carlosfm said:
I've got really scared when I once saw the pimeta PCB.
With that PSU arrangement and very poor bypassing, I don't take for granted anything they say.
The BW resistor could make a difference, I can't deny it because I didn't test. I just can't see a point of using it.
But there's no harm to include it, as long as you don't have to make concessions to PSU layout.
I also can't see a point of biasing this circuit to class-A.
With that miserable PSU arrangement bass is not tight, detail is not even near to what this circuit can produce.
Then they go paralleling buffers to have "tighter bass".
In my oppinion, paralleling buffers should be avoided, stacking them should be avoided like the plague.
It is not needed for a line preamp, and it is not needed for 99.99% of the headphones.
As someone who's built 6 pimetas, including adding your "bypassing" arrangement, I can honestly say that there is NO DIFFERENCE. This is tested with a quality setup, since we last chatted on this item (HD-600 headphones, heavily modified SuperDAC + Monarchy DIP and modified NEC CD-602 transport).
Just to head your arguments off at the pass, I have added electrolytics (10uF cerafine or 100uF cerafine) bypassed by Panasonic MKPs and/or ceramics (depending on the board) at the opamp pins to see if it made a difference. There was none on any of my listening tests. I don't presently have equipment to perform measurements on the differences between the two.
I really think that you criticize others without ever trying or testing their stuff. If you performed tests that were objective and subjective, rather than subjective only, I would lend a lot more credence to your arguments about various circuits. I know that you're all about the listening tests, but you need to give credit where it's due.
Tangent, Morsel, and PPL have all performed comprehensive testing, both objectively and subjectively (listening). Their projects have been built by hundreds, if not thousands, of people around the world. If there really were a difference, I think we all would have heard it by now, or they would've updated the boards.
Regardless, my personal opinion is that the BUF634, as a buffer, is pretty limiting. Most of the headphone amp folks have moved on to either Elantec (intersil) buffers, discrete diamond buffers, or class-A output stages driven by an opamp for their designs. The intersil buffers are on par price-wise with the BUF634, but they don't provide output protection. I can't see why the improvements these folks have made on headphone amps wouldn't apply to a preamp as well.
With a single channel per board, as most of you are doing, I think implementing a discrete diamond buffer or class-a output stage would be pretty easy and lend well to creating a really high-performance preamp. Perhaps it's beyond the scope of this thread, as it seems more dedicated to the OPA627 + BUF634, rather than finding the best opamp + buffer combination. I would encourage those of you that are interested to take a look at tangent's site as well as head-fi to see the new designes described above.
I can personally vouch that the diamond buffers and elantec buffers are far better than the BUF634's in my experience. My amps have used both the OPA627 and the 637 with great results. Also, one more thing you may wish to consider adding, if possible, is rail isolation JFETs for the opamp. Measurement wise, it really helps with cleaning up the power supply to the opamp. I know this would add some additional complications to your layouts, but it may take you that much closer to the ultimate small/inexpensive preamp. The jfets are only $0.20 or so each (you'd need two -- 1 per rail).
motherone said:
That's impressive.
For me the audible difference is clear.
I suggest then, that you leave just the caps near the PSU connectors, entering the board.
Cheaper, smaller board.
Why bother?
You don't have to listen to me, as I tested all this years ago.
What counts is how many they have sold or built, right?
motherone said:
Those are some great thoughts man!
I actually have not limited my thinking to BUF634/OPA627 at all. I am actually not using that combination in my preamp right now. 🙂 I prefer the LT1115 to the 627, and also the AD8610 I think sounds better than the 627.
I am going to order some of theose intersil buffers! The specs look great! As for a diamond buffer, I am wide open to that idea also.
But to your point, I do think this thread was started precisely for the 627/634, but maybe we should start a new thread for "The ultimate GC preamp!" to spur on a more "reference" or "audiophile" design.
Here would be some cool features:
1) Discrete Alcaid type voltage regulators on board.
2) Suitable for any single opamp.
3) Discrete diamond buffer.
4) JFET rail isolation.
5) Option for biasing to class A.
6) Completely seperate SG/PG on board meeting at PS.
🙂
I would certainly want to participate in such a discussion, and would certainly consider that a hugely popular GB. At least as popular as this one.
Cheers!
Russ White said:
Russ, there's no such thing as "Ultimate".
There's always a way to improve or to make something better, and you could only call "ultimate" to anything if you have heard everything else. Mission: impossible.
You can go by two distinct paths: copy a design (like the Pimeta), or make something better (even with the same components).
There are many bad designs flowing on the net, don't be fooled by their success.
While the OPA627 (or other) + BUF634 makes a very good preamp, it's not the best thing on earth, of course.
Also, what can be a very good preamp can be a not-so-good headphone amp, and it seems everything is mixed on this thread.
As a line preamp, the AD815 beats the OPA627+BUF634 by a considerable margin.
Russ White said:
What can I say, I'm a big fan of capacitance. Yes I could remove those caps and shorten the boards a little. Also this could be done one sided if you jumper between the signal ground and power ground on the top layer, my initial design had traces connecting the two together, but I opted to go 2 layers. For me I think this will be the last design using this combination. It fits my needs as a the line amplification section in a home theater amp. Also the boards are small so I can put several into a chassis. The in/out placement should work well with the digital source selection I am working on. Right now I need to focus on the power supply. Does anyone have a PCB layout for the discrete regulator on Nuuk's website, I believe it is an Andrew Rothwell design?
Thanks. 🙂
True, true..... Might be funny to give it a shot, anyways😉
G. your board looks great.
I have been looking for that too, no luck though. If you find it, (or make it), be sure to keep us posted.
Steen🙂
carlosfm said:
Hey, if the circuit has stability under all test environments, and someone throwing it on a scope can't measure any difference, why not?
And yes, it does count how many times they've been sold or built. If you're the only one who can hear the difference, maybe it's all in your head... That is one of the main problems with purely subjective testing.
Russ White said:
Cool. A modular design would be nice, but I don't think that's the ultimate goal of this thread. I suppose another option would be to skip the TO-220 packaged BUF634 and maybe go for a DIP-8 packaging -- you could probably design the pads to accomodate a BUF634 or an Elantec/Intersil buffer.
Russ White said:
Many folks are moving on to Diamond Buffers. However, the difference between a BUF634 and an Intersil buffer is far greater than the difference between the Intersil buffers and a diamond buffer, at least to my ears.
Russ White said:
I'd be game. If I had the time, I'd try to pick up PCB design and help out. Unfortunately, I haven't had much spare time lately. I agree with your features list -- those would be great for a preamp. I think that it would definitely be something worth attempting.
motherone said:
You can't measure, so no difference.🙄
Cool.
You are contradicting everything that most manufacturers say on the datasheets.
Why bypass?
No one needs a stinkin' cap.
motherone said:
Are you telling me that Michael Jackson makes good music just because he sells millions?
I cannot believe this way of thinkin'.
motherone said:
I listen with my ears.
motherone said:
Sure.
That's your problem.
Believing that that your measurements are so reliable that you can say "there's no difference".
Guys, listen to motherone: remove all the caps from the board.
I'm a lunatic.
If you can do it right, why do it wrong?
motherone said:
😱
No measurements this time?😕
Was that BUF634 correctly bypassed?😀
Aaahhh... forget it, move on.
If you are happy...
carlosfm said:
I never said not to bypass. There's a huge difference between not bypassing, and having a neurotic compulsion to put the caps on the leads of the opamp. Afterall, according to most of your posts, these PCB's would still be wrong 😉 They should only provide a pin for the electros to ground near the opamp, and the other leads should be soldered directly to the pins, along with a bypass cap! 😀
carlosfm said:
Are you telling me that Michael Jackson makes good music just because he sells millions?
I cannot believe this way of thinkin'.
Are you saying that his millions and millions of fans are wrong? Are you saying their *subjective* tastes are incorrect, because it doesn't line up with your mode of thinking? C'mon Carlos...
Regardless, I'm not touching musical tastes, as it's beyond the scope of this discussion.
carlosfm said:
I listen with my ears.
Sure.
That's your problem.
Believing that that your measurements are so reliable that you can say "there's no difference".
Guys, listen to motherone: remove all the caps from the board.
I'm a lunatic.
If you can do it right, why do it wrong?
You're right.. I actually stick ear plugs in my ears before I begin any listening sessions. That way, everything sounds pretty much the same.
As usual, you've some how derailed this into something else. You constantly bash on other folks designs, and state that you listen with your ears. What's hillarious about this is that your ears may not be able to tell you if something is going really wrong with the circuit. You may not know if you've really screwed something up. Hell, maybe subjectively, you like a screwed up sound! How are we supposed to know? You never check your circuits with any test equipment, so how do you know how it's really performing?
Afterall, if your ears are that perfect, maybe the audio companies can start throwing away their test gear and just contract out to you. "Hey, it passed the Carlos test, so it must be operating flawlessly!"
Imagine if the spacecraft, sattelites, or computers were designed the way.
carlosfm said:
I've never stated that measurements are the only way. I'm sure per-anders would know where to find them, as I hate trying to search on head-fi's system.
Yes, I have tried the "carlosfm" bypass method on my amps, and there was no difference. I'm sorry if I'm not an golden-eared audiofool enough to detect whatever it is you seem to romanticize on this. I really heard no changes at all.
Have you even tried buffers other than the BUF634?
Anyways, back to your regularly scheduled thread.
motherone said:
Don't play games with me, ok?
Some of the boards are much better now, following the suggestions I gave on PSU bypass arrangement.
If you really don't like it, don't use it.
Don't make me quote some datasheets, I've done that too many times here.
If you didn't hear any difference, that tells a lot of your system.
Your ears may have no fault.
motherone said:
Right.
That's what the caps do, isn't it?
Tighter bass = screwed up sound.
Maybe you like flappy bass.
Who knows?
Excuse me Mr., can I give my opinion on a PCB?
It is supposed to be constructive, even if you can't understand it.
Maybe I'm wrong in everything.
I wonder why there's a lot of people making money with my circuits and my name?😕
Maybe it's all junk...
OK can we please kill this BS and get back to the subject at hand. Lets just say we all have different ideas, and methods of testing. They don't often agree, and seldom do. Black Gate caps sound the best...but only after they have been cryogenically frozen then subjected to gamma radiation, and solderered on PCB's that were blessed by the pope. Being a scientist I have to take any "subjective" measure with a grain of salt, but that is not to say that subjective measures don't have there value...in reality we can often percieve things that we have not yet learned to measure...I have seen it too many times myself. 🙂
Now as for the "constructive" criticism...I believe the idea here is to make suggestions, then let others chime in. If we disagree than so be it, but we don't need 20 posts arguing over who is right. If it is something significant and we really care then we will likely inquire further on the subject, if not drop it an move on.
All of this pointless argument does nothing to advance the design. 😡
Nuff Said!
Now as for the "constructive" criticism...I believe the idea here is to make suggestions, then let others chime in. If we disagree than so be it, but we don't need 20 posts arguing over who is right. If it is something significant and we really care then we will likely inquire further on the subject, if not drop it an move on.
All of this pointless argument does nothing to advance the design. 😡
Nuff Said!
Recieved my boards today (the one with the ALPS on it and the rest) They look ok, but I did a mistake and now i'm angry at myself 🙂 but it's not an electrical fault though 🙂
Hehe, it was just that i want my boards looking nice and proper and i had turned off layer 25 in Eagle (the one with components-name C1,R1.. and so on) because i didn't want them on my boards, and because it was ment not to have them i didn't place them out properly (the names) and now i see that olimex turns that layer on by default i guess so now i have names all over the boards 😀
- Status
- Not open for further replies.