In Active crossover application, OPA2134 and OPA2604, which is better?

[visualwang]

I found that these two ICs had similar parameter, but some difference:

OPA2134: lower noise(8nv, 2604 is 10nv), lower distortion(0.00008%, 2604 is 0.0003%);

OPA2604: high slew rate(25v/us, 2134 is 20V/us), high bandwidth(20M, 2134 is 8M)

In Active Xover, there are input buffer, filter, output with gain changing. Input buffer and filter are all unit gain.

Then, 2134 and 2604, which is better? Or have other ICs be more fitable in this situation?

Thanks in advance!

 

[darkfenriz]

If you are going Sallen-Key filter, that's not much difference, because all op-amps work as followers. I would choose opa2134, because 2604 is known to have much higher THD in reality than its spcs suggest.
Adam

 

[PierreG]

What schematic are you using?

 

[visualwang]

Thanks.

Yes, I use Sallen-key architecture and 4rd L-R filter. Its a two-way system with input buffer and gain ajustable ouput buffer.

I know there is little difference between these two OP, but I want to know which is better.

Actually, 2604 is a little expensive than 2134. If 2134 is better here, I would like select 2134.

 

[ingrast]

T....
Actually, 2604 is a little expensive than 2134. If 2134 is better here, I would like select 2134.

May want to check LM4562 ....
0.00003% THD
2.7 nV
20V/uS
55MHz

Just received 3 free samples today.

Rodolfo

 

[Nelson Pass]

When we were developing the XVR1, I prototyped the
circuit with 2604's, and it did not sound good. I attributed
it to instability in the op amp when used near unity gain.

 

[georgehifi]

I 've tried them all at least half a dozen of the so called high end opamps in this circuit, and the OPA4134 was the best sounding, but still not good enough for me for the mids and tops just for the bass, the mids and tops were better when passivly done still. And as Nelson stated when trying different ones be sure they are unity gain stable before getting them.

Cheers George

 

[mlloyd1]

i think the opa2604 is (at least designed to be) unity gain stable - no minimum gain requirement, such as the OPA637 has.

i think Nelson was pointing to something else in the design/operation of OPA2604?

mlloyd1

 

[clem_o]

In a test that I did with my KEF Kube (a low-frequency active equalizer for older KEF loudspeakers) the 2134s sounded more neutral and less strained than the 2604s. The original ICs are NE5532s BTW.

Cheers!

 

[lineup]

OPA2134 and OPA2604
are very close in data.
It is difficult to say which one is better.
I would say it depends on what circuit.

OPA2604 is a little bit faster. +Gain Bandwidth.
OPA2604 is one of the few op-amps having +24 -24 max supply voltage.

OPA2134 is more stable, as it is not too fast for Audio.
OPA2134 is more used and more liked, than almost any other JFET op
( Maybe still TL072 is most used, but it is far from OPA2134 in quality )

-----

OPA2134 is probably the most used DIY Audio op, all categories.
It is has good qualities and very all-round
so almost never any problems, whatever type of circuit or gain.
Has very good current drive capability for being an OP-amp.
Close to a perfect OP for general Audio designs.

Unless you need that high +24V -24V supply
I would say you buy a supply of 10-20 OPA2134.
They can be good to have at home,
whenever you want to setup some circuit.


lineup

 

[lt cdr data]

tl072 and lf353 are the industry std. ones using in some very good actives

 

[georgehifi]

tl072 and lf353 are the industry std. ones using in some very good actives

Yeah and NE5532's are the industrie standard for output opamps and we all know they sound like ----, so do the TL072's and the LF353's.. It's because they can buy them in bulk for a couple of cents and they do the job, but only just.

My two cents worth

 

[Nelson Pass]

i think the opa2604 is (at least designed to be) unity gain stable - no minimum gain requirement, such as the OPA637 has.

i think Nelson was pointing to something else in the design/operation of OPA2604?

The circuit didn't oscillate - it sounded bad.

This is a topic worthy of a thread all by itself, as it applies to
the sound of op amps, chip amps, and other circuits with
high open loop gains.

It is my observation that as the phase margin gets thin, the
sound starts to suffer, and while many of these circuits don't
show actual oscillation, the approach to that boundary puts
the behaviour into fractal territory.

It's easy to experiment with, as pointed out in the first post
of my thread "7 easy pieces"

The example schematic was:

www.passlabs.com/np/GC-VAR-INV-FDBK-1a.pdf

 

[tiefbassuebertr]

The circuit didn't oscillate - it sounded bad.

This is a topic worthy of a thread all by itself, as it applies to
the sound of op amps, chip amps, and other circuits with
high open loop gains.

It is my observation that as the phase margin gets thin, the
sound starts to suffer, and while many of these circuits don't
show actual oscillation, the approach to that boundary puts
the behaviour into fractal territory.

It's easy to experiment with, as pointed out in the first post
of my thread "7 easy pieces"

The example schematic was:

404 Not Found

possibly there are additional reasons for the unsatisfactory results in sound quality.
Therefore check out my comments in post #66 under
Is the NE5532N Dual Opamp any good?
and in post # 375 under
The Best Sounding DUAL opamps


concerning the internal circuit topology most typicall audio op-amps like NE5534/NE5532, AD744/AD746 or OPA627/OPA637 uses two voltage gain stage in the NFB-loop (LTP-VAS-Buffer) According a lot of listening tests I always note crispy and harsh sound character, that this is one but an important reason for this impressions.
Best sound I get in most cases on OP AMP's with only one voltage gain stage in the NFB loop. Examples for such OP-AMP's are the good known AD797 (still no dual version available), AD829/AD826, THS4031/THS4032,
ADA4898-1/ADA4898-2, AD8021/AD8022 so as the dual versions LM6172, LT1358 and LT1469
Thus I don't like more than a single gain stage in the NFB loop.
The OPA604/2604 and OPA134/2134 uses also only gain stage, and sonic quality is sometimes good and sometimes not, maybe through different factories and different dates of manufacturing.
according this thread under
LT1358 or LME49860 ??
the LT1358 is a good choice.

for "7 easy pieces" check out first pdf file in post #90 under
7 Easy Pieces