Dear Sirs,
I understand that op-amps are all characterized by a wide closed loop bandwidth.
Speaking instead of open loop bandwidth which are the best op-amps in this regard ?
My "feeling" is that the opinion of Mr John Curl about the importance of keeping a high open loop bandwidth in audio amplification stages makes very much sense.
Is there "a op-amp champ" in this regards ?
Thank you very much indeed.
Kind regards,
beppe
I understand that op-amps are all characterized by a wide closed loop bandwidth.
Speaking instead of open loop bandwidth which are the best op-amps in this regard ?
My "feeling" is that the opinion of Mr John Curl about the importance of keeping a high open loop bandwidth in audio amplification stages makes very much sense.
Is there "a op-amp champ" in this regards ?
Thank you very much indeed.
Kind regards,
beppe
sawreyrw said:
Thank you very much Sir for your very kind and welcome reply.
First of all I would like to explain better.
I believe in John Curl and in what HE says.
He says that the open loop bandwidth must be wide on principle.
If the GBW is 10e6, the gain will be 1 at 10 MHZ
OK. But applying this formula, and if I don't understand wrongly, when the gain is made equal to the open loop gain then the bandwidth could be very small.
Much smaller than the audio bandwidth actually.
Mr Curl mentioned the AD825 as an example of op-amp with a good bandwidth even if operated open loop.
Unfortunetely the AD825 comes only in the smd package.
So it is not convenient for me.
I need a op-amp in mini-dip package.
If I am not wrong I read that also the cheap NE5534 is not bad in this regard.
I know that is used also in high-end equipments.
So very bad it should not be.
So I am trying to select a op-amp with a very wide open loop bandwidth.
I will be using it as a buffer (gain=1).
Thank you very much indeed.
Kind regards,
beppe
juergenk said:
Thank you very much indeed Mr Juergenk.
The NE5532 has a large open loop bandwith of app. 1kHz
So these op-amps would not be John Curl approved
If this principle is valid my question is now: do op-amps exist that have a open loop bandwith wide as the whole audio bandwidth ?
If not I would not feel comfortable about using any op-amp in the signal path.
Thank you very much again.
Kind regards,
beppe
P.S. By the way, what do you use as line preamp in your system ?
hi,
Maybe video-opamps have wider open loop bandwiths.
I´ve good experiance with OPA2134 and it has a poor open loop bandwith.
So maybe this specific criteria should not be overestimated.
Regards
Jürgen
currently I´m using a passive preamp eq. a 10k pot.
this is unusual with audio-opamps. The open loop bandwith is always smaller than audio-bandwith.
Maybe video-opamps have wider open loop bandwiths.
I´ve good experiance with OPA2134 and it has a poor open loop bandwith.
So maybe this specific criteria should not be overestimated.
Regards
Jürgen
Beppe - I have noticed you have asked some interesting questions in this forum before, and I think you're on the right track for making good audio. In answer to your question about opamps, you might try just looking (Googling I think might be the term here) at data sheets for yourself just to get a scope of the field of interest. There doesn't appear to be that many as opposed to the regular type of small BW units so they might stand out. Elantec and T.I. make some that look wide but I forget which ones right off hand. Models and numbers change pretty fast sometimes so you'd just have to look to see what's current. Video opamps are others of possible interest. Don't be afraid to venture out on your own there, it's what DIY is all about. I've had very good luck with such an approach.
NE5534 and dual NE5533
As usual
people judge NE5534 by using performance/data of NE5532.
They do not know that NE5534 ... IS NOT a single version of NE5532
They are 2 different op-amps
We who knows this fact, make a difference.
NE5534 is good!
A lot of people judge NE5534 performance
because they have used or tested NE5532 .....
This is unfair!
---------------------------------
OPA2134 is a dual version of OPA134.
But NE5532 is NOT a dual version of NE5534.
The dual version is called NE5533
This is from Philips datasheet:
Regards
lineup
beppe61 said:
juergenk said:
As usual
people judge NE5534 by using performance/data of NE5532.
They do not know that NE5534 ... IS NOT a single version of NE5532
They are 2 different op-amps
We who knows this fact, make a difference.
NE5534 is good!
A lot of people judge NE5534 performance
because they have used or tested NE5532 .....
This is unfair!
---------------------------------
OPA2134 is a dual version of OPA134.
But NE5532 is NOT a dual version of NE5534.
The dual version is called NE5533
This is from Philips datasheet:
Regards
lineup
It might be more beneficial to Beppe to keep this thread concerned with wide open loop BW units as per his request, as I'm sure there are other threads he could reference if needed for other opamps.
Beppe, I also recall seeing something about an opamp that is a "diamond" buffer, or some such name. You might try a search on that.
Beppe, I also recall seeing something about an opamp that is a "diamond" buffer, or some such name. You might try a search on that.
RetroAudio said:
Thank you so much Sir for your kind words.
It is sometime that I am trying to learn the basics about audio electronics but I have to say it is a very demanding task for me.
I usually read with a lot of interest the article of tha most famous designers in the field, as Mr John Curl in this case.
After reading how is important this "open loop bandwidth - let's say OLB for convenience" and knowing how can be convenient to use op-amps (I played with some of them in the past with so so results) I search for this parameter (OLB).
I found nothing for it on the op-amps data sheets.
A so important parameter not found on data sheet
Strange at least, isn't it ?
Sometimes, but not always, I can see this gain bandwidth product - GBP .
Dividing GBP byOLG I get OLB, that I understand it is usually very limited.
This is not good for audio purpose, as Mr Curl says.
Then noticing how preamp based on op-amps often fail to convince on a sonic basis I think he is right.
Elantec and T.I. make some that look wide but I forget which ones right off hand
This is extremely interesting and I thank you for the hint.
Please excuse me if I trivialize the subject but if a goal is to get a wide OLB why do not build op-amps with low OLG ?
If the GPB is kept sufficiently high a low OLG will give as a consequence a very high OLB ?
Maybe I am saying a lot of rubbish.
But I would be extremely interested to get opinion on this.
The convenience to get low OLG op-amps.
Thank you very much again.
Kind regards,
beppe
Re: NE5534 and dual NE5533
Thank you very much Sir.
You state: NE5534 is good!
But how is its bandwidth when operated open loop ?
I mean, we have to fix some principle after all.
Some sort of design requirements.
An open loop bandwidth wide at least as the audio bandwidth it seems to me a reasonable one, all considered.
I think this is an interesting point.
Thank you very much for your explanation.
Kind regards,
beppe
lineup said:
As usual people judge NE5534 by using performance/data of NE5532.
They do not know that NE5534 ... IS NOT a single version of NE5532
They are 2 different op-amps
We who knows this fact, make a difference.
NE5534 is good!
A lot of people judge NE5534 performance
because they have used or tested NE5532 .....
This is unfair!
---------------------------------
OPA2134 is a dual version of OPA134.
But NE5532 is NOT a dual version of NE5534.
The dual version is called NE5533
This is from Philips datasheet:
Regards
lineup
Thank you very much Sir.
You state: NE5534 is good!
But how is its bandwidth when operated open loop ?
I mean, we have to fix some principle after all.
Some sort of design requirements.
An open loop bandwidth wide at least as the audio bandwidth it seems to me a reasonable one, all considered.
I think this is an interesting point.
Thank you very much for your explanation.
Kind regards,
beppe
Hi beppe,
the open loop bandwith is defined by the -3dB point in open loop gain vs. frequency response.
You could extract it yourself from the published figures.
Designers generally don´t care for it. So it is not seperately specified.
It is more important that the amplifier is stable and has enough gain at a specific frequency to give the necessary feedback.
I don´t think a opamp should be judged on its OLB alone.
The comparison between NE5532 and OPA2134 is a good example.
Regards
Jürgen
the open loop bandwith is defined by the -3dB point in open loop gain vs. frequency response.
You could extract it yourself from the published figures.
Designers generally don´t care for it. So it is not seperately specified.
It is more important that the amplifier is stable and has enough gain at a specific frequency to give the necessary feedback.
I don´t think a opamp should be judged on its OLB alone.
The comparison between NE5532 and OPA2134 is a good example.
have a look at the datasheet. It is about 1kHz. Roughly the same as NE5532, when it is compensated for unity gain. A little bit more when it is uncompensated.
Regards
Jürgen
Beppe - most data sheets do indeed have some sort of gain -bandwidth related graph/specs, so I'm not sure what you have been referencing. It is true you may not always get what you want, such as the open loop graph, but that is where reading other specs can help out. It's also true low gain often does go with attaining the large BW, since it all does revolve around that GBW product. Not totally sure if that is what you are getting at or what. Perhaps if you read a data sheet on a wide open loop BW opamp you'd see that they often have low gain as well and would answer any questions you may have. The 825 may be such an opamp, so why not check it out? I have found you can learn much not by just looking at one, but by comparing many. Try looking at the National Semiconductor LM6171 and Elantec 2044 opamps. Hope this helps.
juergenk said:
the open loop bandwith is defined by the -3dB point in open loop gain vs. frequency response
Thank you very much Mr Jurgen, very valuable advice.
This is what I needed to start looking in the data sheets.
I don´t think a opamp should be judged on its OLB alone
Of course not. I agree. Only when is intended to be used in an audio application
(joke ! but not completely).Thank you very much again.
Kind regards,
beppe
You are looking for this kind of graph plot:
Note how the gain drops rapidly with frequency (20dB/decade or 6dB/octave). This is the result of a sigle capacitor (dominant pole) between a collector and base in the internal voltage amplifier. This is necessary to ensure stability at unity gain.
Although this amp (An OP07, as it happens) has an impressive gain at DC, nearly 120dB or 1,000,000 , at 1Khz the gain has fallen to 60dB, only 1,000! At 100Khz it is a rather useless 20dB.
If you were trying to use this op-amp in say as pre-amp with a gain of (say) 40dB, you will see from the graph that the gain starts to fall off above 10Khz!
It is this effect which needs investigation when designing audio with op-amps.
It is possible to get an op-amp without this internal capacitor, but these are not stable at unity gain and need external compensation caps.
Keep reading!
Good luck
Cliff
Note how the gain drops rapidly with frequency (20dB/decade or 6dB/octave). This is the result of a sigle capacitor (dominant pole) between a collector and base in the internal voltage amplifier. This is necessary to ensure stability at unity gain.
Although this amp (An OP07, as it happens) has an impressive gain at DC, nearly 120dB or 1,000,000 , at 1Khz the gain has fallen to 60dB, only 1,000! At 100Khz it is a rather useless 20dB.
If you were trying to use this op-amp in say as pre-amp with a gain of (say) 40dB, you will see from the graph that the gain starts to fall off above 10Khz!
It is this effect which needs investigation when designing audio with op-amps.
It is possible to get an op-amp without this internal capacitor, but these are not stable at unity gain and need external compensation caps.
Keep reading!
Good luck
Cliff
Attachments
RetroAudio said:
The 825 may be such an opamp, so why not check it out?
Well I checked. It seems nothing of special.
Maybe I am confusing graphs. Very limited when operated open loop.
Try looking at the National Semiconductor LM6171 and Elantec 2044 opamps
If I am not wrong for the National Semiconductor LM6171 there is a 10Khz bandwidth at open loop (about 80dB).
Not bad after all. Very interesting.
Thank you very much indeed.
Kind regards,
beppe
cliff said:
Thank you very much indeed Mr Cliff.
Very kind and valuable advice.
On thing I do not understand about this graph.
On the X axis I read Open Loop Gain.
But the OLG is not variable. Is fixed from what I understand.
Should not be the X axis just GAIN with the open loop gain the staring point of the curve on the left ?
There is something I do not catch.
Would not be simpler to say: " The bandwidth of the op-amp operated open loop is let's say 5 Hz ".
Maybe I am just the only one interested in this value after all.
All the other people look usually at distortion spectra and so on.
Thank you very much indeed.
Kind regards,
beppe
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.