[qwerty870]

My preamp uses two Burr Brown OPA132PA opamps. Can I replace them with better ones? They are socketed so instalation should be easy. If so which ones. Thanks in advance.

 

[JMT]

Check out Tangent's Website, he has done a lot of research into various op-amps.

 

[tangent]

Also do a search here and at Headwize. There's no shortage of opinions on op-amps in the archives.

 

[andrzejpw]

Those burr-browns are used in the cmoy pocket amp!

 

[Nezer]

Quote:

Originally posted by andrzejpw Those burr-browns are used in the cmoy pocket amp!

Just about any opamp can be used in the CMOY circuit.

 

[qwerty870]

All I wanted to know was which opamps would be suitable for my monolithic sounds PA-1 preamp which provides 14.5 volts for two mono OPA132PAs. I will probably try a pair of OPA627APs as they have received a lot of praise.

 

[qwerty870]

I actually contacted Monolithic Sounds and they said the OPA627 very well may sound better. I was impressed since most manufacturers dont want you to even open one of their components.

 

[qwerty870]

Does any body have an opinion about the A and B versions of the OPA627 opamps besides tangent? (I already read his)

 

[Nezer]

Quote:

Originally posted by qwerty870 Does any body have an opinion about the A and B versions of the OPA627 opamps besides tangent? (I already read his)

Thjough I havent done an A/B test yet (no oun intended) the B series are supposedly tested for higher tolerances. In otherwords a pair of Bs are guaranteed to match within a very high degree of precission whereas the A seris may not mate as well. Are these differences audible? I would tend to doubt it but I was a disblelieve on differences on ICs.

The only difference between the two is about $15 a pair otherwise and I have decided I'd rather spend the money for the peace of mind.

If I were building several amps I would order the As to save a few bucks and I would be careful as possibly to sonically match (though this may not be necessarry).

If you only need 2 it's only $15 more for the Bs and you shouldn't have any regrets later (should I have gotten the Bs instead, I wonder if they would sound even better?).

For the voltage rails you describe the 627s or 637s would be a better choice. If the gain is over 5 and there is some bandwidth limiting in place (or can easily be put in place) you might consider the 637s over the 627s. It's just more money.

 

[qwerty870]

my preamp provides only 6db of gain. Is this enough for the OPA637s?

 

[Nezer]

Quote:

Originally posted by qwerty870 my preamp provides only 6db of gain. Is this enough for the OPA637s?

If my numbers are correct, and I have every reason to doubt them, you would need a gain of at least 7dB for the 637 to be stable.

Here are my numbers for a G of 5 (the minimum stable gain for the 637):

dB=10 Log(5) = 6.989dB

So, be safe and stick with the 627.

 

[Nezer]

Quote:

Originally posted by Nezer If my numbers are correct, and I have every reason to doubt them, you would need a gain of at least 7dB for the 637 to be stable. Here are my numbers for a G of 5 (the minimum stable gain for the 637): dB=10 Log(5) = 6.989dB So, be safe and stick with the 627.

D'OH! Like I said, I figured I was wrong.

That formula is only valid when the 5 from above is a power ratio. The more generic formula to decribe the gain relation for voltage or current is this:

dB = 20 Log(Vout/Vin)

In this case a G of 5 is a 5::1 ratio of gain so we can say:

dB = 20 Log(5)

which yields a dB of 14 and *well* above your 6dB. For the 637 to work you would need to change the Gain (unless limiting the bandwidth effects the gain stability of the opamp, I dunno what I'm talking about half the time). So the 627, which is unity gain stable (or rather, stable when G=1 or 0dB), would be a better choice for your application where G=2 (or 6dB).

This is making more sense to me all of a sudden.

Now, how does one reversethe equation to calculate G based on dB of gain? I never went very far in my math classes as they bored me.

I can get:
dB/20=Log(Vout/Vin)

But what happens to the Log when moved over?

 

[qwerty870]

Thank you everyone for your replies, especially Nezer. I have just purchased a pair of OPA627BPs based on your recomendations.

 

[tangent]

Quote:

my preamp provides only 6db of gain. Is this enough for the OPA637s?

Um, you're confused. The gain of the upstream source hasn't got anything to do with it. The OPA637 chip itself has to be configured for a gain of 5 or more to be stable. The higher the gain you use, the more stable it will become. Since a gain of 10 is perfectly reasonable for a headphone amp, I would recommend you did that rather than go with the minimum gain you can get away with.

If a gain this high is a problem get the OPA627 instead. It is stable down to unity gain.

 

[tangent]

Quote:

how does one reversethe equation to calculate G based on dB of gain?

The inverse of a logarithm is exponentiation. If:

X = log_subY_(Z)

then:

Z = Y^X

The subY part is the "base" of the logarithm. Since decibels are figured using base-10 logs, the inverse function is:

G = 10 ^ (dB / 20)
[size=xx-small]EDIT: Fixed the equation.[/size]

That is, divide decibels by 20, and then raise 10 to that power to get the gain factor. Any calculator with a base-10 log function will have a 10^x function as well .... and now you know why!