lineup said:
No. 4 transistors made especially: NPN and PNP emitter followers with a common base, directly coupled to PNP and NPN emitter followers with common emitter. Very good matching of parameters and thermal tracking for class AB.
Speaking of class A output opamps, sorry. I don't want to rise prices: I still plan to use them. They do exist.
Wavebourn said:
Are you sure you are at the right forum ?
Prices won't rise. They are hard to find and very dated. The newer opamps sound way better than those noisy old class A opamps who were designed for video/RF purposes.
If they would have been let's say as good as a TDA1541A is in the DAC world they would have been known widely now.
I agree with Patrick here, as long as my limited experience goes.
I'm in the long term process of doing a open loop phono stage. (passive riaa, direct coupled class A etc pp bells/whistels) Only the servo has an opamp The sound is already suberb and better as with my heavily modified phonoclone. There are some reportedly very good opamp based phonostages I have yet to hear, though.
The problems are technical related: wandering and therefore unpredictable offsets relating to less than optimum component matches, shielding issues because of large area circuit (a thick packed euroboard per channel) and similar stuff.
I've seen some european top phono stages have a discrete frontend and opamp based following stages. (just guessing from the looks of the boards)
Technical, but always heartedly debated issues why a certain discrete ckt may sound better, might be:
- open loop possible (it's only the AD844 that could be used this way)
- Heavy biases
- Local degeneration possible
- higher quality passive parts
- only so much active devices as you actually need
- Very good open loop bandwith
- ... (insert something here I forgot...)
If I'd find an opamp that suits a demand and doesn't compromise the sound I would chose it, if the price tag is reasonable (the opa627's tag isn't)
Rüdiger
I'm in the long term process of doing a open loop phono stage. (passive riaa, direct coupled class A etc pp bells/whistels) Only the servo has an opamp The sound is already suberb and better as with my heavily modified phonoclone. There are some reportedly very good opamp based phonostages I have yet to hear, though.
The problems are technical related: wandering and therefore unpredictable offsets relating to less than optimum component matches, shielding issues because of large area circuit (a thick packed euroboard per channel) and similar stuff.
I've seen some european top phono stages have a discrete frontend and opamp based following stages. (just guessing from the looks of the boards)
Technical, but always heartedly debated issues why a certain discrete ckt may sound better, might be:
- open loop possible (it's only the AD844 that could be used this way)
- Heavy biases
- Local degeneration possible
- higher quality passive parts
- only so much active devices as you actually need
- Very good open loop bandwith
- ... (insert something here I forgot...)
If I'd find an opamp that suits a demand and doesn't compromise the sound I would chose it, if the price tag is reasonable (the opa627's tag isn't)
Rüdiger
Mooly said:And how many OpAmps has the music been through before it got to your ampNothing you can do about that !
Onvinyl said:
Fair point. What I was getting at was that unless you have a signal source that has not seen an OpAmp in it's travels, how can you say OpAmps are sonically inferior or a cheap option.
If your discrete "whatever" gain stage sounds superior to an OpAmp (and I am not saying it doesn't) it does not follow that OpAmps are a poor choice. In the same way that some prefer valve amps and some like amps with plenty of distortion of a certain type, your discrete option may well be adding it's own character to the sound.
So unless you use old master tapes or vinyl produced pre opamp, your reference has already been through many OpAmp stages. If the final audible result of this chain + your discrete OpAmp sounds wonderful, by definition the original source must be wonderful too, despite all the OpAmps it's been through.
If your final DIY gain stage changes the sound for the better it is almost certainly because it has added it's own less than perfect but "sonically pleasing to the ears" character ( which may be no bad thing
) but to say OpAmps are inferior -- no way.
Mooly said:Fair point. What I was getting at was that unless you have a signal source that has not seen an OpAmp in it's travels, how can you say OpAmps are sonically inferior or a cheap option.
If your discrete "whatever" gain stage sounds superior to an OpAmp (and I am not saying it doesn't) it does not follow that OpAmps are a poor choice. In the same way that some prefer valve amps and some like amps with plenty of distortion of a certain type, your discrete option may well be adding it's own character to the sound.
So unless you use old master tapes or vinyl produced pre opamp, your reference has already been through many OpAmp stages. If the final audible result of this chain + your discrete OpAmp sounds wonderful, by definition the original source must be wonderful too, despite all the OpAmps it's been through.
If your final DIY gain stage changes the sound for the better it is almost certainly because it has added it's own less than perfect but "sonically pleasing to the ears" character ( which may be no bad thing
Agree 100%. Finally some unbiased thinking.
Jan Didden
Mooly said:
...thing
You have so many (far too many to handle) things to factor in, that finally one references his private experiences one made with a certain project. This references should be made explicit, as much as possible.
There is always the danger of generalization (or, even worse, simple repeat what has been said elsewhere without a practical experience).
So, yes, your points are sensible.
OTOH, the musicians and techs I knew where in a way as crazy with equipment as the average audiophile, just for different things...
Rüdiger
Hi,
I can't add much more really. It's down to personal choice ultimately what you use, I just wanted to get the point across that it's unfair to criticise the performance of OpAmps as a generalisation as so many seem to do.
Audio is such a personal thing, and folks have such strong opinions
I would add that if you use OpAmps the implementation is critical with such things as HF compensation etc. So many designs fail on this alone. OpAmps aren't a cure all or easy option but they are capable of outstanding performance when used sypathetically.
So if you prefer the sound of a discrete design thats fine -- just remember that you probably like the extra or particular type of distortion it produces -- and if I am being honest that's why I like the sound of my MOSFET power amp so much.
Regards Karl
I can't add much more really. It's down to personal choice ultimately what you use, I just wanted to get the point across that it's unfair to criticise the performance of OpAmps as a generalisation as so many seem to do.
Audio is such a personal thing, and folks have such strong opinions
I would add that if you use OpAmps the implementation is critical with such things as HF compensation etc. So many designs fail on this alone. OpAmps aren't a cure all or easy option but they are capable of outstanding performance when used sypathetically.
So if you prefer the sound of a discrete design thats fine -- just remember that you probably like the extra or particular type of distortion it produces -- and if I am being honest that's why I like the sound of my MOSFET power amp so much.
Regards Karl
Onvinyl said:
Nor is the sound IMO. If you get the chance to listen to the top of the range ASR it will be interesting to know what you think. Although battery powered equipment sounds so dramatically different that it's difficult to compare.
I have hundreds of recordings which have obviously not been molested by opamps, but then i don't listen to any kind of pop.
Even if there were opamps in the recording chain they were certainly not required to work at microvolt levels.
It is always a good idea to incorporate an amplifiying block into an existing chain, trim gain to unity and see how much it affects the signal. I first did this in the eighties, using discretes, opamps and tubes and the results were quite unexpected. I don't claim theat an amp based on maximum transparency stages will ultimately sound good but it's still a very interesting experiment. Maybe the time is ripe to repeat it.
Onvinyl said:
Or it's you who likes the specific sound of an open loop bandwith of 10Hz -- who knows?
Or this one: THS4631 JFET input, 10KHz open loop BW, +/-15V supply: http://focus.ti.com/lit/ds/symlink/ths4631.pdf
Or this one: AD8065/66 JFET input, 10KHz open loop BW, +/-12V supply, -110dB 3rd harmonic at 100KHz: http://www.analog.com/static/imported-files/data_sheets/AD8065_8066.pdf
Guys, get real, this is the year of 2008. The uA741 times are gone.
Oh man, where to start.
There aren't many mixing consoles left that don't have opamps in them. Some of the older Neves are around and some of these were passive with gain stages on the back side. These are still desirable because " the studios like the color they add to the mix." And they still bring HUGE bucks on the used market.
One of the decisions made when making a CD is what console does the studio have because; "the producer or musician is looking for a certain sound."
A local studio I work with has over 20 different mic pre's because; "each one sounds different." Some are tube, some opamp, some discrete, some a mix.
See where all this is going?
So please be careful, opamp bad, discrete good. Does not always apply, it depends on the mission.
As a side note; the opamp of choice in a product line of mic pre's, eq's, compressors that I'm willing to bet is used on the majority of CDs made today is... the LM356. I'm not defending the '356, I am again saying, please be careful with blanket statements.
There aren't many mixing consoles left that don't have opamps in them. Some of the older Neves are around and some of these were passive with gain stages on the back side. These are still desirable because " the studios like the color they add to the mix." And they still bring HUGE bucks on the used market.
One of the decisions made when making a CD is what console does the studio have because; "the producer or musician is looking for a certain sound."
A local studio I work with has over 20 different mic pre's because; "each one sounds different." Some are tube, some opamp, some discrete, some a mix.
See where all this is going?
So please be careful, opamp bad, discrete good. Does not always apply, it depends on the mission.
As a side note; the opamp of choice in a product line of mic pre's, eq's, compressors that I'm willing to bet is used on the majority of CDs made today is... the LM356. I'm not defending the '356, I am again saying, please be careful with blanket statements.
I have posted a Sim schematic of a discrete amplifier
that had simulated Open Loop -3dB at 20 kHz
The distortion in open loop was very low, probably as I recall it below 0.2 %.
But of course this was only a preamp .. not a power amplifier.
The topology was arranged to produce less 3rd than 2nd.
I use so called current feedback for this.
The feedback goes directly into the Emitters of the input complementary pair.
I have this circuit somewhere.
This is the nicest open loop performance I have designed, so far.
The level of feedback is of course very low .. like 30-40 dB usually for my best Class A designs.
This is a level that has been used by Nelson Pass, Hiraga and JLH, too.
They all have been moderate with feedback factor.
that had simulated Open Loop -3dB at 20 kHz
The distortion in open loop was very low, probably as I recall it below 0.2 %.
But of course this was only a preamp .. not a power amplifier.
The topology was arranged to produce less 3rd than 2nd.
I use so called current feedback for this.
The feedback goes directly into the Emitters of the input complementary pair.
I have this circuit somewhere.
This is the nicest open loop performance I have designed, so far.
The level of feedback is of course very low .. like 30-40 dB usually for my best Class A designs.
This is a level that has been used by Nelson Pass, Hiraga and JLH, too.
They all have been moderate with feedback factor.
Mooly said:And how many OpAmps has the music been through before it got to your amp
The question really should not be: how many OpAmps has the music been through before it got to your amp
but rather is adding one more opamp to the chain audibly worse than not adding one more?
back before FET input op amps, there was a technique for using bipolar input op amps to make low noise amplifiers. the trick was to use a matched pair of discrete JFETs as a diff amp input stage to the op amp. one version of it is in Nat Semi AN-32 figure 30.
the things to consider when deciding whether or not to use discrete designs:
1>do i need absolute control over parameters such as open loop gain, slew rate, pole frequency, and input characteristics, etc...? (if yes to any of these, build a discrete)
2> is there an "off the shelf" solution to what i want to accomplish? (if yes, use an op amp)
3>can i accomplish what i want to do by combining an op amp with discrete circuits? (if yes, build a hybrid)
the things to consider when deciding whether or not to use discrete designs:
1>do i need absolute control over parameters such as open loop gain, slew rate, pole frequency, and input characteristics, etc...? (if yes to any of these, build a discrete)
2> is there an "off the shelf" solution to what i want to accomplish? (if yes, use an op amp)
3>can i accomplish what i want to do by combining an op amp with discrete circuits? (if yes, build a hybrid)
Some are still hung up on open loop BW. It is completely irrelevant to almost anything. There are amplifiers that on a lucky die have 120dB of OLG and on a less lucky one 100dB and a subsequent factor of 10 difference in OLBW. This makes no difference what so ever in performance in normal closed-loop applications. One could add a bunch of DC transfer function fixes to an amp like the AD8065 and give the same basic amplifier an OLG of 140dB and much lower OLBW, again no difference in real applications. Gain at frequency of interest is what matters and that does not change.
scott wurcer said:
Why have I been told the O-Loop B-Width matters?
Is it relevant at all?
You say almost, Scott.
Gain at frequency of interest.
You mean between 20 -20.000 Hz, I suppose.
I can still perceive tiny bits of sounds >=12.000
But close to nothing. As I have tested myself.
So, when making an amplifier,
I would, so to say, be interested of <=15 kHz performance.
And the Gain at those frequencies.
You mean if the amplifier will have a flat gain upto 15 kHz Open Loop
I can be satisfied.
Right?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.