[Nezer]

As I reported on another post I ordered a crapload of opamps to test.

Among those ordered were a few OPA637.

The package arrived this morning and I have't designed my test circuit yet.

But I was most intrested in hearing the OPA637 after reading the data sheet. Basically, it's a OPA627+. After hearing high-remarks over the 627 (another opamp in the batch) I had the highest hopes for the 637.

I really have nothing to compare it to as of yet except the LM6171s that are in my modded Corda HA-1 (upgraded power caps, posted in another thread).

Being the impatient SOB I am I just *had* to drop in the 637 (this is the one of the ones I was really intrested in) into my modded Corda HA-1. After lifting pins 1 and 5 out of the way (why does Dr. Meier have these connected on the PCB; they aren't even used on the 6171s he uses) I gave em a test drive for a few minutes (after plugging them into the first stage of course).

The first thing I noticed is that the current draw needs to drop. These things got pretty hot, though well within operating conditions (I could touch them for about 10-15 seconds before I couldn't take it any longer); certainly hotter than the 6171s sitting on the second stage. This was not unexpected based on some recent posts about dropping-in the 627s into the first stage of the Corda.

In listening to music I found that they appeared to pull a bit more detail out than the 6171s. Certainly not to the EAR HP4-level but we're getting closer. It was a significant improvement over the 6171s and I would imagine lowering the class-a current draw would help. Things I had never heard before included someone knocking something around during the opening passage of Miles Davis' So What, the opeing track from Kind of Blue. The shaker from Dire Straits Why Worry from Brothers in Arms was clearer. I'm a detail freak and I really got off on this.

The next thing, and this just clobbered me upside the head, was the soundstage. These little puppies really seperated *everything* and put each instrument in it's place! On Ani DiFranco's Everest from Up up up up up up, her voice was *clearly* behind and to the left of me. It almost sounded like she was standing over my shoulder singing in my ear (which has long since been a fnatasy of mine but I digress).

The thing that really blew me away was how this added detail and more-accurate soundstage really brought-out the little things. For example, in Sting's Fill Her Up from A Brand New Day the acoustic guitar comping chords was both acurate and well-placed. It's back in the mix a bit but it really stood-out in the subtle way it should. I should add that I *know* what a comping guitar sounds like, I've played guitar for 15 or so years and started by playing that style in a high-school jazz band. This opamp brought it out and presented it the way it *should* sound.

Needless to say I'm *very* impressed by the 637! I think the other opamps in the package are going to have a hard-time standing-up to them.

These single-channel opamps in thier 'a' class that I sampled (the 'b' class is supposed to have tighter tolerences across the prodcution lines) go for a little over $30 for a pair from DigiKey. The B incarnation goes for just over $44 a pair.

IN comparison the LM6171s go for just over $6 a pair. Quite a difference in price but also quite a differenc in sound, IMHO. I would personally gladly shell-out the $30 for the 637s any day over the 6171s but at the same time I can't fault Dr. Meier for his choice to use the 6171s as it's a damned fine opamp and especially fine considering it's price!

For now the 637s have been removed from the Corda and nestled neatly back into the packing tube awaiting my building of a test circuit. I'm thinking a Meta42 equipped with these opamps might just be as close to Nirvana for a reasonable price I'm going to get in this lifetime. Only one way to find out... Tangent, could you quote me a...

 

[erix]

I have a pair of those around here somewhere.

For research purposes....

ok,
erix

 

[adewolf]

Nezer,
Three days ago I posted at Headwize on some of my impressions about LM6171 vs. OPA627.

I tested opa627 in my DIY meier's amp HA-1.

OPA 627 is a better sounding op-amp than the one by National.
Yes they're getting hot within seconds, BUT I'm doing the mods suggested by Jan Meier and PPl and hope it works: replacing the 1.3 kOhm to 3.3 kOhm between voltage rails and the opa's output; and using small heatsinks over the opa's surface.

No question. OPA627 sounds better (I mean, playing classic music, I don't know with other kinds of music).

Now I'm going to ask PPL on how to use the pair of AD8620's which arrived at home today. I'm very excicted about to test these glorified ones.

 

[Nezer]

I've decided to hold off on any further testing until I build a Meta42-based test circuit. I have a *ton* of opamps here from TI to test.

I'd like co compare it to some of the other offerings in the world, but I think there will be a hard time holding-up to the 637! I mean, if the 637 sounded that good with that load on it with a 6171 as a follower, how great would it sound if biased with a better current drain and using a buffer as a follower??

I do have some 627s to test as well.

 

[adewolf]

Nezer
------
if the 637 sounded that good with that load on it with a 6171 as a follower, how great would it sound if biased with a better current drain and using a buffer as a follower
------

I'm just using four 627; do you think this is wrong?

 

[Nezer]

Quote:

Originally posted by adewolf Nezer ------ if the 637 sounded that good with that load on it with a 6171 as a follower, how great would it sound if biased with a better current drain and using a buffer as a follower ------ I'm just using four 627; do you think this is wrong?

There's nothing 'wrong' with anything if you like what you hear!

Just from what I've been reading using a buffer in the second stage with a buffer can help with high-impedance cans and doesn't hurt lower impedance ones. From what I understand the opamps do a great job and voltage amplification (the purpose for the first stage) while buffers perform better on the second stage to feed current to current hungry headphones.

You might try building one wiht 627s on the first stage (or 637s if you want to hear em) and an EL2001 buffer on the second stage. It *might* sound better to you than using the 627s on the second stage and would likely be cheaper to boot.

I'm jsut starting to get into the whole DIY thing but these opamp circuits are really simple electrically speaking. I may soon have a hard time justifing buying a Max when I can make something myself that I might like even better!

Right now my 'dream' am is a Meta-42-based circuit with the CRD in the bias circuit running 637s (maybe 627s maybe something else, have to test) on the first stage and an EL2001 on the second stage configured in a Jung multiloop with a gain of about 5-6 and a bandwidth limit of about 100khz. Add to that a total-overkill power supply with some HUGE Cerafine caps, voltage regulators for the rails and probably an LM6x21 or an EL2009 as the power supply buffer to stabalize the rails and provide some decent current. I'm thinking this could be the amp for me!

 

[ppl]

If using the OPA-637 i would use a widebandwidth buffer like the EL-2002,EL-2009,OPA-633 BUF-634 operated in the wideband mode (Pin1 Connected to V-) as the Output stage. the Slower EL-2001 will handel the OPA-627 but to expect a 70 MHZ buffer like the 2001 to be stable with a 80MHz. opamp like the OPA-637 is asking alot.

 

[andrzejpw]

nice review! I've had my eye on the 637 for a while now. . .

 

[Nezer]

Quote:

Originally posted by ppl If using the OPA-637 i would use a widebandwidth buffer like the EL-2002,EL-2009,OPA-633 BUF-634 operated in the wideband mode (Pin1 Connected to V-) as the Output stage. the Slower EL-2001 will handel the OPA-627 but to expect a 70 MHZ buffer like the 2001 to be stable with a 80MHz. opamp like the OPA-637 is asking alot.

Ideally, for a real amp, wouldn't you want to limit the bandwidth to something way more same? I have pretty good hearing and I top-out about 19k, 8-0 M seems a bit overkill.

 

[tangent]

The bandwidth of op-amps (and buffers) has little to do with audio bandwidth. It's more related to slew rate -- the faster the chip can slew from one end of its range to the other, the greater its bandwidth. Since you want high slew rate, you get high bandwidth as a result.

Limiting bandwidth of the op-amp is only practical in multilooped amps like the META42. There are a great many other op-amp applications where you can't limit the bandwidth, and so ppl's advice for those cases still stands.

Even in a multilooped amp, I would still consider using the EL2002 with the OPA637. You're intentionally choosing a high-spec chip here, so you might as well use a high-spec buffer along with it.

 

[Nezer]

Quote:

Originally posted by tangent The bandwidth of op-amps (and buffers) has little to do with audio bandwidth. It's more related to slew rate -- the faster the chip can slew from one end of its range to the other, the greater its bandwidth. Since you want high slew rate, you get high bandwidth as a result. Limiting bandwidth of the op-amp is only practical in multilooped amps like the META42. There are a great many other op-amp applications where you can't limit the bandwidth, and so ppl's advice for those cases still stands. Even in a multilooped amp, I would still consider using the EL2002 with the OPA637. You're intentionally choosing a high-spec chip here, so you might as well use a high-spec buffer along with it.

This sounds reasonable.

Maybe I'll cough up the cash somehow to get some parts from you. I wonder if I have anything left to sell... Maybe I could sell my Porta Corda to get the parts for a new portable amp that I'll build.

 

[braillediver]

I just ordered samples of the: LM7171 Very High Speed, High Output Current, Voltage Feedback Amplifier

This is supposed to have improved bandwidth, slew rate, input bias current and output current over the LM6171.

Has anyone ever tried it? These free samples are addictive.