High-end preamp for my GC

[peranders]

sometimes you can even hear the same chips packaged using Korean dips and Indonesian dips: the indonesian dips tend to sound positively gray.

millwood, you are absolutely right but what about those IC's packaged in Thailand and Taiwan?

 

[carlosfm]

I take millwood's post as a joke.😀
Please keep this thread "clean".
😀

 

[peranders]

I take millwood's post as a joke.😀
😀

A joke? I thought he really ment it. It sounded so serious.

 

[matjans]

carlos, you're jinxed! 😀

 

[peranders]

I have some DIL versions of the BUF634 but never used them, because:

1. I can get a much tighter layout with the TO220 version (shorter feedback loop).
2. Heatsinked it has more current output (250ma against 200ma) and runs cooler (yes, this things can get hot without heatsink).
3. Some people reported me tighter bass with paralleled DIL BUF634s (one on top of the other, not ideal), while I have very good dynamics and tight bass, and I don't feel the need to parallel more buffers.
I tested several amps with my pre, and when I don't have very tight bass, the fault is not on the pre.
Also very important is to bypass each chip.
And to use a small cap (200pf is fine for me) on the feedback loop, as you can see on the datasheet, figure 4.
Oh, BW pin goes directly to V-.

It rather interesting how small thing some people claim that can hear. I don't doubt that a very few have very sensitive ears.

When I go to myself I have a very pratctical expample. I tested my headphone amp QRV-04 with OPA627 in the left channel and AD8610 in the right. Couldn't hear the slightest difference altought I wanted to. As you can see at the picture I have change idle current of the BUF634 with a jumper. Jumper off is "medium speed" and on is "high speed". Also in the case I can't hear a thing in difference and hear is so easy to change but still I can't hear. Two more people have tested this on their own. The same results.
At the picture you can se the jumper pins.

http://home5.swipnet.se/~w-50674/hifi_pics/hifi_100pr/qrv04r0_overview.jpg

What do I want to say with this? Some people are VERY self confident in what they hear. How many can put money on what they claim?

I think we all must take a very humble attitude when we talk about listening experience and which parts that are really involved.

 

[Andypairo]

i won't use an alps or standard pot here as i can get my hands on pretty cheeap (4 eur) 19 step 4-deck rotary switches with which i will build a stepped ladder attenuator.

An externally hosted image should be here but it was not working when we last tested it.

Hello matjans,
can you put your hands on a couple of these switches for me too?
Are they make-before break or break before make?

Cheers

Andrea

 

[millwood]

Some people are VERY self confident in what they hear. How many can put money on what they claim?

isn't that interesting? those people are so confident in their abilities to hear things that they are afraid to put money on themselves.

🙂

yes, peranders, I meant it a joke.

 

[carlosfm]

It rather interesting how small thing some people claim that can hear. I don't doubt that a very few have very sensitive ears.

What are you talking about, P-A?
I don't quite understand...
The 200pf cap on the feedback loop makes a huge difference in sound.
Without it, treble is a little on the bright side, and with less diffinition (a little splashy).
In some conditions (like higher capacitance interconnects or/and lower input impedance of the next stage) you may hear differences with both versions of the BUF634, heatsinked is better, and also TO220 is better for the layout if you compare to the DIL version.
My oppinion, take it if you want.

Note: I din't test the AD8610 with the BUF634, so I can't tell if you may find any difference in the sound.
Isolated, yes, they are different.

***EDIT***
I didn't say that direct BW to V- sounds better, it's just how I use it.

 

[peranders]

What are you talking about, P-A?
I don't quite understand...

I ment mostly in general....

At the moment I like AD8610 + BUF634, but it's only a feeling.

 

[matjans]

they're non-shorting (aka break before make). I'll try to get a few more.
I'm going on vacation to italy in a few days (lago maggiore). i'll send you an email if i have them and if i'm there i can mail them from there (or you can come pick them up).

We'll see.

 

[Nuuk]

they're non-shorting (aka break before make).

Might be a bit noisy when you change volume matjans!

May I suggest that you have a look at this ? 😉

 

[Andypairo]

they're non-shorting (aka break before make). I'll try to get a few more.
I'm going on vacation to italy in a few days (lago maggiore). i'll send you an email if i have them and if i'm there i can mail them from there (or you can come pick them up).

We'll see.

Thank you for your offer, let me know if you can get 'em.
BTW can you take Paypal (or do you prefer a trade with some components)?

Cheers

Andrea

 

[matjans]

nuuk: one parallel resistor solves the whole problem (or at least part of it). not ideal but it should work.

sorry i don;t take paypal.

 

[carlosfm]

I ment mostly in general....

And you believe in millwood's post?😕

At the moment I like AD8610 + BUF634, but it's only a feeling.

Although you can't find any difference between that and the OPA627 on your circuit.
Fine, sotimes it's just a matter of good feelings.😀
Sometimes intuition proves right, sometimes not.
BTW P-A, as the AD chips are cheaper, you may have a point.😉

 

[carlosfm]

Hi

My other thread is closed, so I post this here, I hope you don't mind.
I want to thank John (dhaen) and all the mods for taking the right decision.
I arrived home and was just posting a proposition for that thread to be closed while the mods do their cleaning, as this was not going anywhere.
You had already decided that way, congratulations.

I also want to thank publicly to George for sending me those Siemens caps for my Epos speakers.
Thanks George, I've just received them!😎

I also want to thank Pedja for being a pioneer.
Pedja, your work is very appreciated.

And Nuuk, he keeps making a huge public service to the DIY community, RESPECT!

I thought I just had to say this, I'll keep coming here if this 😎 s down.

Thanks for the attention.😀

 

[analog_sa]

Korean dips

Milwood

I was aware of your culinary expertsie but the question was directed to Carlos.

Carlos

Also, you get better results with the TO220 version of the BUF634, heatsinked.

I have some DIL versions of the BUF634 but never used them

So you don't really know then, do you?

2. Heatsinked it has more current output (250ma against 200ma) and runs cooler (yes, this things can get hot without heatsink).

A perplexing statement. 250mA is max available current in typical applications with no more than a few mA actual current. Will the difference between 200 and 250mA be audible?
Why would running cooler (with an unnecessary heatsink) be better? Often semis sound better when hot. Have you compared the sound without heatsinks?

Sorry i'm asking such a lame question instead of comparing on my own but my only current design using 634 is very awkward for testing.

 

[millwood]

I was aware of your culinary expertsie but the question was directed to Carlos.

I knew that but just wanted to help.

So you don't really know then, do you?

it sounded positively grey.

Sorry i'm asking such a lame question instead of comparing on my own but my only current design using 634 is very awkward for testing.

no need to be sorry about asking lame questions. we are used to them by now.

 

[carlosfm]

analog_sa,

Running the BUF634 at +/-18v (yes, no problem), without heatsink you can't even touch it.
Even at +/-15v it gets very hot on some conditions.
Why use the DIP (or DIL if you want), it you can get a better layout with the TO220?
I don't know about you, but I can get a much shorter feedback loop.
I had reports of untight bass with the DIP version, but I don't have that problem, and I've figured out by now that I can't trust everybody's oppinion, because this thing demands a good layout.
Even more demanding than an OPA627 alone.
Anyway, I'm a little surprized to see you talking about op-amps.
On this thread, some months ago, you said that since the NE5532/4, nothing has changed.😕
Please, I tried to help everybody that came to this thread, but as things go I'm really ****ed off.
If you find it hard to belive that a product has a TO220 version for some reason, then😎 .

If current doesn't matter (a few ma is enough), stick to a single op-amp.
Why use buffers?😕

 

[MalichiConstant]

Using buffers for low current

It seems to be a pretty popular practice for audio circuits with op amps. The buffer does not have to be an IC if you don't want. PS Audio used a jfet as a follower inside the feedback loop of an op amp based preamp nearly two decades ago. I think I have seen a circuit with a small power mosfet in the loop as a buffer as well. It is not just layout and decoupling but the relative bandwidths of the op amp and buffer are critical and the buffered op amp circuit is still sensitive to capacitive loading and benefits from a series output resistor.


From Walt Jung:

http://home.comcast.net/~walt-jung/wsb/PDFs/WTnT_Op_Amp_Audio_1.pdf

http://home.comcast.net/~walt-jung/wsb/PDFs/WTnT_Op_Amp_Audio_2.pdf

http://home.comcast.net/~walt-jung/wsb/PDFs/Composite_Line_Driver_with_Low_Distortion.pdf

 

[carlosfm]

On the BUF634 datasheet

1. "Although all four package types are tested for the same output performance using a high speed test, the higher junction temperatures with the DIP and SO-8 package types will often provide less output voltage swing. Junction temperature is reduced in the DDPAK surface-mount power package because it is soldered directly to the circuit board. The TO-220 package used with a good heat sink further reduces junction temperature, allowing maximum possible output swing."

2. "Bandwidth can be extended to approximately 180MHz by connecting the bandwidth control pin to V–. This increases the quiescent current to approximately 15mA."

Just a comment: why bias an op-amp to class-A here, like on the Pimeta circuit?😕

3. "The –3dB bandwidth of the BUF634 is approximately 30MHz
in the low quiescent current mode (1.5mA typical).
Output voltage swing is somewhat improved in the wide
bandwidth mode. The increased quiescent current when in
wide bandwidth mode produces greater power dissipation
during low output current conditions."

4. "High speed or fast-settling op amps generally require the wide bandwidth mode to remain stable and to assure good dynamic performance."

5. Check this table, and the recommended mode for some op-amps. It's all there, isn't it? If you can use the best type (TO220) and the recommended mode, why not use it?