Are High quality opamps important in active crossovers?

[dfdye]

I use OPA4134 (same but quad) and work nice for a 4th order L-R crossover.

How is the cross channel interference in those? Do you have a frame of reerence as to those vs the 2134's? I've never used the 4134's but they do sound really convenient.

Thanks!

 

[jacco vermeulen]

OPA2134 is a very nice op-amp. You want low noise and low distortion

So, what is the minimum noise level of opamps for use in a crossover ?
IIRC, noise level of the OPA2134 is worse than that of the NE5534.

 

[peranders]

So, what is the minimum noise level of opamps for use in a crossover ?
IIRC, noise level of the OPA2134 is worse than that of the NE5534.

... yes.... but how much in dB's? Not very.

 

[AndrewT]

Hi,
it depends a bit on your output level and gain/sensitivity of the poweramp/speakers.

For conventional gain/sensitivity I would guess that 10nV/rootHz is quiet enough you might even get away with 15nV/rootHz particularly if you optimise towards low resistances throughout the filter system.

High sensitivity speakers and/or high gain power amps will require much lower input signals and these noisy opamps even when optimised may be a bit hissy and you may then require <6nV/rootHz.

As a guide 30db amp gain and 86db speakers giving 116dbA/2.8Vac input is my interpretation of conventional.

Someone using 26db+82db=108dbA is therefore low and 34db+90db=124dbA is high. There is a very high category but generally these are valve(tube)/horn based.

I hope this helps towards your decision.

 

[ingrast]

How is the cross channel interference in those? Do you have a frame of reerence as to those vs the 2134's? I've never used the 4134's but they do sound really convenient.

Thanks!

I could not find crosstalk specification, but I guess anything above 40 dB - most certainly easily met by these devices - should be of no consequence. I do not have direct experience in this respect, I use the 2134 for input stages in my amplifiers.

Rodolfo

 

[ingrast]

So, what is the minimum noise level of opamps for use in a crossover ?
IIRC, noise level of the OPA2134 is worse than that of the NE5534.

In my case I have 37 dB overall gain, 89 dB woofer and 91 dB tweeter. 4th order L-R crossover with 4134 (0 dB passband gain) and power amp input 2134.

In this conditions, noise is noticeable only putting your ear directly against the drivers. (and some residual hum I will supress anyway). Half meter away and they are completely quiet.

Rodolfo

 

[peranders]

In this conditions, noise is noticeable only putting your ear directly against the drivers.

Try to disconnect the filters and short the input of the power amp. How much difference in noise?

 

[darkfenriz]

These are Sallen Key filters mostly, they are not very demanding to op-amp. Moreover you may find it nicier to use some other buffer circuits, like source follower of Sziklai pair instead of opamp follower. Some like avoiding class B high feedback opamps wherever possible.

regards

 

[ingrast]

I could not find crosstalk specification, ....
Rodolfo

Update, I found the spec, it is 90 dB min, 130 dB typ for all multiple types (OPA 2134, 4134) ...

Rodolfo

 

[ingrast]

Try to disconnect the filters and short the input of the power amp. How much difference in noise?

Thanks, will check and report. Anyway, I do not find this an issue, at least in my environment which is suburbia and rather quiet.

Rodolfo

 

[ingrast]

Try to disconnect the filters and short the input of the power amp. How much difference in noise?

Well, tried it and noted no difference. In fact most of what I could hear was mains related ant then only placing the ear against the drivers. A hint of hiss could be perceived in the tweeter.

Yet just stepping back about half meter all noise vanishes. I took care to close doors and turn off computers, radio of course, and observe only when there was no street traffic (about 100 m away and sparse).

As said, system gain is 37 dB (for the woofer channel, the tweeter has 2 dB more conversion gain and being 4 ohms against 8 for the woofer, the corresponding amplifier channel was adjusted for 29 dB voltage gain).

4134 crossover is up front and 0 dB gain. I may probably change this anyway, for it does not make much sense to process crossover at a nominal - 10 dBV level, may be I lower the amplifier gains say to 27 dB (for the woofer) and put 10 dB in front of the crossover.

Rodolfo