The best sounding audio integrated opamps

Andrea, please!! they're not ready for the whole op-amps numerology/colors, just yet :p

what does LT1124ACN8 say btw? coz I'll prolly get it soldered together w/ the burson anytime soon.
Hmm, not the most interesting of numbers... the most "musical" sounding of them being the 3 and the 5, for my ears. At any rate, there's also the sum of the numbers to consider - it'll give the overall "nature" of the opamp (or other). For instance...

The LT1028 = 11, which is a key number in numerology; meaning sort of the "reunion of the opposites". No wonder it's that good?

Also, take the LME49723 = 7 [you sum until you get a prime number], which is another very meaningful number... no wonder this one sounds better than the LM4562 and the rest?

And, LT1124 = 8, that is something like "fame & good fortune" :) but not perfection...

and LT1115 = 8 as well


Note that I'm not an expert in numerology by any means... but I've consistently found proof of what I'm saying (like it or not, you scientific minds).


Ah, and strangely I'm not surprised that the LT1115 is better known and/or more considered on audio forums than the LT1028 :)
 
Last edited:
BTW, burson told me "There are two LM7809 and LM7909 on the Prodigy HD2 soundcard, you can upgrade them to LM7812 & 7912. You may worry the drop off voltage of the power supply but from my own experiment it has been prevent to be very effective and our opamp couple perform even better with this mod."

but I've spoken to the EE who will solder them for me, and he's said that these parts need +4V headroom to regulate...so 12V in>12V out is not the best theoritical idea in his opinion, better get regulated 8.8V than unregulated 12V from the PC PSU(I've carefully chosen a PSU w/ the lowest ripple ever, though)

and he said that I need 0.1uf caps w/ the lowest value(the 630V Mundorfs MKP cap the original owner put is WAY overkill): http://www.bursonaudio.com/opamp101.htm
Please solder a good quality foil capacitor (MKP caps) across the power intake pins (V+ and V-).

how about the best mundorfs money can buy w/ a low voltage :p
 
Last edited:
Last edited:
Administrator
Joined 2007
Paid Member
Fairchild quote 2 volts as the headroom required but that has to include ripple and spike etc. In other words if the input goes to 13.9 you may be in trouble. That 2 volts is the same for pos and neg regs and at the quoted current (1 amp)

Just measure it and see what you have to play with.
 
Administrator
Joined 2007
Paid Member
Surface mount caps are better as there is no lead inductance that negates their effectiveness at high frequencies. Mount them between the legs of the reg on the pcb. And being smaller they are less prone to pick up radiated noise etc....

You do know with non polarised caps of foil construction there is a "right" way round to connect them to take advantage of the outer layer of foil for shielding. Connect the outer to the lowest impedance or ground end :)
 
Administrator
Joined 2007
Paid Member
12.3 is no good. You can raise the output of any 78/79 reg by adding a zener into the ground lead to lift the output. So a 7805 and 6 volt zener makes an 11 volt reg. There are better ways still using a small transistor and two resistor to "lift" the ground pin of the reg but the zener approach is fine.
 
oh ok! thanks for the expert advices :)

so should I ditch the mundorf MKP 630V/0.1uf that I have(it's huuuuge), and put 3 WIMA caps or sumthing? got any reference in mind please? I appear to really need 0.1uf

so what would be the wisest to replace the stock LM7809/LM7909 in your opinion? some 7805/7905 + a 6V zener?

you kinda lost me on "Connect the outer to the lowest impedance or ground end" but I'll tell the EE who will install them for me :)
 
Administrator
Joined 2007
Paid Member
This is the sort of cap I use,
--|U0805F104ZCT|CAPACITOR, 0805 0.1UF 50V 10PK | CPC

As to the regs, with only 12 volts input to play with you are better staying as you are with the 7809 etc.

You could :) always use a DC DC convertor and regs to get whatever you wanted, -/+ 18 is easily achieved but you have to remember the convertr has to implemented correctly so as not to contaminate signal grounds etc.
Such as these,
MURATA POWER SOLUTIONS|NMJ1215SC|CONVERTER, DC/DC, SIL, 1W, +/ | CPC

TBH I would stay as you are.
 
Administrator
Joined 2007
Paid Member
Oh yes... the caps. If you imagine rolling a piece of paper into a tube you can connect a wire at the outer wrap or the innermost. A capacitor is two such "sheets" rolled together. So the outer wrap connection if it is the "ground" or low impedance point in the circuit can be used to help shield the "inner" connection to the other sheet. Poly caps (small signal ones) have a red band on them and that is the outer. In small signal circuitry it can make a big difference to noise pickup etc.
 
alright, that's also what he advised me :)

I was kinda hoping for "audiophile" grade caps tbh, the mundorf sounds amazing :p

how can I fit these tiny caps on the back of a DIP8 socket exactly?

I haven't seen any way to differentiate the 2 ends of the mundorf cap, I'll look better :)
 
Last edited:
Administrator
Joined 2007
Paid Member
how can I fit these tiny caps on the back of a DIP8 socket exactly?

You can't. Next question :)

They are ideal for rail decoupling, adding between the legs of standard 5mm pitch electroylitics etc.
For decoupling an opamp you want something like,
--|MCRR25104Z5UM0050|CAPACITOR, 100NF 50V | CPC
soldered direct to the supply pins on the IC. You can do that neatly and it won't stop you using a socket. In fact I sometimes (depending on application) do that and have a small electroylitic as well, say 4.7uF

What you have to remember is that so called "super" components cause as many problems as they solve in some applications. Size is an issue. Decoupling a slightly noisy rail with a component with long leads will just negate it's effect. You get better reults with a normal commercial grade component deployed correctly. Add a centimeter of PCB print too and that then suddenly becomes the dominant feature. All this is easy to measure and prove to yourself with a scope. You can actually slide a scope probe up and down a long lead to a cap and see the ripple get bigger as you get further from the cap.
 
very nice, thanks man! I'll carefully examine the mundorf for any marking on either end..so it's connected between V- and V+...I take it the outside foil should go to V+?

oh ok, that's the same kind of cap they use on the burson itself...might indeed be a better choice, Andrea also told me that long leads were ruining any benefit of using the cap in the first place :eek:

discrete op-amp manufacturer claim that they use highly matched components...I'd need to find the same sort of cap but milspec if any possible? and this is CFC, Burson advised MKP...not a problem?

http://shop.ebay.com/i.html?_trkpar...&_sticky=1&_trksid=p3286.c0.m14&_sop=15&_sc=1

this one then? http://cgi.ebay.com/100N-250V-Metallised-polyester-capacitors-x-10_W0QQitemZ270436973887
 
Last edited: