|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
71.83.42.181
In Reply to: RE: OPA2134 vs OPA2604 vs OPA2107 posted by KanedaK on December 22, 2016 at 16:38:35
I got best results in my active xovers by mixing 2134 and 4562. The 4562 can make some hiss and get a bit too detailed but has nice air. The 2134 is very quiet but pretty dull and closed in. Don't know anything about the others you mentioned. I am now in the process of converting to OPA627, a pretty expensive process, but I hope one that will result in the highest fidelity.
Peace,
Tom E
berate is 8 and benign is 9
Follow Ups:
LM4562 in my revox player took more than 100 hours to sound good.
Before that, it was as sharp and cold as a swiss knife.
OPA2134 seems inoffensive and defined enough but it's got this flat, slightly recessed in the midrange presentation - not quite what I'd call "dull" but not bright, just a bit "hifi", it sounds like an entry-level Japanese solid-state amplifier, and there's not much openess and width to speak of. I've heard my system imaging from wall to wall so this sounds a bit uninspiring. everything stays between the speakers.
Say I'd mix LM4562 and OPA2134. I guess I have to try it myself but, what do you think would work best? LM4562 as input buffer and OPA2134 at the outputs? or the opposite? Or, maybe, OPA2134 on the lowpass outputs and LM4562 on the highpass?
I have this thought that the signal coming out of the lowpass should go trough the same active devices as the signal coming from the highpass to keep coherency at its max, but somehow I feel that's just overthinking and it's not going the feed the same amplifiers anyways so... maybe something to try.
Call it what you will, but you get the idea that the 2134 leaves something out of the signal, while the 4562 adds something to it. I see using them together as an opportunity to extract the best balance from both. Coherence between the bands is determined by filter configuration, not parts. Coherence within the bands is the result of which parts are used. I used 4562 for input buffer and high pass, 2134 for low pass. You'll need to experiment. Do you use sockets or solder direct? You might consider sockets for experimentation, then direct solder the final configuration.
Of course, everything needs playing time to settle. Not sure it's a hundred hours (seems to be a number regularly stated but seldom confirmed), but mine played for months.
You must be aware that the passive parts produce sonic effects as well as the opamps. I have found filter component caps and resistors to contribute, especially in the high pass section. Hint: Wima caps, although good, are not the ultimate solution, primarily because they have magnetic leads, as do most cheap metal film resistors. Even the power supply caps can make a difference. Ultimate fidelity is a complex problem, with no simple solutions. Fiddling with opamps is only a start, but it's better than doing nothing. Don't neglect power supply bypassing and decoupling of faster opamps. You can add a .01uf cap right across pins 4 & 8 on the bottom of the board.
Peace,
Tom E
berate is 8 and benign is 9
Thanks for your answer.
Right now I'm gonna let the opa2134 burn-in a little bit more but I know myself, I won't resist long the temptation to order a few LM4562.
I was thinking the same: 4562 as buffer and on the high-pass.
I'm using gold plated, machined sockets - I'm not giving up on using SS3602 from Sparkos labs one day and this is all a "temporary" solution. But I know I might well find the perfect balance with regular opamps and leave it as it is forever. I've stuffed a 0,1uF MLCC capacitor under each socket between pins 4 and 8.
I've replaced the coupling caps with back-to-back Silmics II (-++-) for now but I might add some bypass or squeeze some bigger film caps instead but that can wait. Back to back Silmics sound surprisingly good.
I've replaced the PSU caps with Nichicon muse KZ. One would argue that you don't need "audio" caps in that position but it sounds better than before so I'm happy with it.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: