The OPA637AP isn't well suited for this circuit and coupled with the class A biasing I'm sure it runs quite hot.

I don't know that it is recommended by the manufacturer, but it is indeed sold by a reseller who has a good rep (though his OPA627s are suspiciously cheap).  OPA637 is a fast opamp and can have problems with oscillation, especially at unity gain.  I haven't used it in this application so I can't offer any more than that.  There are a number of cheaper, more forgiving options out there though.  OPA602BP is my favorite at the moment.

Quote:

Originally Posted by AllsWell 

Well, could you maybe elaborate to what IS well suited ? I bought it according to the manufacturers ( of the amp ) recommendations . And it sounds great, but what would your (nullpo) reccomendations be ? (Just of out plain curiosity from that one liner you gave us, (mystery) .

 

**Not trying to be hostile fyI**

Class A really runs hot, just look up Schiit's write-up on the Asgard since they anticipate that many buyers will be younger people who probably started with iPods. We have old Rockford Fosgate Power series amps in cars before and the running joke was we could leave the BBQ equipment at home when we tailgate, because you can at teh very least cook an egg on those after an hour with no airconditioning in the cabin (worse in the trunk); at some point most of the people using them had SUVs or minivans and they'd mount the amps anywhere the rear vents could reach them. Fullrange Class D amps started appearing a few years ago and while they were 'colder' than A/B's (in both sound signature and operating temperature), even some biased more to B than A, these eventually took over since the A/B 'warmth' is usually just harmonics. Car audio competitions are more objective more than subjective, and they needed amps with a flatter response. The Musical Fidelity A1 driving speakers with less than 90db of sensitivity would run just as hot; the Pass Labs 10- to 40watt per channel pure Class A amps are the size of 200watt per channel Class A/B's, but most of this bulk is just the chassis heatsinks and the huge torroidal power supply. Either way, compare those to NuForce or Bel Canto fullrange Class D amps, which pump out over 150watts per channel and come in chassis that are the same size as a slim DVD player or midi-size CDPlayer.

That said, op amps do have limitations, especially since they're designed primarily for A/B operation, not pure Class A or for much greater Class A bias. Some amps come biased into Class A already from the manufacturer and that should attest to them knowing what the limits are, not to mention that they probably designed the chassis to help with cooling. Best would be to look up the data sheet of the op amp and measure the temp, if it's not there shoot the manufacturer an email. No need to panic waiting to confirm that though, because like I said above, running considerably warmer (as long as the chassis isn't too hot to touch) is normal for Class A.

I used the 627/637sm before and they heated up pretty badly. Now I'm using the LME49710HA, which IMO sounds better. But I used all of them for a long time and can say that they will work fine and won't burn anything

What resistor value is being used to bias the opamp?

3.2KΩ - too hot
3.9KΩ (3.85mA) - seems to be a good safe choice.
4.3KΩ - cooler, but may not fully bias.

http://www.head-fi.org/t/475618/matrix-m-stage-amp-review-simple-cheap-and-excellent/1065#post_6988664

Quote:

Originally Posted by AllsWell 

 

This is what is say's on Tams website about the product I got;

 

"3.9K is the recommended value for the resistors. You can also choose other values if you know how it works in this mod. Otherwise, just go with the default."

 

So, I'm guessing it's 3.9  ..... which is good by your explanation there .

 

Guys, im no tech guy so I barely have an Idea of what anyone's talking about (Especially ProtegeManiac's post).... I'm more of a intermediate level "tech" guy with some attention issues... lol . anyways.

 

I leave the specific math to engineers and lay out the concepts. I'll go into a little history lesson on topologies: Class A keeps the circuit "open"; any changes in the signal and it's ready to amplify that. Problem is even with no signal coming in it's still "open" - that turns into heat. Class A/B was meant to manage this, and Class D is basically a very, very fast way of switching "open" and "closed," so it stays cool if left running, and much cooler than A or A/B when running. The ygive you the numbers and values but you'd have to get the concept of what biasing it to Class A actually does.

Think of them in terms of anime, but take not this is not all the aspects of these amps, but just to give you a general idea of the heat generation:
Class A - guy makes a fireball in his hands while teammates engage the enemy; if the opportunity to use it comes, he can just let it rip. However, the longer he doesn't use it, he gets weaker and sweats a lot
Class A/B - he only charges his fireball when the opportunity comes; however the result is probably that the enemy will see it coming and block it.
Class D - Ninja/expert katana or super demon class. It's like standing in attention, or even attending to something else, then whipping out a blade, even a katana (that's why it's curved by the way) as a super-fast reaction and blocking the attack; or as a super-demon, like a "Class A" or "Class S" demon in some shows, while some need to "charge" a fireball powerful enough to blow a mountain off the map*, this kind of character basically just flicks it out of his hands**


*but for some reason it can't vaporize a flesh and blood opponent
**they all miss, and the area that used to look like the Rockies ends up looking like Kansas with rocks thrown all over it

Quote:

Originally Posted by AllsWell 

Lol ProtegeManiac, that is a great post/ explanation. Definately gave me a better idea on what exactly I'm dealing with. ( Now I wonder why teachers in real life can't explain things like that .... hmmm)

There are probably older teachers with a lot of experience but can't "connect" - that's why younger teachers with less experience compensate with rapport. An older teacher even if using abstract parallels is part of his style won't think of using Dragonball/YuYuHakusho/Flame of Recca/etc if he's never seen those. Coincidentally I think I'm entering that phase of getting "older," because by now the only anime I'm inclined to watch are the reincarnations of Gundam, "reincarnation" being the keyword here. I'm old enough that I watched the original on (a dozen) VHS casettes :P

In any case glad I was able to clarify my point. Most people normally use hoses and floodgates when explaining these though.

Quote:

Originally Posted by AllsWell 

Ok so I finally received my "Burr Brown OPA637AP with Class A Biasing Mod" for my Matrix M Stage which I'm VERY pleased with.

 

The soundstage on my DT990 (600ohm) has received more depth, there is more of a viscerality to the music, more of a "grandeur" . everything is just more "powered up" and better lol. I am pleased, except for one thing.

 

The new Burr Brown Opamp I purchased from Tams Audio generates (or adds to) a bit more heat on my Matrix Amp. It's much warmer now than it was before at stock configuration.

 

Has anyone noticed this / experienced this ?

 

Is this a normal thing (nothing to worry about ?) It's not HOT to the touch , just warmer than before.

 

Anyways, I think it was well worth the extra 50 bux for the upgrade IMHO.

 

 

I assume you are talking about the Class A biased OPA637 Op Amp running hotter than the stock Op Amp?

Yes the Class A biased OP Amp will run hotter than the stock one and hotter than a standard OPA627 or OPA637.

The Matrix M Stage itself always runs very hot regardless of which Op Amp you use.

But the OPA637 in any format is not recommended for use in the M.

 

Try using a standard OPA627, personally I don't hear a huge difference between the two versions (Class A and non Class A) myself, in fact, I think the stock OPA627 is a touch clearer.

 

Quote:

Originally Posted by nullpo 

The OPA637AP isn't well suited for this circuit and coupled with the class A biasing I'm sure it runs quite hot.

He's using the OPA637 which is NOT unity gain stable (i.e. 0 dB gain stable), but you are correct, it will oscillate in this circuit at 0 or 10 dB of gain.

OP should run the M Stage at 20 dB or 18 dB of gain to avoid oscillation with the OPA637 (bad, very bad!).

The output stage of the M Stage uses two output transistors per channel.

One is used for amplifying the positive half of the signal, the other is used for the negative half of the signal.

In Class A, both transistors are ON all the time for all of the signal.  Class A runs hot.

In Class A/B the positive transistor is ON only for the positive half of the signal and negative transistor is only ON for the negative half of the signal.  Class A/B runs much cooler.

In Class D the output transistors are switching ON and OFF at a speed much, much higher that the audio signal, think 500 kHz or 100 kHz.  It's a bit like recreating a signal from ONEs and ZEROs in a DAC.

Hmmmmmm,

Maybe I'll add something to the Glossary Of Terms.

Quote:

Originally Posted by nullpo 

I don't know that it is recommended by the manufacturer, but it is indeed sold by a reseller who has a good rep (though his OPA627s are suspiciously cheap).  OPA637 is a fast opamp and can have problems with oscillation, especially at unity gain.  I haven't used it in this application so I can't offer any more than that.  There are a number of cheaper, more forgiving options out there though.  OPA602BP is my favorite at the moment.

Ooops.

One of these days I have to learn how to read!

The OP was referring to the OPA637.     I read the title too quickly and thought he was referring to the OPA627. Doh!

Nullpo is quite right.

The OPA637 IS NOT stable at gains less that 5.

This means you must always run the Matrix M at a gain of either 18 dB or 20 dB to avoid oscillation.

This could be the reason why your OPA637 is running so hot.

In addition, the oscillation is not good for your headphones and may fry them.

Due to the limitations of the gain settings, I do not believe the OPA637 is suitable for the M Stage.

Stick with the OPA627.

You don't need the OPA637's high speed in a headphone amp.

And I agree, his OPA627s are suspiciously cheap! Could be fakes!

My personal favourite Op Amp for the M is the LM4562. But I certainly haven't tried every Op Amp out there!