Also Toroids will give you different results than other types of trafos. For example with all the speakers I have tested and a singfle 300VA toroid per channel I have crystal clear mids and I have never heard tighter bass(even down lowwwwww) ;). Now remember that is with a PCB that is not the same as Rudi's.

I actually fully trust Rudi's review and he has been very straightforward and reliable in the past, I just don't think he is listening to the same amp I am because mids and bass are very much strong points of the REV_C monoblocs I am listening to. Now Rudi has some very cool speakers, mine may indeed not be as sensitive as his, but they are no slackers. :D

I would like to see what the other beta builders think as well, but I dont see how any mod could get you significantly better performance (I have tried). Sure there are certainly small gains to be had here or there at the cost of more components and greater expense, but there is always a point of diminishing returns.

Remember that we hobbiests tend to color our opinions a lot by the amount of work we put into something. I certyainly find that to be true. It is human nature. It is very tough to be totally objective. Some would say impossible. :)

So please don't take my word for it (or anyone's really), my opinion is just that. Find out what yours is, but it is good to start with some baseline, and then compare to things at the same time. A/B tests(particularly blind) can be very revealing. :)

In any case all this discussion is excellent, and great food for thought. If you read a lot earlier in the thread you will see Mauro's characterization of his choices for PSU components, agree with them or not, they have proved very sound and they are based on very sound engineering. I think this has really been demonstrated by scores of happy builders. :) You can take the freely available PDF and etch it and have yourself one heck of an amp.

I was actually very very skeptical of the zener regulators, but I find that I was dead wrong now that I have tried 78xx/79xx and 317/337 combos I can say that mauro hit the sweet spot of simplicity with excellent performance, sure they used less current, but the audio performance was not improved substantially(some listeners thought it worse some better) so I was left scratching my head. I will try the seperate trafo mod as it is easy enough, but would I expect a huge improvemnt... no. Some... Yes! Does this make it worth doing sure!!! To some people like Rudi and I even little incremental improvemnts can be very rewarding, even if only a select few ever notice them. ;)

Nothing will ever keep me from experimenting :). Thats what DIY is all about. So those of you who are doing these kinda of mods, I love it! It helps keep things alive and interesting. :D We all just have to keep in mind that your milage may vary. ;)

So cheers to you all and happy modding. :)
Russ
 
Hi Russ

just a confirmation from my side..

the first amp i build were 100% rev a except for the bridge rectifiers

second one. Rev A with ref C mods. also dual rectifiers.

third Ref C but my layout. i found this a touch better than the previous two. but not that much

then i build this one with all its modifications and it was a completely different amp.

as a test i always give the amp to Ryan "Dr H" to test. just to get his opinions. he was looking at upgrading his 3865 combo with all the snubbers etc.

he also had exactly the same findings as myself so far. he was about to go for the Tripath tk2350 till he heard this incarnation.

the difference on this amp is not just a little bit it is a HUGE improvement. so much so that i have to use this instead of my digital. little improvements is equal to going from a E88CC-SQ valve to a CCa valve on my DAC output. huge improvements is some books. small in mine


i just use a bigger input cap than you , a Wima MKP10 160v the rest is the ordinary components, nothing fancy nothing special. and always the same


why not build a amp based on my arrangement and test it, ;) you would be happy you did. and remember that you need to use 18v for the opamp

cheers
Rudi
 
Russ White said:
Huge difgference? Wow I hope so, if so I think I will shoot straight to self actuallization in a euphoric climax of pure audio bliss.... I wait a second, I am already there! :D Haha JK.


I know what you mean.

well it was a eye opener for me a Ryan. try the bigger cap aswell. what you you using now.

what we found with the previous versions were that they did have a good mid. but not to the liking of both if us. we were very used to the digital that had a valve type sound and with that i do not mean woolly and warm ( that is a very bad valve) but without any grain and lots dimension

i am sure the setup with be even better with toroidals instead of the cheap e-core i was using (test bed trafo's) actually this is not my final product it stull have to be build into a nice case

cheers
Rudi
 
this is what i am using to compare it with

left - Tripath RB-tk2350
middle - mosfet amp 1000va per rail ( crazy I know )
right - JLH 1996 version
 

Attachments

  • amps.jpg
    amps.jpg
    55.8 KB · Views: 1,642
currently using 2.2uf KimberCap 200V :)

Nice amps. :up:

Just a few minutes and I should have a working "Rudi" channel, well as close as I can easily approximate today. Everything is essentially the same as yours but the channel is on its own PCB and its own heatsink and the byapassing is like my monoblocs. Also I am using low ESR caps(some nichicon some panasonic FC) for all the bypassing electrolitics.

The NFB cap I will try both bypassed, and solo.

Cheers!
 
mikelm said:
Hi Rudi

How does it now compare to the JLH ?

Does your JLH have DC feedback ?

mike

Hi Mike

the JLH is great amp but unfortuneately can't modify them any more.

the new one is Much better than the JLH (hate to say that - seeing that i loved these amps since i was 12) :D and it was my benchmark for a long time

I do not know of a JLH without DC feedback :xeye:

cheers
Rudi
 
rudi said:
hi Mike

I see what you are getting at. Mine uses a cap but to get same results as just a resistor i used two caps to make a non polar cap and the bypassed it with a 100nf. this way you get the same effect as removing it but without the DC offset problem

cheers
rudi

Mmm.... are you sure about that ?

I will do all of the mods you suggest and compare it to my balanced JLH with DC feedback and report my findings.

mike
 
Class A and LM318

Since the LM318 appears to be a standard op-amp and responds well to the usual op-amp tlc (low noise power supplies and good decoupling), has anyone here tried the Class A bias trick?

Using a 3-5k resistor from output to V- is a simple enough solution, but it may sound even better by gioing the JFET route.

Also, with such poor PSRR (-60dB at 10k), perhaps the Goudreau triplets (search usenet for the name and DAC-1) are worth pursuing as well?

Ryan