| carlosfm |
At request of several families:D , my PM is full of questions:bawling: , here's the trick.
It's easier than you think.
On some op-amps you notice better sound by doing this, on other ones you don't.
But the OPA627 really needs this to sing properly.
As simple as a resistor from V- to output, biasing the op-amp in to class-A at between 5 to 10ma.
For +/-18v PSU I use 2k, and for +/-15v PSU I use 1.8k.
You can use anything that bias it from 5 to 10 ma, it's not critical, although I usually go for around 10ma.
No need to use jfets to do this, just try it with a resistor at around these values and be amazed.;)
:cool: |
|
|
| Sunsun22 |
Hi Carlos,
Do you need any special type of resistor to do this biasing?
Sunny |
|
|
| Algar_emi |
Hi Sunsun22. Let us know if you implement this trick on the ML38 if it improve the ML38 sound.
Didi you found the round can IC to DIP adapter for the ML38 upgrade?
Thanks...
SB |
|
|
| Algar_emi |
Hi Carlos. Do you think this trick can also improve the OPA2604 sound?
Thanks... |
|
|
| analog_sa |
| At the time when i tested the OPA627 and decided it's not my cup of tea i used a FET CCS at output set to 2 mA. Indeed, it does improve the sound, but not enough to make it listenable. |
|
|
| Sunsun22 |
I will try this trick on my ML38 and let you know later. However, I am still waiting for some parts from Farnell to upgrade the PCB. According to Farnell, my parts will arrive by next Monday, I'll give it a judgement after a few days of sit in.
Up to this moment, I still cannot find the adapter for the MAT02/03 but I think I am not going to change it at this moment because MAT02/03 are pretty expensive and I need a few more of them to match. Too costly.
Sunny |
|
|
| Nuuk |
When I biased my OPA627's into class A, I thought they actually sounded slightly worse!
AFAIK, the circuit was fine and there was an improvment with the NE5534's when used in the same circuit.
Strange thing this hobby! ;) |
|
|
| PMA |
| quote: | Originally posted by carlosfm
But the OPA627 really needs this to sing properly.
As simple as a resistor from V- to output, biasing the op-amp in to class-A at between 5 to 10ma.
|
Carlos,
I am going to check it this evening.
Cheers, Pavel |
|
|
| Upupa Epops |
| To PMA : By biasing you must use Caddock or bulk metal Vishay :D , with cheaper ones will not results convincing :whazzat:, remember on this ! |
|
|
| matjans |
| and, of course, *proper* supply bypassing. |
|
|
| carlosfm |
Biasing at around 10ma with "normal" 1% film resistors works like a charm for me.:cool:
| quote: | Originally posted by matjans
and, of course, *proper* supply bypassing. |
That should always be done, or it won't sound as it should.;)
But that, I think, should be clear by now, after so much talk.:D
Note: this server has more crashes than Windows.:clown:
To post this note not I took some 10 minutes, plenty of patience and retries.
See you maby tonight, it's no good now.:bawling: |
|
|
| carlosfm |
| quote: | Originally posted by Nuuk
When I biased my OPA627's into class A, I thought they actually sounded slightly worse!
AFAIK, the circuit was fine and there was an improvment with the NE5534's when used in the same circuit.
Strange thing this hobby! ;) |
Nuuk, that's not strange at all.
The OPAs, and most of the modern/fast op-amps, need a good layout and PSU bypassing. They are very picky.
The fact that an NE5534 works better on a circuit doens't mean nothing to me, I find it normal, I've experienced that too many times.
Put the right caps where they are needed and youl'll have the OPA627 (and many more) largely outperforming the NE5534.
But then again, if you like the NE5534's sound, I don't discuss that.:angel: |
|
|
| mlloyd1 |
thanks for the tip!
have you noticed any negative impacts on thermal performance, dc offset/drift and noise?
mlloyd1
| quote: | Originally posted by carlosfm
...
As simple as a resistor from V- to output, biasing the op-amp in to class-A at between 5 to 10ma.
...
|
|
|
|
| PMA |
| quote: | Originally posted by PMA
Carlos,
I am going to check it this evening.
Cheers, Pavel |
OK - I have done it. It does not work for every opamp. But it does work for the OPA627.
My application is the voltage stage in my new amplifier project (error correction amp). The opamp is supplied from +/-18.5V (LM117+LM337). I have chosen the 2.7k resistor from the opamp output to -Vs.
For OPA627, the improvement is substantial. The effect is in clearer mids and highs and the overall resolution for the complex music. The music is more liquid. And the "tail" dissappears. Absolutely positive, nothing to complain. Only very very well designed discrete circuit is capable to give same sound quality as OPA627 with increased Iq.
The biasing against -Vs does not work for AD797 and brings no effect for OPA134. |
|
|
| carlosfm |
| quote: | Originally posted by mlloyd1
thanks for the tip!
have you noticed any negative impacts on thermal performance, dc offset/drift and noise?
mlloyd1
|
No.
It just gets a little hot, but nothing to worry about.
DC-offset remains the same.
| quote: | Originally posted by PMA
The biasing against -Vs does not work for AD797 and brings no effect for OPA134. |
My experience with the OPA132/2132 too, I feel it doesn't make a difference with this one.:angel:
| quote: | Originally posted by PMA
OK - I have done it. |
You're fast.:D
| quote: | Originally posted by PMA
It does not work for every opamp. But it does work for the OPA627. |
:angel:
| quote: | Originally posted by PMA
For OPA627, the improvement is substantial. The effect is in clearer mids and highs and the overall resolution for the complex music. The music is more liquid. And the "tail" dissappears. Absolutely positive, nothing to complain. Only very very well designed discrete circuit is capable to give same sound quality as OPA627 with increased Iq. |
Oh yes!
Amen. :angel:
:happy1: :cheerful: :nod:
:usd:
Note: I have here several new op-amps for some time to test but don't really feel the need... I'm lazy...
I'm not so convinced if it's possible to better a well implemented OPA627/37...
I use them for years and I just love them.:hbeat: |
|
|
| Sunsun22 |
| Has anyone try this biasing on AD823 or OPA2604? |
|
|
| Dr.H |
After all the benefit shown from a simple R between V- and output, read how bad this is for sound...
http://tangentsoft.net/audio/opamp-bias.html
Then try the cascoded JFET baising. I biased to 2.5mA using the JFET method and found the OPA627 sound to open up, especially in highs and mids. Bass was about the same. I think my experience is described in the High-end pre-amp for my GC thread.
Maybe I'll try the higher levels Carlosfm suggests, but I have seen a post on headwize suggesting that higher than 3mA added grain to the sound.
Having said all that, I think that the THS4061 UNBIASED still sounds better than a 3mA biased OPA627...;) , especially in the highs. Clear, but delicate, this being in the IV stage of a DAC.
Maybe I'll try the bias on THS4061 tonight.... |
|
|
| Nicke |
| quote: | Originally posted by PMA
The biasing against -Vs does not work for AD797 and brings no effect for OPA134. |
Perhaps you should try biasing with a little lower current and use a JFET CCS instead.I tried that years ago on a AD797 with good results.I used a 2mA CCS biased to -rail.I guess that the 0,5mA output stage bias in AD797 isn´t enough...
Once I tried 2mA CCS´s with LM6172 and that really improved alot.
Nicke |
|
|
| PMA |
| quote: | Originally posted by Dr.H
After all the benefit shown from a simple R between V- and output, read how bad this is for sound...
[ |
Ryan,
I do
1) design circuits
2) measure
3) listen to the results
Omitting any of the 3 mentioned points is a fatal error ;) |
|
|
| Nicke |
| quote: | Originally posted by Sunsun22
Has anyone try this biasing on AD823 or OPA2604? |
No,but I would not bias the AD823 too heavy.
It does not have the best current capability,17mA max.
But I´ll guess that a couple of mA wouldn´t hurt.
Nicke |
|
|
| Pan2 |
PMA,
"The music is more liquid. And the "tail" dissappears. "
What do you mean with ""the tail"?
/Peter |
|
|
| PMA |
Let's forget over-masterred music (techno, funk, rock etc.) and consider acoustical instruments, classical music.
If you listen e.g. to the sound of the recorded piano, sometimes you may hear a sound that comes "behind" the piano tone. That's what I call "tail". It means unclear, masked, blurred sound. |
|
|
| 5th element |
This is very interesting, I use AD8610's in the tweeter section of my active xover, OPA627's in the midrange and OPA2134's on the bass.
Id be interested in biasing my OPA627's in class A etc if it really is worthwhile. Bare in mind this is 10 opamps lol. Now I take its exactly the same with an active filter with the resistor use?? Also if you are using 12volt supply lines what resistor value would you suggest? Baring in mind some of the stages within the xover have high gain (open baffle compensation etc).
Cheers for any help. |
|
|
| carlosfm |
Instead of reading anything you find on the net and take it serious, to have an oppinion you should test the method I described, with similar values, like Pavel (PMA) did, and listen .
You won't feel the need to use fets, believe it or not.
Sometimes people find it hard to make it simple...:rolleyes:
| quote: | Originally posted by Dr.H
Clear, but delicate, this being in the IV stage of a DAC. |
Don't take conclusions just by testing an op-amp on an I/V stage of a dac, as in some cases a bipolar op-amp may work better there.
In that case, you're comparing different things.
Try the output buffer/filter stage too.
Try a pre.
etc. |
|
|
| matjans |
basic stuff: v=i*r thus r = v/i
choose a bias current (say 7,5mA or 0,0075A) and fill in the formula:
r=12/0,0075 = 1600 or 1.6k ohms.
do check the datasheets for the different opamps for current capability, some opamps (like nicke said about the ad823) can't deliver too much. |
|
|
| Upupa Epops |
| To Pan : If you will have several amps, which every will have " tail ", you hardly recognize it, 'cos you will think, that is normal. But if you once will hear amp without tail, you will know this forever. Unfortunately absolutely most of amp tail have. |
|
|
| carlosfm |
You don't know the real thing until you hear it.:D
Some people I know lived happy with midi systems and all-in-one AV:dead: packages until the day they came to my home.:bawling:
I feel petty, because some of them became very unhappy.:clown: |
|
|
| carlosfm |
| quote: | Originally posted by 5th element
Now I take its exactly the same with an active filter with the resistor use?? |
Yes
| quote: | Originally posted by 5th element
Also if you are using 12volt supply lines what resistor value would you suggest? |
A standard value of 1.5k will do the job.
Wonderfuly.:D |
|
|
| air |
Tried it... its playing... its great :D
First I tested with my preamp running only one OPA627AP on each channel, biased into class A with 8mA. Wasnt really impressed, until I changed the opamps to OPA627BP. Yes, I first tested the BP´s without class A biasing, and they sound better with bias! |
|
|
| Upupa Epops |
| Guys, whatabout trimpot - small class A, middle class A, high class A, overdimensioned class A etc. :D . |
|
|
| 5th element |
| OK well ive looked thru all my resistors and ive got enough 1k's for both the DAC filter (OPA627BP's) and the filter stages in the xover (OPA627AP's) As the 627's can output 45mA. |
|
|
| Upupa Epops |
| P. S. : ... plus switch for switching between rails - class A against +, class A against - :D . |
|
|
| TimA |
Qoute...''The OPAs, and most of the modern/fast op-amps, need a good layout and PSU bypassing. They are very picky. The fact that an NE5534 works better on a circuit doens't mean nothing to me...''
Carlos, correct me if I am wrong but the OPA627, which has a unity gain bandwidth of 16MHz, is hardly any faster than the NJM2114 (15MHz) or the NE5534 (10MHz). To put it another way the difference is one semi-tone (insignificant) and six semi-tones (hardly significant) respectively.
Sunsun 22.
In my experience both the OPA2604 and the AD823 benefit from a 2mA c.c.s. on their output. I have not found an op-amp that doesn't...yet.
Tim. |
|
|
| PMA |
| quote: | Originally posted by TimA
Qoute...''The OPAs, and most of the modern/fast op-amps, need a good layout and PSU bypassing. They are very picky. The fact that an NE5534 works better on a circuit doens't mean nothing to me...''
Tim. |
Untrue. OPA627, OPA134 are not any fast and do not need special care. What you are saying is true for AD8065/8066 etc.
BTW - 5534/5532 is a horrible amp concerning sonic results. |
|
|
| Dr.H |
Carlosfm, PMA,
Gentlemen, it appears that I have been misunderstood, but then I may not have expressed myself clearly.
Your very positive findings on class A biasing using a simple resistor is a well understood and accepted mod for most op-amps. And I can confirm that I have tried it as well on the humble OPA2604.
I was simply, in the name of progress, suggesting the reading on the website to show that, when using the simple R, modulation of the op-amps output can occur, that some of the benefits of the high input impedance of the buffer can be negated, etc.
And so, I was suggesting that as good as resistor-based biasing is (and there is no doubt at all that it brings meaningful sonic benefits), there are better methods, like JFET for example.
So, I agree, R biasing is good, but if you liked that, you owe it to yourself to try JFETS, I think you'll be happier with the sound, whatever the theory.
Yes Carlosfm, I agree that I refer to the use of the THS4061 in a specific application, but in fairness remember that I have posted my impressions of the same op-amp in the very pre that made all gaincloners so happy and I came to the conclusion that in my specific set-up, I preferred THS4061 over OPA627 in the pre.
Of course I will be posting my results on the biased THS4061...
Happy DIYing and lets continue sharing impressions, opinions, experiences etc. That's what makes this a good place.
Ryan |
|
|
| carlmart |
| quote: | Originally posted by Dr.H
I was simply, in the name of progress, suggesting the reading on the website to show that, when using the simple R, modulation of the op-amps output can occur, that some of the benefits of the high input impedance of the buffer can be negated, etc.
And so, I was suggesting that as good as resistor-based biasing is (and there is no doubt at all that it brings meaningful sonic benefits), there are better methods, like JFET for example.
So, I agree, R biasing is good, but if you liked that, you owe it to yourself to try JFETS, I think you'll be happier with the sound, whatever the theory.
|
Sorry, Ryan, but we better look for a real light at the end of our audio tunnel, because better specs (modulation, distortion) seem not to be in the same field as good audio results. By audio results I mean sounds which are more life like, instruments that sound natural and not amplified.
Constant current sources seem to be contradictory, perhaps very much like current mirrors. Specs look better, but sounds worst. So what do you go for?
The best compromise seems to be to listen to it and see how it sounds. Using a FET with a resistor is easy to arrange, to anyone who is willing to try. But a resistor is so simple!...
Carlos E. Martinez |
|
|
| TimA |
PMA
Please get your quotes from the correct source!
We have both made a similar point regarding OPA op-amp bandwidth, yet you have incorrectly attributed a qoute to me that you should have attributed to Carlos.
Tim. |
|
|
| carlosfm |
| quote: | Originally posted by TimA
Carlos, correct me if I am wrong but the OPA627, which has a unity gain bandwidth of 16MHz, is hardly any faster than the NJM2114 (15MHz) or the NE5534 (10MHz). To put it another way the difference is one semi-tone (insignificant) and six semi-tones (hardly significant) respectively.
|
Check slew rate too, and settling time.
Anyway, my point and my experience is: put these op-amps on, for instance, a cdp or a pre, and you won't be impressed.
Because most commercial products have miserable layout and psu bypassing.
Put some caps under the circuit, where they are needed, and it sings.
AD op-amps are not so picky, im my experience.
And the NE5532/4 works everywhere, even with psu caps at 5 cm distance. :clown:
Why do you think manufacturers like and use them so much?
Less parts needed!
| quote: | Originally posted by PMA
Untrue. OPA627, OPA134 are not any fast and do not need special care. What you are saying is true for AD8065/8066 etc.
BTW - 5534/5532 is a horrible amp concerning sonic results. |
The OPAs NEED special care.
I'm talking PSU bypass here.
The OPA627 and OPA134 will have overblown bass that mucks up everything on a bad circuit.
Then some people don't like it...:rolleyes:
The only BB OPA that I know and tested that works everywhere without special care, like the NEs, is the OPA227 and 228.
| quote: | Originally posted by PMA
BTW - 5534/5532 is a horrible amp concerning sonic results. |
Oh yes, measures well, not bad spects, etc. but sounds like:dead: .
My oppinion too.:angel: |
|
|
| Pan2 |
| quote: | Originally posted by Upupa Epops
To Pan : If you will have several amps, which every will have " tail ", you hardly recognize it, 'cos you will think, that is normal. But if you once will hear amp without tail, you will know this forever. Unfortunately absolutely most of amp tail have. |
I understand what you mean. Was pussled for a while though since I use the term "tail" most often regarding lowlevel reverberation and maybe fading of acoustic instruments.
No tails (by your definition) in my home though, no opamps, no feedback loops. Balanced class A single end in SACD and class A power amps with only metalfilm shunt attenuation in between. BW DC-1MHz in amps.
/Peter |
|
|
| carlosfm |
| quote: | Originally posted by Dr.H
Yes Carlosfm, I agree that I refer to the use of the THS4061 in a specific application, but in fairness remember that I have posted my impressions of the same op-amp in the very pre that made all gaincloners so happy and I came to the conclusion that in my specific set-up, I preferred THS4061 over OPA627 in the pre.
Ryan |
Ryan, I haven't tested the THS, so I can't comment.
I have them here to make a test.
With those beautiful specs I have a feeling...
It may even be that working with the BUF634 they outperform the OPA627.:eek:
And alone?
Possibilities are plenty, but I have no time now.:bawling: |
|
|
| carlosfm |
| quote: | Originally posted by Upupa Epops
Guys, whatabout trimpot - small class A, middle class A, high class A, overdimensioned class A etc. :D . |
A double rotary switch with several resistors?:D
| quote: | Originally posted by Upupa Epops
P. S. : ... plus switch for switching between rails - class A against +, class A against - :D . |
:D :cool: |
|
|
| mandat |
Are there some fundamental differences between OPA627 and OPA637 ICs (excluding SR and stability for gain=1)? Because one of experienced audio designer has told that OPA627 is a completely misunderstanding in sonic quality terms, while OPA637 is quite qood.
There are some coincidence between increasing an A class biasing current for opamps like AD847, AD825 and LT1362 and its "musicality" (information from the web). From the same source - the compensating current for AD825 should be 2.4 mA. |
|
|
| Upupa Epops |
| How I see, by meaning most of yours, manufacturers of audio integrated amps are idiots. Whatabout to make any public inquiry and by results let to make an realy " musical " one ? This neverending discussions are still rather tedious :xeye: . |
|
|
| PMA |
Well - I have simulated the OPA627 - an output stage and preceding BJT LTP with current mirror and current source. I did not include input cascode. I agree that my simulation will be quite far from the real circuit.
But one thing can be seen clearly. Loading the output against -Vs by resistor changes the mode of operation of the output stage. Now only the upper transistor is working, the lower one having negligible current. The output stage works as an SE class A one transistor output stage with emitter load resistor. The conditions of the output stage are changed and an investigation of the new output stage drive capability should be done. |
|
|
| PMA |
| quote: | Originally posted by Upupa Epops
How I see, by meaning most of yours, manufacturers of audio integrated amps are idiots. Whatabout to make any public inquiry and by results let to make an realy " musical " one ? This neverending discussions are still rather tedious :xeye: . |
These opamps are not optimized for audio use, but for instrumentation and measuring purposes. |
|
|
| Nuuk |
We post our opinions of certain opamps, in certain circuits, in certain pieces of equipment, here and on other forums.
At best we can hope to give others a suggestion of something to try for themselves but we can never give a definitive recommendation of which will be the best opamp for somebody else.
Most opamps are not too expensive and by using a socket, it is not too difficult to try different types (and different circuits). What is pointless is arguing that one opamps is better than another, or one circuit is better than another.
As Carlos says about his class A resistor mod - try it and see for yourself and I would add that goes for opamp make and model as well. ;)
Even describing an opamp as musical is open to interpretation as 'musical' means different things to different people. I am a little worried that newbies who read threads like this one try to decide on which opamp to use based on what is written, rather than on what they could hear if they tried the opamp for themselves.
Fashions abound in this game but the wise hi-fi builder will not be swayed by them. Never discount an opamp because it is either cheap or because it has been around a long time. You may be ruling out your 'perfect opamp!
We all know (or should know) about system matching by now. I am sure that the same rules apply to choosing an opamp. Much will depend on all the other factors in the system.
Finally, let's not forget that many opamps were never designed for the purpose of reproducing music (although some may do that job well) Having said that, some of those that make that claim don't seem to do a very good job, in my experience. |
|
|
| sss |
| quote: | Originally posted by Nuuk
I am a little worried that newbies who read threads like this one try to decide on which opamp to use based on what is written, rather than on what they could hear if they tried the opamp for themselves.
Fashions abound in this game but the wise hi-fi builder will not be swayed by them. Never discount an opamp because it is either cheap or because it has been around a long time. You may be ruling out your 'perfect opamp!
|
Nuuk i totally agree with u
guys dont forget that before the music is recorded on a cd it goes through 50 NE5532 op amps , even today most pro audio equipment got those . |
|
|
| carlosfm |
Guys,
I opened this thread to share with you all a trick that I always use on the OPA627.
Yes, this thread is about the OPA627.
The op-amp I love, the one I always use, when possible.
The trick is very easy to try, I just wished that all the tweaks were so easy as throwing a resistor into a circuit.:bawling:
This is, I think, valuable information for who uses the 627, and even those that don't, with a socket they can easily test the 627 again, with the resistor, and listen .
Even those who tried to bias the 627 into class-A with other methods and/or values can try again, with the method I describe here.
Sometimes I also ask myself why is it so complicated to open a thread here and share information.:confused:
It usually degenerates into a nonsense and endless discussion.:bawling:
Just try it if you want, listen and tell us what you think.:angel:
Like Pavel did. |
|
|
| Franz G |
I inserted this 2k resistors 30 minutes ago, and actually Marla Glen is singing like never before!
It is an interesting hint, Carlos! Would be great, to drive a tube with this OPA627...
As for different opamp types: I tried NE5534AP (maybe 18V is a little bit to high for this chip?), OPA134 and OPA627 in my amp and my favourite is the OPA627.
The NE5534AP, Philips, where second choice for me, OPA134 third and NE5534AP Ti are at the end.
BUT: I am very happy with my NAD Phono Preamp PP 2 and I opened it. What is in there: an NE5532!
So, it depends really on the circuit and other factors, wich opamp you like.
Just try, for every new circuit!
My two cents
Franz
BTW: my OPA627 are still sitting in sockets, but no problem! I bypassed them at the socket with 2x0,1uF foil, 1uF Tantal and 0,15uF foil between +Vs and -Vs. |
|
|
| carlosfm |
| quote: | Originally posted by Franz G
I inserted this 2k resistors 30 minutes ago, and actually Marla Glen is singing like never before! |
:D
| quote: | Originally posted by Franz G
BUT: I am very happy with my NAD Phono Preamp PP 2 and I opened it. What is in there: an NE5532! |
Change that for an OPA2228.
Maby you like it even more.;)
Or maby not...:rolleyes:
| quote: | Originally posted by Franz G
It is an interesting hint, Carlos! |
I've bee fiddling again with my amp and another hint is on it's way, on a new thread.
It's an easy one, like this!:D
Stay tuned.;) |
|
|
| Franz G |
Carlos
The sound is now really good, but I want to see the results on my scope.
What would happen, when we do the same for the LM3875? Just bias to a certain level, so you have class A up to 1 or 2 watt, what is enough for my speakers.
I think one of the bottleneck of all LM38xx amps are the quasi symetrical outputstages. So, you could add the best buffer, but this problem resists.
Franz |
|
|
| carlosfm |
If you like the sound, I hope the scope doesn't make you go back.:rolleyes:
You can try the same on the LM3875, I never did.
But you may need big heatsinks.
I don't know if it makes a difference, I think I've read somewhere that it doesn't. |
|
|
| PMA |
| I have checked the OPA627+resistor against -Vs yesterday on the scope. There was no problem. |
|
|
| PMA |
| quote: | Originally posted by Franz G
Carlos
What would happen, when we do the same for the LM3875? Just bias to a certain level, so you have class A up to 1 or 2 watt, what is enough for my speakers.
I think one of the bottleneck of all LM38xx amps are the quasi symetrical outputstages. So, you could add the best buffer, but this problem resists.
Franz |
You are right. The output circuit of the LM3875 considerably affects the sound, especially when compared with a good class A amplifier. |
|
|
| Joe Rasmussen |
I also would like to share a little 'secret' of mine when it comes to the OPA-627. Especially since on this thread there has been several references re the need for good decoupling etc.
Her is my trick:
What you are seeing here is a little difficult to describe. Essentially this idea applies to all opamp ICs - even LM3875 et al. But this trick works a treat on the OPA-627.
The key here is the signal return. But more importantly, it is the return signal current.
In many opamps only one of the caps shown is needed. In most cases it is the 'neg' one, but not always. If memory serves me right, the OPA-627 requires the positive cap. Best to add both and if using ICs in plug'n'play comparisons, you are covered.
Which cap is necessary actually depends on the internal comp cap, whether it has a neg or positive voltage across it. This indicates whether an IC is referenced to which voltage rail. Think about it, the IC does not have a ground terminal, right? Exactly! So where does it get its signal ground reference.... from one of either rails.
Now it may not become immediately apparent, but think about : Does it not make sense that the signal currents return back to this point via the decoupling cap shown above? It kinda does.
This is a bit simple, it gets rather more complicated as the open loop gain falls with increasing frequency we lose power supply rejection and now unwanted (disturbed) HF signals can appear right across the comp cap. Does it affect the sound? You betcha, and not to mention stability.
So analyse how you are using the 627 (and others too, but I agree the 627 is special IMHO), we know where the signal (and its currents) are going out - but where do the return?
Bring 'em home. :)
Joe R. |
|
|
| carlosfm |
Hi Joe, welcome!:cool:
It makes sense, very nice.
What caps do you use there, for instance, on the OPA627?
I suppose it would not be the same for the 637?
Do you notice any improvement in sound by doing this?
Thanks.:angel: |
|
|
| carlosfm |
Mmmm...
Joe, why not a pair of resistors instead of caps to make a virtual ground for the signal return? |
|
|
| analog_sa |
| quote: | | Think about it, the IC does not have a ground terminal, right? Exactly! So where does it get its signal ground reference.... from one of either rails. |
I seem to feel lost with the above statement. Don't the inputs provide ground reference? What has the PS to do with that? And how are the shown caps different to any othe PS bypassing caps? |
|
|
| PMA |
| Just separate signal return of the input non-inverting input resistor, small stabilizers (LM317L etc.), bypass capacitors etc. to separate PCB traces and connect it into just one point of the PCB - that's my way. No problems with ground induced noise. Left and right channel have separate power supplies, no ground loops. |
|
|
| analog_sa |
| quote: | | Just separate signal return of the input non-inverting input resistor, small stabilizers (LM317L etc.), bypass capacitors etc. to separate PCB traces and connect it into just one point of the PCB - that's my way. No problems with ground induced noise. Left and right channel have separate power supplies, no ground loops. |
Agreed. But these are just normal grounding techniques. Nothing groundbreaking :) |
|
|
| PMA |
| quote: | Originally posted by analog_sa
Agreed. But these are just normal grounding techniques. Nothing groundbreaking :) |
Agreed ;) . Haven't said any word about groundbreaking ;) . |
|
|
| Franz G |
This is my "real" gainclone, e.g. noninverted but with (to?) big bypass caps:
2 x 18000uF (out of a Fostex PA-amp)
4 x 10000uF (Jamicon)
4 x 2200uF (Philips)
4 x 1 uF foil Siemens
4 x 0,1uF foil Wima
The first impression is impressive :D
First, I used a trafo out of an Kenwood KA-1500, result: Smoke. With a 225VA torroid, 2x22V, it works fine and cool.
But give me some time. I will compare soon with the regulated amp and the tube buffered versions, together with some friends. And then write about.
Franz
P.S.
One result is clear: for subwoofers use BIG caps! |
|
|
| aparatusonitus |
| quote: | Originally posted by PMA
Just separate signal return of the input non-inverting input resistor, small stabilizers (LM317L etc.), bypass capacitors etc. to separate PCB traces and connect it into just one point of the PCB - that's my way. No problems with ground induced noise. Left and right channel have separate power supplies, no ground loops. |
I have seen this in same Philips cd players, if I remember correctly.
Could you post same shematics about this arrangement, please.
Regards, Ivo |
|
|
| Franz G |
Sorry Folks, I puted the last reply in this thread: but actually I switched back to my regulated, OPA627 Class-A buffered amp:
This is music, not the NIGC with BIG caps!
So, i will reduce the caps and listen to it.
In theory, the best buffer must be no buffer (not true for the inverted version, I know).
Franz |
|
|
| matjans |
I'm not sure i get it. Are you creating a virtual ground especially for signal return *in addition* to normal supply bypassing or is it 'just' a ground layout tip?
/m |
|
|
| Joe Rasmussen |
| quote: | Originally posted by carlosfm
Hi Joe, welcome!:cool:
It makes sense, very nice.
What caps do you use there, for instance, on the OPA627?
I suppose it would not be the same for the 637?
Do you notice any improvement in sound by doing this?
Thanks.:angel: |
That depends on what the previous implementation was like (and I wouldn't be able to count the different possibles). If it was pour to start with and THEN this idea was used - just add two film caps, bring the signal return back to the junction of the two before it sees ground - THEN it could make a significant improvement in sound.
Re 627 vs 637, yes, the same applies. But the 627 has a more aggressive internal com cap. This means less open loop gain at HF. This means less PSR at HF and keeping voltage changes/modulations away from the comp cap that just bit more difficult. That would be my logic.
| quote: |
Mmmm...
Joe, why not a pair of resistors instead of caps to make a virtual ground for the signal return? |
Need something that has reactance going down with increasing frequency i.e. a cap. The value of the cap? It ought to be larger if the load Z is low value and vice versa. It must be film as it needs MHz plus performance.
A 0.1uF is 16 Ohm @ 100KHz. So that should suffice for signal ICs, but a little higher value wouldn't go astray...
But this will not do with a gainclone (power) chip and 8 Ohm Load. That would require 4.7uF film or similar. This will return HF currents which now are much higher. These should be on the chips' pins and NOT across the larger remoter reservoir caps.
Or...
Now let us revise the above idea to a gainclone chip and Low Z Speaker Load:
Since most of us are likely using Panasonic 1000uF/50V electros, which have good HF performance, low ESR etc. So they do a good job and 1000uF is enough to deal with 4-8 Ohm Loads.
But the bottom line is the same. Where do you return the now much higher power return currents? The problem is the same even if the scale is different.
Now we can also see WHY the 1000uF caps needs to be close to plus and neg on the chip. In fact right on the pins.
The speaker return? Bring it right back to the chip and those caps.
Joe R. |
|
|
| Joe Rasmussen |
| quote: | Originally posted by matjans
I'm not sure i get it. Are you creating a virtual ground especially for signal return *in addition* to normal supply bypassing or is it 'just' a ground layout tip?
/m |
:scratch: :scratch: :scratch: :scratch:
It's gets complicated when we get beyond just one chip - all I can say is you need to analyse what's happening and be aware of the return HF currents getting back to the chip via the rail it uses as its signal ground for those HF currents.
Like I said, this is not an easy subject to explain but when dealing with it you must try to analyse and find ways to deal with it. Even very experienced techs and designers have found themselves scratching their heads on this, especially in more extended multiple IC topologies. And they may not even be dealing with sonic qualities, just trying to keep things stable. "Where is that weird oscillation pattern coming from?"
The concept of 'ground' is a fleeting subject.
There is no universal ground.
For example: Grounds can lie.
NEVER TRUST A 'GROUND' !!! TREAT ALL WITH SUPREME SUSPICION.
Really, when you think of it, I ain't got any dirt in my amps, so there are NO grounds in any of my amps. :cannotbe:
Grounds are common connections to which signals and power are referred or referenced. Nothing more than that. That is textbook.
Grounds do not LOCK DOWN signals or currents. Grounds can move.
As for a 'layout tip' - you are right. I can say that for single chip situations, you can consider it thus. And that was my idea as a 'trick' to improve the 627, which I like (if it has to be an IC). Apply it to the gainclone chip too. I bet that the original Gaincard has the speaker return coming back to the electro cap x 2 junction.
Joe R. |
|
|
| carlosfm |
| This is getting interesting.:D :cool: |
|
|
| carlosfm |
| quote: | Originally posted by Franz G
Sorry Folks, I puted the last reply in this thread: but actually I switched back to my regulated, OPA627 Class-A buffered amp:
This is music, not the NIGC with BIG caps!
Franz |
Use those big caps on the URGC only for a sub amp, Franz. |
|
|
| matjans |
i think i'm sort of getting the point here. looking at a 'ground' purely as a current return path and/or point of reference explains many, at first sight, complicated (not only stability) problems.
Non EE people (like me) tend to forget/overlook this pretty often. Looking at a preamp pcb i am designing i see a number of stupid layout errors already... :mad:
/m |
|
|
| carlosfm |
| quote: | Originally posted by matjans
Looking at a preamp pcb i am designing i see a number of stupid layout errors already... :mad:
/m |
If you're designing, you're still in time to fix it.:D
If it's op-amps, don't forget to include a place for a resistor from output to V-!:cool:
Joe's trick seams easy to adap without too many changes on an existing circuit, but on an already tight layout I guess those caps will be floating with a leg in the air.:xeye: |
|
|
| matjans |
| quote: | Originally posted by Joe Rasmussen
I bet that the original Gaincard has the speaker return coming back to the electro cap x 2 junction.
Joe R. |
Don't know about the original gaincard but the 47labs 4717 shigaraki seems to have it that way:
follow the green wires from the black speaker post to the space in between the two caps. Pictures borrowed from this dutch review. in the review the pics are clickable which will give you higher res pics.
stupid Q : carlos, wont't you get a negative dc offset by putting a resistor from v- to output? |
|
|
| PMA |
| quote: | Originally posted by Joe Rasmussen
Re 627 vs 637, yes, the same applies. But the 627 has a more aggressive internal com cap. This means less open loop gain at HF. This means less PSR at HF and keeping voltage changes/modulations away from the comp cap that just bit more difficult. That would be my logic.
|
Do not want to be much offensive, but according to Burr-Brown's datasheet the 637 has much worse +Vs PSRR than the 627, even for high frequencies (datasheet, page 5, PSR vs frequency curves). The difference is nearly 20dB.
Please take notice that +Vs PSRR is almost always different than -Vs PSRR for every opamp. |
|
|
| PMA |
| quote: | Originally posted by matjans
Don't know about the original gaincard but the 47labs 4717 shigaraki seems to have it that way:
|
matjans,
these are horrible pictures. Are they 5 years old children to assembly and solder this??
I was thinking that we are a bit serious here. |
|
|
| Nuuk |
I may have misunderstood what Joe is saying here but I thought the return wire from the speaker always went to the junction of the two 1000 uF caps anyway (as shown below where the thicker blue wire is the one in question). :xeye:
 |
|
|
| matjans |
| pma: i'm sorry but these are pictures of the insides of the real, actual 4717. ;) :bigeyes: i wasn't joking |
|
|
| PMA |
| quote: | Originally posted by Joe Rasmussen
It's gets complicated when we get beyond just one chip - all I can say is you need to analyse what's happening and be aware of the return HF currents getting back to the chip via the rail it uses as its signal ground for those HF currents.
The concept of 'ground' is a fleeting subject.
There is no universal ground.
For example: Grounds can lie.
NEVER TRUST A 'GROUND' !!! TREAT ALL WITH SUPREME SUSPICION.
Grounds do not LOCK DOWN signals or currents. Grounds can move.
Joe R. |
Joe,
how about using groundplanes even for audio circuits. I have the best experience in a preamp. |
|
|
| Joe Rasmussen |
| quote: | Originally posted by PMA
Do not want to be much offensive, but according to Burr-Brown's datasheet the 637 has much worse +Vs PSRR than the 627, even for high frequencies (datasheet, page 5, PSR vs frequency curves). The difference is nearly 20dB.
Please take notice that +Vs PSRR is almost always different than -Vs PSRR for every opamp. |
Point taken. I was merely supposing. I didn't check the spec sheet, so I got it back to front. But in either case, the principle is till the same, PSR deteriorates at HF and less able to suppress HF return currents. This is one area where feedback does a good job but also needs a bit of help.
| quote: |
how about using groundplanes even for audio circuits. I have the best experience in a preamp. |
Lot to be said for groundplanes and we have used them at VSEI. But they shouldn't, as I'm sure you would agree, be viewed as a perfect cure but as one to put in one's armaments. But in later years we have preferred in straight audio (AF) circuits to analyse loops, then find the best point to reference and 'tap' that point via a wire (or even better a foil or 300x0.1mm tin stranded copper) that has zilch inductance etc., then make sure that neither AC or DC (ideally) actually flows through this connection. If you do it right, everything forms loops that are stable.
But as always, everybody has to decide what or how they feel best able to deal with it. But having a suspicious mind is a good thing, IMHO.
Joe R. |
|
|
| Joe Rasmussen |
| quote: | Originally posted by Nuuk
I may have misunderstood what Joe is saying here but I thought the return wire from the speaker always went to the junction of the two 1000 uF caps anyway (as shown below where the thicker blue wire is the one in question). :xeye:
[/IMG] |
You are right Nick. This is good practice, but more than that, as I'm trying to illustrate. By having those caps right on the chip V+ and V- they now perform an additional task. Look at a lot of amps where big reservoir caps are many inches away. In ALL opamp ICs this becomes much more critical than discrete amps. It has to be because of the internal compensation used by all opamps. The way it's done means that one of the rails have to become a 'ground.' Normal grounding is bad enough, but in opamps... yikes!
The beauty of those 1000uF caps being so close to the 3878 power in pins, they now perform an additional task that normally reservoir caps/electros don't. It greatly simplifies critical opamp grounding. I use Panasonic, many others do as well. Does Sakura/47 Labs use 'naked' Black Gates or something? If they do, these are very good electro caps with good HF performance. Now HF return currents are ably dealt with and no additional components required. On the other hand, if we had remote reservoir caps, we would have to put our thinking caps on (the other kind of cap - or should I say 'hat').
BTW, the 3875 spec sheet shows clearly the internal comp cap has a negative' voltage across it, thus the V- is our pseudo ground, or Clayton's Ground, the ground you have when you don't have a ground.
Or as Forest Gump would have said: Ground is what Ground does. ;)
Those internal pics are eye-openers. Gosh, I can imagine what Rod E. would be thinking. The school-master in him would cause him to deeply frown.
"Rip-off."
Now when a large company gets rid of its CEO because of poor performance, and gives him 10 million dollar handshake, that's what I call a real rip-off! :headshot:
Joe R. |
|
|
| garbage |
| quote: | Originally posted by PMA
OK - I have done it. It does not work for every opamp. But it does work for the OPA627.
My application is the voltage stage in my new amplifier project (error correction amp). The opamp is supplied from +/-18.5V (LM117+LM337). I have chosen the 2.7k resistor from the opamp output to -Vs.
For OPA627, the improvement is substantial. The effect is in clearer mids and highs and the overall resolution for the complex music. The music is more liquid. And the "tail" dissappears. Absolutely positive, nothing to complain. Only very very well designed discrete circuit is capable to give same sound quality as OPA627 with increased Iq.
The biasing against -Vs does not work for AD797 and brings no effect for OPA134. |
Off topic, but PMA, can this be done using your pcb?
Can i just put a 1800R resistor across pins 2 and 6 of the opa627 (if i am using +/-15v)?
Rgds |
|
|
| PMA |
V.,
As you have probably noticed, I have not tried it for audio buffer circuit, but power amp voltage amplifying stage. There is a considerable difference, in audio buffer the opamp is not loaded (high input impedance of the BUF634T) and the opamp output stage class is not an issue that much.
But - there is no technical obstacle to try it. Please feel free to make an experiment and let us know. You do not need to be afraid of any technical problem in your PCB.
Cheers, Pavel |
|
|
| ABO |
| quote: | Originally posted by garbage
Off topic, but PMA, can this be done using your pcb?
Can i just put a 1800R resistor across pins 2 and 6 of the opa627 (if i am using +/-15v)?
Rgds |
Make sure you put the resistor across pins 4 and 6, not 2 and 6. |
|
|
| PMA |
| quote: | Originally posted by ABO
Make sure you put the resistor across pins 4 and 6, not 2 and 6. |
Thanks, of course. The resistor must be connected between the output and -Vs. |
|
|
| carlosfm |
I must say that I don't have biasing into class-A on my OPA627+BUF634 pre, because the buffer makes part of the work and I never thought it was necessary, even sonically speaking.
I don't open my pre for a long time, I've been busy with other things.
I will be trying soon other chips and implementations, but completely different from this pre.
If they prove superior, I'll change the pre.:D |
|
|
| PMA |
| I have finally simulated the whole OPA627 circuitry and can confirm that resistor from output to -Vs changes the mode of operation of the output stage so that only upper transistor works, in SE class A. This is true until load does not suck that much current in negative polarity that lower transistor must open and deliver it. That means that proper value of biasing resistor depends on load resistor, heavier load needs lower value of biasing resistor. But do not overheat the chip. |
|
|
| Nuuk |
Slightly OT but can anyone identify this chip for me?
A search on Google doesn't bring anything up. I would guess it is a dual but I found three of them on the output of an old Pioneer CDP. |
|
|
| PMA |
It is a standard dual opamp. Hope that might be replaced by OPA2134.
The 5238 were used as I/V stages of CD players (in Sony XB930, e.g.). |
|
|
| Nuuk |
| Thanks PMA. I want something to use in an active crossover circuit so I will try them. |
|
|
| garbage |
| quote: | Originally posted by PMA
V.,
As you have probably noticed, I have not tried it for audio buffer circuit, but power amp voltage amplifying stage. There is a considerable difference, in audio buffer the opamp is not loaded (high input impedance of the BUF634T) and the opamp output stage class is not an issue that much.
But - there is no technical obstacle to try it. Please feel free to make an experiment and let us know. You do not need to be afraid of any technical problem in your PCB.
Cheers, Pavel |
Experiments with opa627/buf634:
My opa627 are on sockets, so I was able to plug the 1800R in and out. I A/B it for about 4 times. Cannot really tell the difference.
Diana Krall's voice seemed to be focused with the resistors in. Not too sure. It's not a big impact, maybe very subtle.
I played Temptation from The girl in the other room album. |
|
|
| Franz G |
I changed my mind and switched back to normal operation mode, in the meantime.
Why? Because this chip ist not designed (and evaluated) to be a single ended class a amp. There is to much asymetry and may be the working point of the single output transistor is not optimal.
This is not true class A!
And: they get to hot for my taste.
Franz |
|
|
| PMA |
[QUOTE]Originally posted by Franz G
This is not true class A!
And: they get to hot for my taste.
Franz [/QUOTE
The mode of operation depends only on ratio of load resistor and biasing resistor. In case you load it by 10k input impedance of the following amplifier and you use 2.7k resistor from output to -Vs, I can assure you that it is a true class A and that there is no problem with power dissipation. This is not a question of feeling, but circuit technology. |
|
|
| carlosfm |
I have good results from around 5 to 10ma biasing and that's what I recommend.
I gave some standard values for the resistor that you can test, at close to 10ma, but you can test other values too.
From 5 to 10ma there's some values you can test.
I wouldn't pass 10ma biasing with the 627.
Franz, what's important is what sounds better for you.
You seam to be biased (no pun intended) by theory.
If unbiased sounds better for you on the circuit you are testing, that's ok, but don't quit at first try.
Biasing an op-amp in class-A in most cases makes it operate in single-ended, I think that it doesn't happen only with the 627.
Another thing: the 627 alone, biased at around 10ma doesn't get hotter than the OPA627+BUF634 in wide BW mode, with +/-18v psu.
It's a little hot, but nothing special.
You can touch it with your finger for the time you want, it's not really hot.
No problems here. |
|
|
| garbage |
| quote: | Originally posted by carlosfm
Another thing: the 627 alone, biased at around 10ma doesn't get hotter than the OPA627+BUF634 in wide BW mode, with +/-18v psu.
It's a little hot, but nothing special.
You can touch it with your finger for the time you want, it's not really hot.
No problems here. |
that is true, i checked the temperature of the opa627 with my fingers with the 1800R resistor (at 8mA plus) . it is only slightly warmer than without the resistor(no heatsink will be required). the buf634 is much hotter in comparison, even with heatsink. |
|
|
| Nuuk |
| quote: | | Diana Krall's voice seemed to be focused with the resistors in. Not too sure. |
I think you are referring to the album 'The Girl In The Other Room'! But on some of her other recordings, I have found that her voice is not precisely located and I have assumed this is due to her voice being picked up by both her mike and the mike in the piano. |
|
|
| carlosfm |
Listen to Jacintha "Jacintha is her name" and in a while you'll be crying.:bawling:
:cool: |
|
|
| Franz G |
No, really folks: forget about class A, in this case!
It would be nice to have a single ended class A audio opamp, of course. And a single ended class A power opamp.
But you cannot change this chip in a true, balanced class A mode just by bypassing one output transistor!
And the circuit to reduce thermal feedback is out of order.
Please, forget it or like it! But maybe, you cheat yourself?
BTW: the NIGC, reduced caps (removed the 18000uF) and good bypassing seems to be better than an inverted, buffered, regulated version.
I am very sorry! My conclusion: the best buffer is no buffer. Thats the reason, not to use a reversed version, if there is a poti in the input.
Franz |
|
|
| carlosfm |
| quote: | Originally posted by Franz G
No, really folks: forget about class A, in this case! |
People can try, listen and decide.
In some few cases (in some circuits) it may not work so well, but in most cases it sounds better for me.
Are you recommending what sounds better for you?
| quote: | Originally posted by Franz G
BTW: the NIGC, reduced caps (removed the 18000uF) and good bypassing seems to be better than an inverted, buffered, regulated version. |
Franz, if it works better for you, fine.
How much capacitance per rail are you using now?
| quote: | Originally posted by Franz G
I am very sorry! |
You don't need to.:) |
|
|
| PMA |
| quote: | Originally posted by Franz G
No, really folks: forget about class A, in this case!
It would be nice to have a single ended class A audio opamp, of course. And a single ended class A power opamp.
But you cannot change this chip in a true, balanced class A mode just by bypassing one output transistor!
Franz |
You should bring any evidence for what you are saying, not only feelings. Me, I can show 2 images.
1st: this is an OPA627 circuit with NFB for gain +10. Rload = 1.6k, R bias = 2.7k. |
|
|
|
