Very interesting suggestion. Although not that easy to implement. Last time i tried doing that was with a 5534 where it really made a difference. I guess it's the only way to get the full CMRR in a practical circuit.
Instead of giving my oppinion and trying to help, I should shut my mouth.
Thanks for helping me figuring this out.
Go on... [/B][/QUOTE]
-----------------------------------------------------------------------------------
How do you know that findings in your system apply to others? To me, you are not so much giving opinion as asserting that you speak the truth.
Now, most of us accept that othetr people have different, not inferior, and often superior systems.
The performance of an opamp in a dac more than 20 years old doesn't say much for me.
Thanks for helping me figuring this out.
Go on... [/B][/QUOTE]
-----------------------------------------------------------------------------------
How do you know that findings in your system apply to others? To me, you are not so much giving opinion as asserting that you speak the truth.
Now, most of us accept that othetr people have different, not inferior, and often superior systems.
The performance of an opamp in a dac more than 20 years old doesn't say much for me.
Read my post, I don't limit my tests to one system.
The dac in question has HDCD, and the much beloved AD1862 dacs, and with some mods (including regulated PSU) is serious business.
And it's not 20 years old, it's around 10 years old.
You are talking without knowing the product in question.
But forget it, go on...
The dac in question has HDCD, and the much beloved AD1862 dacs, and with some mods (including regulated PSU) is serious business.
And it's not 20 years old, it's around 10 years old.
You are talking without knowing the product in question.
But forget it, go on...
hint
Terry's post on the previous page gives a clue of one of the tricks I use to make the OPA627 sing.
What is it?
When they re-open my thread, cleaned of all the off-topic thing, I'll say it.
Terry's post on the previous page gives a clue of one of the tricks I use to make the OPA627 sing.
What is it?
When they re-open my thread, cleaned of all the off-topic thing, I'll say it.
analog_sa said:
It can be implemented in just about any circuit
that does not have variable conditions ie; after a
volume control. Hint, the OP looks like gnd WRT components
in FB path. Yes, I know this sounds wrong but if you
do a careful analysis it is right.
Cheers
Terry
I don't know how to compare op-amps. I just use them in several circuits
with varying gain. Some are biased some are not. But OPA627 is never been
my favourite. I use them (BP/AP) for error correction in power supply 😀
May be Carlos has something I don't know. I guess OPA627 is not designed
specifically for audio. And I think BB/TI had explained that the OPAx134,
while inferior in numbers, is very musical. And this is true according
to my ears.
But some audio manufacturers use OPA627 in their higher-end stuff, so I
think they have a reason to use this opamp (because the cost may not be
proportional to the improvement). Maybe by looking at these circuit
things going to be clearer. So far, the answer is around the bias option.
with varying gain. Some are biased some are not. But OPA627 is never been
my favourite. I use them (BP/AP) for error correction in power supply 😀
May be Carlos has something I don't know. I guess OPA627 is not designed
specifically for audio. And I think BB/TI had explained that the OPAx134,
while inferior in numbers, is very musical. And this is true according
to my ears.
But some audio manufacturers use OPA627 in their higher-end stuff, so I
think they have a reason to use this opamp (because the cost may not be
proportional to the improvement). Maybe by looking at these circuit
things going to be clearer. So far, the answer is around the bias option.
Re: hint
Perhaps it is time to remind of Walt Jung´s articles about audio opamps...
Walt´s articles
analog_sa said:
carlosfm said:Terry's post on the previous page gives a clue of one of the tricks I use to make the OPA627 sing.
What is it?
When they re-open my thread, cleaned of all the off-topic thing, I'll say it.
Perhaps it is time to remind of Walt Jung´s articles about audio opamps...
Walt´s articles
OPA134
Hi Jay: I recently swapped out a pair of OPA134's that I put in a California Gamma DAC (PCM1710) with OPA627's. I found the OPA627's produce a very different, much clearer, but less mucical sound than the OPA134. I ended up swapping the OPA134's back in. I looked all over the internet and could found only you who reported a similar experience. Whew! What a relief. I thought I was losing my hearing marbles. Why do you think that I had such a radically different impression on this most highly regarded OPAMP? Personal preference or uniqueness of the Gamma circuitry? Is there any OPAMP that you consider as or more musical than the OPA134? Best regards, Tim King
Hi Jay: I recently swapped out a pair of OPA134's that I put in a California Gamma DAC (PCM1710) with OPA627's. I found the OPA627's produce a very different, much clearer, but less mucical sound than the OPA134. I ended up swapping the OPA134's back in. I looked all over the internet and could found only you who reported a similar experience. Whew! What a relief. I thought I was losing my hearing marbles. Why do you think that I had such a radically different impression on this most highly regarded OPAMP? Personal preference or uniqueness of the Gamma circuitry? Is there any OPAMP that you consider as or more musical than the OPA134? Best regards, Tim King
Well, since you're all so polarised could I suggest you analyse the effect of your biassing efforts on the residual spectra of each device under it's specific loading. +ve or -ve rail can make a HUGE difference to the residuals.
Perhaps that's where your answer lies.
Cheers,
greg
Perhaps that's where your answer lies.
Cheers,
greg
OPA627 Class A
Greg:
Thank for your comments. I have biased the OPA627's to class "A" with a 2.7 kohm resister and the sound is much better now. Nobody seemed to have memtioned that OPA627 operating in class "AB" may have somewhat harsh high's and restricted, unnatural mid's. I plan to keep the OPA627 in my DAC now. I wonder why OPA134 sounds pretty good in class "AB".
Best regards,
Tim King
Greg:
Thank for your comments. I have biased the OPA627's to class "A" with a 2.7 kohm resister and the sound is much better now. Nobody seemed to have memtioned that OPA627 operating in class "AB" may have somewhat harsh high's and restricted, unnatural mid's. I plan to keep the OPA627 in my DAC now. I wonder why OPA134 sounds pretty good in class "AB".
Best regards,
Tim King
Re: Re: hint
This is before the output buffer, which is used to take the load (AC-load) OFF the input and gain controlling OP-amp.
This output buffer is within the global feedback loop.
Buffer is making first OP-amp only have to drive a very small AC-load (high impedance input of buffer)
but driving a small constant current source, DC-load, like 1 mA.
A simple resistor from output to negative supply
will increase the AC-load of a single OP-amp (without output buffer.
And we all know that the the more such load we put to an OP-amp output, the higher distortion. Until it can not drive anymore.
The lower the resistance of the load, the worse figures.
If you want to put in an assymetrical output load to an OP-amp, it is prefered to use a constant current source.
Between output and positive or negative supply rail.
This will not add to AC-load. For AC-signals.
Walt Jung mentions to put a load to the output of OP-amp. (see attached image)Nicke said:
This is before the output buffer, which is used to take the load (AC-load) OFF the input and gain controlling OP-amp.
This output buffer is within the global feedback loop.
Buffer is making first OP-amp only have to drive a very small AC-load (high impedance input of buffer)
but driving a small constant current source, DC-load, like 1 mA.
A simple resistor from output to negative supply
will increase the AC-load of a single OP-amp (without output buffer.
And we all know that the the more such load we put to an OP-amp output, the higher distortion. Until it can not drive anymore.
The lower the resistance of the load, the worse figures.
If you want to put in an assymetrical output load to an OP-amp, it is prefered to use a constant current source.
Between output and positive or negative supply rail.
This will not add to AC-load. For AC-signals.
Attachments
And not only Walt Jung, but also Linear Technology suggests biasing op amps, so it is not just something that floats around in the audio DIY community.
http://www.linear.com/pc/images/products/thumbnails/1293.jpg
(see LT1115 datasheet for larger picture)
http://www.linear.com/pc/images/products/thumbnails/1293.jpg
(see LT1115 datasheet for larger picture)
Guys, the OPA627 sounds very good with just a resistor, as long as you bias it between 5 and 10ma (around 10ma is better).
I know it's not the ideal solution, that's very old story, but it does sound good.
If it was audible that the op-amp couldn't drive a 1.5~2.7k impedance I would not even use it.
A transistor can be used, but first try the resistor at these values. You'll think: why bother?
I know it's not the ideal solution, that's very old story, but it does sound good.
If it was audible that the op-amp couldn't drive a 1.5~2.7k impedance I would not even use it.
A transistor can be used, but first try the resistor at these values. You'll think: why bother?
Hotter than hell? Are we joking or what???
Are you sure, P-A?
Because it just gets slightly warm.
Sometimes I don't know what's on your mind, really.
peranders said:
Are you sure, P-A?
Because it just gets slightly warm.
Sometimes I don't know what's on your mind, really.
Upupa Epops said:
Check your maths.
And check what your friend Pavel Macura had to say about it:
http://www.diyaudio.com/forums/showthread.php?postid=452342#post452342
He biased at 6.85ma, so I suppose near to hell?
To P-A I say this: it's much less warm than a BUF634 in wide BW mode, without heatsink, at near +/-18V.
And P-A doesn't use heatsinks, he uses the SMD chip version.
I had a hard time x'plaining him that he should use the TO-220 version, heatsinked. For some reason there's a TO-220 version.
Now he is worried about the OPA627, which only gets slightly warm at around 10ma biasing (and I mean near 10ma, that's UP TO), nothing to worry about.
- Status
- Not open for further replies.