OPAMP upgrades
Michael Jeffreys
some of the components.
The first items to go will be the NE5532 and NE5534s. What are the best
replacements for these opamps?
There also seems to be an LM637 in the phono stage. Is this (opamp?)
worth keeping?
The amp also uses IRFP240 MOSFETs in a quasi-complementary mode.
Should I fit the higher current versions of these items or just parallel
them up.
Thanks.
--
MIKE JEFFREYS.
Michael Jeffreys
>There also seems to be an LM637 in the phono stage. Is this (opamp?)
>worth keeping?
At the risk of being seen 'talking to myself' I have just rxed a fax from
Nat Semi, and the LM637 is an improved (faster) OP37. This is not as
quiet as an LT1028, so out it comes :-)
--
MIKE
------------------------------------------------------------------------------
NB E&OE (TM)
Monte P McGuire
Michael Jeffreys <jeffre...@bt-web.bt.co.uk> wrote:
>In article <3tr037$7...@pheidippides.axion.bt.co.uk>,
>jeffre...@bt-web.bt.co.uk (Michael Jeffreys) says:
>
>>There also seems to be an LM637 in the phono stage. Is this (opamp?)
>>worth keeping?
>
>At the risk of being seen 'talking to myself' I have just rxed a fax from
>Nat Semi, and the LM637 is an improved (faster) OP37. This is not as
>quiet as an LT1028, so out it comes :-)
Are you sure it will be quieter in that particular circuit? The
LT1028 achieves low noise by heavily biasing the input stage and the
tradeoff is increased input _current_ noise. If you use it with
source impedances greater than about 1K or so, it will be noisier than
an OP37.
Is the LM637 unity gain stable? Do they spec any distortion figures
for the LM637? The OP37 is not an awful amp, but I can hear the OP27
pretty easily and there's only 10dB improvement between the two; this
makes me uncomfortable about the OP37.
Another one to try: the Burr Brown OPA627 and 637 have less distortion
and less input current noise than the OP37 at the expense of slightly
greater voltage noise. I'd compare those to the LM637.
As far as 5532 and 5534 replacements go, the Analog Devices OP176 and
OP275 are pretty good, as are the Burr Brown OPA 604 and 2604. The
latter have more voltage noise and less current noise, so they may not
be as quiet as the 553[24], but they will have less distortion.
Best of luck,
Monte McGuire - N1TBL
mcg...@world.std.com
Kalman Rubinson
: I have an ARCAM Delta 290 integrated amplifier and want to upgrade
: some of the components.
: The first items to go will be the NE5532 and NE5534s. What are the best
: replacements for these opamps?
Depends on the circuit configurations.
: There also seems to be an LM637 in the phono stage. Is this (opamp?)
: worth keeping?
Ditto.
: The amp also uses IRFP240 MOSFETs in a quasi-complementary mode.
: Should I fit the higher current versions of these items or just parallel
: them up.
What do you hope to achieve?
Kal
Kalman Rubinson
: At the risk of being seen 'talking to myself' I have just rxed a fax from
: Nat Semi, and the LM637 is an improved (faster) OP37. This is not as
: quiet as an LT1028, so out it comes :-)
Yes but have you 'heard' or used an LT1028? Do you know that the offset
and trim pins are different? Is there a noise problem now?
Kal
k
PETER LARSEN
MJ> I have an ARCAM Delta 290 integrated amplifier and want to upgrade
MJ> some of the components.
Hmm ...
MJ> The first items to go will be the NE5532 and NE5534s. What are the
MJ> best replacements for these opamps?
According to their specs Analog Devices OP275 (dual) and OP176 (single)
are designed to be replacement devices.
MJ> There also seems to be an LM637 in the phono stage. Is this (opamp?)
MJ> worth keeping?
Oh yes .. put in a small box with a glass window and hang it on the
wall! - unless of course it is a quad, because finding alternatives to
those is not that simple. Suggestions gladly received ... must be unity
gain stable!
MJ> The amp also uses IRFP240 MOSFETs in a quasi-complementary mode.
MJ> Should I fit the higher current versions of these items or just
MJ> parallel them up.
Dunno, look for unbiased electrolytics in the signal path first, usually
one can get away with using components an order of magnitude smaller,
but finding the room for 'audiophile grade' capacitors isn't always
simple, and can require some ingenuity, it takes thorough understanding
of the circuit to determine whether they are totally redundant, and
without that thorough understanding of the circuit it may be better to
replace them than to simply remove them. As far as the current goes,
well it does come from the power-supply doesn't it .... doubling the
capacitors in it or better may well be the next step. Also doing that
has the advantage when compared to adding/changing output devices that
you don't have to worry about the risk of needing a shortwave licence to
use the amp for the number of microseconds it will last under worst
instance conditions of instability.
All of the above just my USD 0.02, and void if wrong or inapplicable,
other people may have better suggestions.
Get an application note for the OP176 in case there is 'dc midpoint'
adjustment circuit around the NE5534, if so, it will have to be either
removed or rewired to a different pin.
Kind regards
Peter Larsen
******************************************************************
* Internet....: pla...@dkb.dk, Copenhagen, Denmark *
* Fidonet.....: Peter Larsen@2:235/334.4 *
* Relaynet(tm): Address to Peter Larsen, route ->1117 *
* Conference Co-host for the Relaynet Amiga Conference *
******************************************************************
---
ş RM 1.3 03104 ş Internet: pla...@dkb.dk, Fidonet: Peter Larsen@2:235/334.4
Leon Vaicius
jeffre...@bt-web.bt.co.uk says...
>
>The first items to go will be the NE5532 and NE5534s. What are the best
>replacements for these opamps?
>
Analog Devices has an improved replacement for the 5534 called SSM-2134.
These are claimed to have lower noise and used in professional audio
equipment. I have seen these marketed by Newerk and by Allied electronics.
Analog Devices has a publication called Audio/Video Reference Manual that
contains data and application notes for their high end audio/video
components. This manual is interesting, & informational. The application
notes opened my eyes. Worth getting.
Leon
Bernhard Muller
PETER LARSEN (pla...@dkb.dk) writes:
> MJ> I have an ARCAM Delta 290 integrated amplifier and want to upgrade
> MJ> some of the components.
>
> Hmm ...
>
> MJ> The first items to go will be the NE5532 and NE5534s. What are the
> MJ> best replacements for these opamps?
>
> According to their specs Analog Devices OP275 (dual) and OP176 (single)
> are designed to be replacement devices.
What is wrong with 5534's or, for that matter, 32's? I have used them for
many years and they have many excellent properties. I find they sound
very neutral.
--
We can choose to Throw Stones
to Stumble over them
Bern Muller to Climb over them
or to Build with them.--William Arthur Ward
Monte P McGuire
Bernhard Muller <bp...@FreeNet.Carleton.CA> wrote:
>
>What is wrong with 5534's or, for that matter, 32's? I have used them for
>many years and they have many excellent properties. I find they sound
>very neutral.
They're pretty good op amps and at the time they were introduced, they
were completely state of the art, one of the first few monolithic op
amps that actually sounded good. The 5534 and the TL070 series were
basically 'it' in the 70s and 80s and I'd estimate 80% of the
currently sold professional and semi professional audio equipment
still uses one of the two op amp families.
However, 20 years have passed and the semiconductor industry has moved
forward. Not only has process technology advanced, but op amp circuit
design and topologies are more advanced as well. The current state of
the art offers 10x less distortion, several times less noise, better
DC precision and better output stages. Not always at low cost, but
that's the state of the art today. These newer chips (like the AD797)
do perform better in older circuits designed for the 5534 and if you
want the best possible performance, there's no reason not to use them.
Regards,
Bernhard Muller
Subject: Re: OPAMP upgrades
Organization: The National Capital FreeNet
PETER LARSEN (pla...@dkb.dk) writes:
> MJ> I have an ARCAM Delta 290 integrated amplifier and want to upgrade
> MJ> some of the components.
>
> Hmm ...
>
> MJ> The first items to go will be the NE5532 and NE5534s. What are the
> MJ> best replacements for these opamps?
>
> According to their specs Analog Devices OP275 (dual) and OP176 (single)
> are designed to be replacement devices.
What is wrong with 5534's or, for that matter, 32's? I have used them for
many years and they have many excellent properties. I find they sound
very neutral.
--
We can choose to Throw Stones
to Stumble over them
Bern Muller to Climb over them
or to Build with them.--William Arthur Ward
--
|Fidonet: Bernhard Muller 2:283/618.99
|Internet: Bernhar...@p99.f618.n283.z2.fidonet.org
|
| Standard disclaimer: The views of this user are strictly his own.
Monte P McGuire
Subject: Re: OPAMP upgrades
Organization: The World Public Access UNIX, Brookline, MA
In article <DBqrB...@freenet.carleton.ca>,
Bernhard Muller <bp...@FreeNet.Carleton.CA> wrote:
>
>What is wrong with 5534's or, for that matter, 32's? I have used them for
>many years and they have many excellent properties. I find they sound
>very neutral.
They're pretty good op amps and at the time they were introduced, they
were completely state of the art, one of the first few monolithic op
amps that actually sounded good. The 5534 and the TL070 series were
basically 'it' in the 70s and 80s and I'd estimate 80% of the
currently sold professional and semi professional audio equipment
still uses one of the two op amp families.
However, 20 years have passed and the semiconductor industry has moved
forward. Not only has process technology advanced, but op amp circuit
design and topologies are more advanced as well. The current state of
the art offers 10x less distortion, several times less noise, better
DC precision and better output stages. Not always at low cost, but
that's the state of the art today. These newer chips (like the AD797)
do perform better in older circuits designed for the 5534 and if you
want the best possible performance, there's no reason not to use them.
Regards,
Monte McGuire - N1TBL
mcg...@world.std.com
--
|Fidonet: Monte P McGuire 2:283/618.99
|Internet: Monte.P...@p99.f618.n283.z2.fidonet.org
Bernhard Muller
Monte P McGuire (mcg...@world.std.com) writes:
> In article <DBqrB...@freenet.carleton.ca>,
> Bernhard Muller <bp...@FreeNet.Carleton.CA> wrote:
>>
>>What is wrong with 5534's or, for that matter, 32's? I have used them for
>>many years and they have many excellent properties. I find they sound
>>very neutral.
>
> They're pretty good op amps and at the time they were introduced, they
> were completely state of the art, one of the first few monolithic op
> amps that actually sounded good. The 5534 and the TL070 series were
> basically 'it' in the 70s and 80s and I'd estimate 80% of the
> currently sold professional and semi professional audio equipment
> still uses one of the two op amp families.
>
> However, 20 years have passed and the semiconductor industry has moved
> forward. Not only has process technology advanced, but op amp circuit
> design and topologies are more advanced as well. The current state of
> the art offers 10x less distortion,
At gains of 60 dB the 5534 still has inaudible distortion. So less really
doesn't matter.
several times less noise,
The noise performance of 5524's is, again, sufficiently good. The noise
in an RIAA input stage of typical gain built around a 5534 is completely
inaudible. For a moving coil preamp you might want to use a different unit.
> better
> DC precision
DC precision is generally unimportant in Audio work since the power amps
generally are not DC coupled.
> and better output stages.
5534's drive 600 ohms with ease. No reason to need more unless you are
building a power amplifier.
Not always at low cost, but
> that's the state of the art today. These newer chips (like the AD797)
> do perform better in older circuits designed for the 5534 and if you
> want the best possible performance, there's no reason not to use them.
You mentioned one reason: cost. The benefits of the newer chips are
present, just as you say, but you get those benefits in certain specific
circuits, not generally in audio circuits. Wholesale replacement of
5534's is a waste of time and money unless there is a specific
identifiable problem that you can cure with the new chip. And noise,
distortion, DC performance and output drive capability are not usually
problems in audio circuits using 5534's.
Monte P McGuire
Bernhard Muller <bp...@FreeNet.Carleton.CA> wrote:
>
>At gains of 60 dB the 5534 still has inaudible distortion. So less really
>doesn't matter.
So why does a 797 sound much more clear and open than a 5534 in my
console's stereo mix bus stage? The 5534 barely has 60dB of open loop
gain at 20KHz, so if you can't hear an op amp running nearly open
loop, something's wrong in your listening setup. Clean up the rest of
the system and try again.
> several times less noise,
>
>The noise performance of 5524's is, again, sufficiently good. The noise
>in an RIAA input stage of typical gain built around a 5534 is completely
>inaudible. For a moving coil preamp you might want to use a different unit.
All audio circuits are not MM RIAA preamps!! A 797 is quiet enough
for a 150 ohm dynamic mike, the 5534 is not. For impedances around
1-10K, the 5534 is a reasonably good choice noise wise, but there are
JFET amps like the OPA627 that have the same voltage noise and a tiny
fraction of the current noise. Likewise, the LT1028 and AD797 have
more current noise but far less voltage noise, essential for use with
sources under 500 ohms.
>> better
>> DC precision
>
>DC precision is generally unimportant in Audio work since the power amps
>generally are not DC coupled.
If you can remove one coupling cap, I think it's a good thing. Mixing
consoles have lots and lots of coupling caps because they have lots of
amps, pots and switches all in series. If you can simply eliminate
the offset with a good amp, you can remove lots of coupling caps that
are downstream.
>> and better output stages.
>
>5534's drive 600 ohms with ease. No reason to need more unless you are
>building a power amplifier.
What about a signal multed to several 600 ohm loads?? At +15dBV??
How about long cables? Even with a 50 ohm build out resistor, there's
still a bit of capacitance to drive sometimes and there's no reason to
get extra distortion from the output stage. Again, for consumer gear
with 50K loads and 2' of cable, anything will drive that, so it might
not matter there. There are some spots in pro audio gear that do
require hefty current drive with no additional distortion and modern
chips work better there.
> Not always at low cost, but
>> that's the state of the art today. These newer chips (like the AD797)
>> do perform better in older circuits designed for the 5534 and if you
>> want the best possible performance, there's no reason not to use them.
>
>You mentioned one reason: cost. The benefits of the newer chips are
>present, just as you say, but you get those benefits in certain specific
>circuits, not generally in audio circuits. Wholesale replacement of
>5534's is a waste of time and money unless there is a specific
>identifiable problem that you can cure with the new chip. And noise,
>distortion, DC performance and output drive capability are not usually
>problems in audio circuits using 5534's.
Please, go ahead and use 5534s. Use LM301s with feedforward
compensation if you'd like - they both have negligible distortion, if
negligible means <2% like the 60dB case you claim above. Don't claim
that the distortion is inaudible in every circuit though, because it
is not. If you really ruin a signal with a few bad stages then yes, a
few more percent of trash isn't really going to be that audible, but
what about completely clean signals straight off of a condenser
microphone? They do exist sometimes! And gear that handles them
should at least try to be better than the storage medium that's down
the line a few stages.
Also, just because a designer used a 5534, it does not automatically
become the best performing chip available for all time. My mixing
console was designed 14 years ago and the 5534 was the best choice
then for a mix amp. There are better chips now, so I use them.
Should we have stuck with the LM301 just because it's a reasonably OK
amp? Why is it a waste of time to expect the distortion floor of a
system to be below the noise floor??
Bernhard Muller
> Bernhard Muller <bp...@FreeNet.Carleton.CA> wrote:
>>
>>At gains of 60 dB the 5534 still has inaudible distortion. So less really
>>doesn't matter.
>
> So why does a 797 sound much more clear and open than a 5534 in my
> console's stereo mix bus stage?
Beats me. You made the change, I assume you made some kind of measurement
to support your assertion. Tell me.
> The 5534 barely has 60dB of open loop
> gain at 20KHz, so if you can't hear an op amp running nearly open
> loop, something's wrong in your listening setup. Clean up the rest of
> the system and try again.
I don't hear well at the 2nd harmonic of 20 KHz. I admit I probably would
cascade a couple of op amps if I were seeking 60 dB of flat gain. But, as
you say below, 2% at 20 kHz. I can't hear 2% harmonic distortion of a 20 kHz
wave on ANY system. I I doubt you can.
>
> >> several times less noise,
>>
>>The noise performance of 5524's is, again, sufficiently good. The noise
>>in an RIAA input stage of typical gain built around a 5534 is completely
>>inaudible. For a moving coil preamp you might want to use a different unit.
>
> All audio circuits are not MM RIAA preamps!! A 797 is quiet enough
> for a 150 ohm dynamic mike, the 5534 is not. For impedances around
> 1-10K, the 5534 is a reasonably good choice noise wise, but there are
> JFET amps like the OPA627 that have the same voltage noise and a tiny
> fraction of the current noise. Likewise, the LT1028 and AD797 have
> more current noise but far less voltage noise, essential for use with
> sources under 500 ohms.
Sure. Like I said, there are special circumstances where other op ams are
better. The above is one.
>>DC precision is generally unimportant in Audio work since the power amps
>>generally are not DC coupled.
>
> If you can remove one coupling cap, I think it's a good thing. Mixing
> consoles have lots and lots of coupling caps because they have lots of
> amps, pots and switches all in series. If you can simply eliminate
> the offset with a good amp, you can remove lots of coupling caps that
> are downstream.
Most mass replacers of op amps don't do what you suggest: replace great
bunches of caps. and anyway, why is replacing caps a Good Thing? better low
frequency response, perhaps. Replacing electrolytics and ceramics in the
signal path is a good thing, perhaps. But a polycarbonate? This is
beginning to sound like mysticism.
>
> >>> and better output stages.
>>
>>5534's drive 600 ohms with ease. No reason to need more unless you are
>>building a power amplifier.
>
> What about a signal multed to several 600 ohm loads?? At +15dBV??
> How about long cables? Even with a 50 ohm build out resistor, there's
> still a bit of capacitance to drive sometimes and there's no reason to
> get extra distortion from the output stage. Again, for consumer gear
> with 50K loads and 2' of cable, anything will drive that, so it might
> not matter there. There are some spots in pro audio gear that do
> require hefty current drive with no additional distortion and modern
> chips work better there.
Again, a special purpose. I completely agree with your last 2 sentences.
>>circuits, not generally in audio circuits. Wholesale replacement of
>>5534's is a waste of time and money unless there is a specific
>>identifiable problem that you can cure with the new chip. And noise,
>>distortion, DC performance and output drive capability are not usually
>>problems in audio circuits using 5534's.
>
> few more percent of trash isn't really going to be that audible, but
> what about completely clean signals straight off of a condenser
> microphone? They do exist sometimes! And gear that handles them
> should at least try to be better than the storage medium that's down
> the line a few stages.
It is easy enough to measure if any stages are 'dirtying' up the signal.
deal with those stages specifically. Your vigorous reply suggests some
strong feeling. I am sorry if I offended you. But you have given me no
reason to change my statement above. I repeat: If you have a specific
problem that the 5534 can't handle, they by all means replace it. But
wholesale replacement is a waste of time and money.
>
> Also, just because a designer used a 5534, it does not automatically
> become the best performing chip available for all time. My mixing
> console was designed 14 years ago and the 5534 was the best choice
> then for a mix amp. There are better chips now, so I use them.
> Should we have stuck with the LM301 just because it's a reasonably OK
> amp? Why is it a waste of time to expect the distortion floor of a
> system to be below the noise floor??
There are applications where the 301 is probably just fine (I wouldn't use
it in audio) There are applications in audio where the newer and more
expensive chips will make no measurable or audible difference. Why use
them, particularly when the adequate chip is already in place. I didn't
comment on noise versus distortion floors. Why do you claim I did?
In any case, you can do what you wish in your studio. But to advise
others to spend a lot of money without good solid justification (specific
problems to solve) is unkind.
Klaus
Does anyone have a good suggestions for a replacement (upgrade) for an RC
4558? I have an older amp that uses a couple of these, and I'd like to
clean it up a bit (without re-designing the circuitry).
Thanks.
-- Klaus (heil...@math.berkeley.edu, http://math.berkeley.edu/~heilmayr)
Monte P McGuire
Klaus <heil...@math.berkeley.edu> wrote:
>OK, as long as we're talking about op-amp upgrades...
>
>Does anyone have a good suggestions for a replacement (upgrade) for an RC
>4558? I have an older amp that uses a couple of these, and I'd like to
>clean it up a bit (without re-designing the circuitry).
There are a few choices: Burr Brown OPA2604 if you need low input
current noise but can tolerate slightly higher input voltage noise,
Analog Devices OP275 if you want lower voltage noise and similar
current noise, and Analog Devices AD712 if you want the euphonic
'rising 2HD with increasing frequency' distortion that many BiFETs
happily generate.
All three sound good but the first two are probably cleanest; the
AD712 is basically like a souped up TL072 and while it sounds good, it
does add euphonic (IMHO) distortion when used in noninverting mode.
The 4558 is really a horrible sounding amp, so if you can't find any
of the above, go to Radio Shack and get some TL072 as a stopgap.
Best of luck,
Monte P McGuire
Bernhard Muller <bp...@FreeNet.Carleton.CA> wrote:
>Monte P McGuire (mcg...@world.std.com) writes:
>> So why does a 797 sound much more clear and open than a 5534 in my
>> console's stereo mix bus stage?
>
>Beats me. You made the change, I assume you made some kind of measurement
>to support your assertion. Tell me.
I think it's because the 797 has 10x less distortion. They both have
adequate open loop gain for the task at hand and also sufficient
bandwith, so that's the only spec I can muster up as a justification.
>> The 5534 barely has 60dB of open loop
>> gain at 20KHz, so if you can't hear an op amp running nearly open
>> loop, something's wrong in your listening setup. Clean up the rest of
>> the system and try again.
>
>I don't hear well at the 2nd harmonic of 20 KHz. I admit I probably would
>cascade a couple of op amps if I were seeking 60 dB of flat gain. But, as
>you say below, 2% at 20 kHz. I can't hear 2% harmonic distortion of a 20 kHz
>wave on ANY system. I I doubt you can.
That's not really the point. If you understand what dominant pole
amplifier compensation means, then if I say that "an op amp is open
loop at 60dB @ 20KHz" then you can instantly know that a flat 60dB amp
made from such a device has only 12dB of feedback at 5KHz and 24dB of
feedback at 1.25KHz and so on. What I'm saying is not so much related
to 20KHz and it's harmonics, but the fact that you're running the amp
with very little feedback over the high frequency range and that's not
good.
Now, if you're talking RIAA stages, which are essentially integrators,
then the above is irrelevant, since such a circuit has uniform
feedback over the entire band. So, 60dB of gain at DC with a 5534
really means 40dB of feedback everywhere, an entirely rosy
situation...
[snip]
>> few more percent of trash isn't really going to be that audible, but
>> what about completely clean signals straight off of a condenser
>> microphone? They do exist sometimes! And gear that handles them
>> should at least try to be better than the storage medium that's down
>> the line a few stages.
>
>It is easy enough to measure if any stages are 'dirtying' up the signal.
>deal with those stages specifically. Your vigorous reply suggests some
>strong feeling. I am sorry if I offended you. But you have given me no
>reason to change my statement above. I repeat: If you have a specific
>problem that the 5534 can't handle, they by all means replace it. But
>wholesale replacement is a waste of time and money.
Now that my feet touch the ground, I can see that the original
poster's request really has more to do with consumer circuits that are
really not stressed as much as some of the pro circuits that I work
with. For those uses, it may be difficult to hear the benefits of a
cleaner op amp than a 5534.
However, some people wish to explore the outer limits of performance
and don't really care about whether the little bit of residual
distortion is audible or not in any particular circumstance. They
would rather remove the distortion than play devils advocate and see
if they are able to find a program that allows them to hear the
difference. I think that's why a person asks about which amps sound
good; they don't have the ability to run good listening tests and
don't want to buy lots and lots of chips just to find one that sounds
good. Advice is cheap...
Personally, I find it easier to just listen to the amps directly. I
compare the sound of a circuit (adjusted for unity gain) to the sound
of the system with that circuit patched out. If there's a difference,
then the amp that makes the difference smallest is the winner. If
there's no audible difference, preference goes to the circuit that
specs lower distortion, lower noise or lower offset. Not a perfect
method, but over time it has basically steered me in the right
direction.
John 015
(Bernhard Muller) said to Monte P McGuire:
>
>We don't basically disagree all that much. You got caught up in my strong
>feelings about folks advising neophytes to do all kinds of expensive
>difficult things to their systems that have little likelihood of making
>significant improvement. Clearly, if someone wants to have fun and
>experiment, while knowing what they are doing, I'm all for it.
>
I heartedly agree but in *some* circuits the opamp makes all the
difference: cloning the (infamous?) CD3400 "sound" proved impossible
with the any of the standard high-end opamps (1028, 797, 2406, 827,
275, 712, 249 etc etc).
The AD275 is probably the cleanest and most sterile sounding
op-amp I have tried. I have yet to leave it in any of my circuits
but I can see why people recommend it on the recording side.
John
Sheldon D. Stokes
> We don't basically disagree all that much. You got caught up in my strong
> feelings about folks advising neophytes to do all kinds of expensive
> difficult things to their systems that have little likelihood of making
> significant improvement. Clearly, if someone wants to have fun and
> experiment, while knowing what they are doing, I'm all for it.
>
My $0.02:
I typically replace all my op-amps (I only have them in non critical
appliances like my tuner, tape deck and toaster) with sockets. That
way I can try any chip I feel like. The two I keep large stocks of are:
Motorolla MC34082. This is a really nice sounding opamp. It's a
little on the noisey side, but the sound quality is very nice. The
other one I keep on hand is the Burr Brown OPA2604. It is a little
more sterile sounding, but it's very nice too, and has a lower noise
floor. The nice thing about these two devices is that they are pretty
cheap.
I always socket op-amps so I can play with new ones when they come out.
Sheldon
PETER LARSEN
Hi Bernhard,
References: <DBqrB...@freenet.carleton.ca> <3tr037$7...@pheidippides.axion.bt.co.uk> <8AD1517.1406003765.uuout@
BM> > According to their specs Analog Devices OP275 (dual) and OP176
BM> (single) > are designed to be replacement devices.
BM> What is wrong with 5534's or, for that matter, 32's? I have used
BM> them for many years and they have many excellent properties. I find
BM> they sound very neutral.
There is no disputing that the 553x's are quite good, also they have
some advantages in some circuits compared to the OP275/OP176 in case
demands for output swing are large, unless the supply voltage is raised
to +/- 18 to +/- 22 Volts, can't remember what the upper limit is. So,
and this is according to info received via this forum, there may be
applciations where it isn't just 'plug and play'.
As for the sonic difference ... and of course for whether it matters for
whatever device ... why not simply try them. In my personal opinion the
OP275/OP176 sounds much cleaner in the treble, and has better
intertransient silence, better perspective and a very nice bass. It has
many of the excellent characteristics of a really well designed circuit
made with discrete components.
So if you have some device with say 10 of the 553x's, and they are
socketed, try to get 10 of the suggested alternative, and plug them in.
There may well be other great opamps out there, but the suggested
alternative has a very nice price, and if you use them in multitude
(such as the Denon DE70 recently modified ... 21 OP275's was what it
took, and there are some quad opamps left I haven't found an alternative
to, but it was 4558 and 4560 that was in it .... ) that is nice, and it
may also be nice that they use less power than the 553x-series. Other
things may also matter .... but I would be very interested to hear from
you in case you have any application in which you say on trying them
that they were not better. Not because of any affiliation to Analog
Devices of any kind, but as I usually do recommend them as in my
experience good for the purposes I have used them for, I very much would
like to know if there is some caveat about them, some situation in which
they are not a cost-efficient improvement of some device.
Bernhard Muller
Monte P McGuire (mcg...@world.std.com) writes:
>
> That's not really the point. If you understand what dominant pole
> amplifier compensation means, then if I say that "an op amp is open
> loop at 60dB @ 20KHz" then you can instantly know that a flat 60dB amp
> made from such a device has only 12dB of feedback at 5KHz and 24dB of
> feedback at 1.25KHz and so on. What I'm saying is not so much related
> to 20KHz and it's harmonics, but the fact that you're running the amp
> with very little feedback over the high frequency range and that's not
> good.
>
> Now, if you're talking RIAA stages, which are essentially integrators,
> then the above is irrelevant, since such a circuit has uniform
> feedback over the entire band. So, 60dB of gain at DC with a 5534
> really means 40dB of feedback everywhere, an entirely rosy
> situation...
I got carried away with hyperbole in my original post. I _was_ thinking
of RIAA circuits. We basically agree.
>
> Now that my feet touch the ground, I can see that the original
> poster's request really has more to do with consumer circuits that are
> really not stressed as much as some of the pro circuits that I work
> with. For those uses, it may be difficult to hear the benefits of a
> cleaner op amp than a 5534.
-snip-
>
> Personally, I find it easier to just listen to the amps directly. I
> compare the sound of a circuit (adjusted for unity gain) to the sound
> of the system with that circuit patched out. If there's a difference,
> then the amp that makes the difference smallest is the winner. If
> there's no audible difference, preference goes to the circuit that
> specs lower distortion, lower noise or lower offset. Not a perfect
> method, but over time it has basically steered me in the right
> direction.
We don't basically disagree all that much. You got caught up in my strong
feelings about folks advising neophytes to do all kinds of expensive
difficult things to their systems that have little likelihood of making
significant improvement. Clearly, if someone wants to have fun and
experiment, while knowing what they are doing, I'm all for it.
--
PETER LARSEN
Hi Bernhard,
References: <DBsJq...@freenet.carleton.ca> <3tr037$7...@pheidippides.axion.bt.co.uk> <8AD1517.1406003765.uuout@
BM> At gains of 60 dB the 5534 still has inaudible distortion. So less
BM> really doesn't matter.
if it is inaudible, how come the OP275/176 sounds better? ... anyway,
you mention a lot of tech spec points, and not at all listening tests,
and after all those are in fact the final test to make a decision by, at
least when 'all things are equal' in terms of tech specs ... not that
you say they are, just that in your opinion the tech spec differences do
not matter and would not be associated with any sonic difference.
PETER LARSEN
Hi Bernhard,
References: <DBuFv...@freenet.carleton.ca> <3tr037$7...@pheidippides.axion.bt.co.uk> <DBqrB...@freenet.carlet
BM> Most mass replacers of op amps don't do what you suggest: replace
BM> great bunches of caps. and anyway, why is replacing caps a Good
BM> Thing? better low frequency response, perhaps. Replacing
BM> electrolytics and ceramics in the signal path is a good thing,
BM> perhaps. But a polycarbonate? This is beginning to sound like
BM> mysticism.
being one of those 'replacers of opamps' I will remind you that I don't
advocate replacing any and all caps, but only unbiased electrolytics.
But it is technical absurd to the best of my knowledge to have them
coupling caps all ower the place ... one in the input, one between each
amplifier stage and one at the output ... what's the point of having
them .... it seems to me that they are _relevant_ to protect transducers
and potentiometers, but other than that? ... I ask because I'm seriously
considering to simply remove them where it is difficult to find room for
a quality replacement for an unbiased electrolytic!
your local Dave
place a larger order for something more audiophile-quality.
However, my local rat shack only has TL082.. will this work as
a relatively low-noise audio op-amp in a crossover? I don't
know the characteristics (or how to interpret if I did)..
Dave Douglas
dm...@virginia.edu
tINY
>
>Hi Bernhard,
>
>References: <DBsJq...@freenet.carleton.ca> <3tr037$7...@pheidippides.axion.bt.co.uk> <8AD1517.1406003765.uuout@
>
>BM> At gains of 60 dB the 5534 still has inaudible distortion. So less
>BM> really doesn't matter.
>
>if it is inaudible, how come the OP275/176 sounds better? ... anyway,
>you mention a lot of tech spec points, and not at all listening tests,
>and after all those are in fact the final test to make a decision by, at
>least when 'all things are equal' in terms of tech specs ... not that
>you say they are, just that in your opinion the tech spec differences do
>not matter and would not be associated with any sonic difference.
Remember, unless you are changing the rest of the circuit (scaling
impeadances, optimizing for Vos, etc), the sound is likely better with the
op-amp which more closely resembles the original part. Also, the Vos vs Vcm is
crucial to linearity. Without typical curves for this parameter, a single
number for CMRR and Vos is almost meaningless.
Also, to find out what is going on, you may want to look at the slew rate vs
the freq and signal level in your situation (you may like a little even order
harmonic distortion). The OP-x series of devices tend to have a lower slew
rate.
There is no perfect op-amp, all have trade-offs. The integration of the
device and the design are what make exceptional sound.
-tINY
Rich Hotaling
>
>Monte P McGuire (mcg...@world.std.com) writes:
>
>>
> We can choose to Throw Stones
> to Stumble over them
> Bern Muller to Climb over them
> or to Build with them.--William Arthur Ward
Whew! Boy am I glad you guys kissed and made up. Can't we all be nice here?
If nothing else, it's more entertaining than T.V.
See Ya, Richard C. Hotaling
we...@cloud.com
tINY
Raidio Schlack usually carries 5532's. These are well suited to active
x-overs. Keep the load on each output at about 1k. You may want to balance
the source impeadances, as they draw a fair amount of bias current. Use good
1% metal film resistors and MPF or Poly caps.
If you look in your phone book, I bet there is a real electronics parts store
not too far from you.
-tINY
Bernhard Muller
PETER LARSEN (pla...@dkb.dk) writes:
> Hi Bernhard,
> BM> electrolytics and ceramics in the signal path is a good thing,
> BM> perhaps. But a polycarbonate? This is beginning to sound like
> BM> mysticism.
>
> being one of those 'replacers of opamps' I will remind you that I don't
> advocate replacing any and all caps, but only unbiased electrolytics.
> But it is technical absurd to the best of my knowledge to have them
> coupling caps all ower the place ... one in the input, one between each
> amplifier stage and one at the output ... what's the point of having
I agree. But mass removers of coupling caps need to have op amps with
very low DC offsets so that the accumulated DC offset doesn't eventually
cause the output of one stage to go to the rail. Low DC offset was
mentioned as a benefit of op amp replacement by a writer who I think was a
cap remover. I prefer to leave good quality caps in as you do, and not
worry about DC offset.
> them .... it seems to me that they are _relevant_ to protect transducers
--
Michael Jeffreys
Further to my requests for knowledge on 'super' opamps, it is
obvious that I must know the source resistance. Can anyone
work it out for me from the diagram below showing the initial
stage of the MM phono input to my ARCAM Delta 290.
My guesses are either ~2K or ~47K with a gain of 7.
If the source resistance is less than 2K I think I can use
some of the ultra low noise opamps LT1028 etc.
| \
PHONO IN ---- 330 ----------| + \
| | \
| | -------- OUT
| | / |
---| - /
47K | | / 1K8
|
| | |
| ---------------
| |
|
| 330
|
| |
| |
--- ---
0V 0V
Thanks!
--
MIKE
------------------------------------------------------------------------------
NB E&OE (TM)
john 015
jeffre...@bt-web.bt.co.uk (Michael Jeffreys) said:
>
>Further to my requests for knowledge on 'super' opamps, it is
>obvious that I must know the source resistance. Can anyone
>work it out for me from the diagram below showing the initial
>stage of the MM phono input to my ARCAM Delta 290.
>
>My guesses are either ~2K or ~47K with a gain of 7.
>
>If the source resistance is less than 2K I think I can use
>some of the ultra low noise opamps LT1028 etc.
>
First the MM: It will have a DC source resistance and
numerous frequency dependent terms. I speculate the
inductance in your average MM is large enough to dominate
the DC resistance. So your source impedance can probably
be written R_dc + jwL with L rather big. I have no numbers
for MM cartridges but speculate maybe 500 ohms + huge L
terms which could bring you up towards 10 kOhms at 20 kHz.
(This is a bit of a dilemma in especially tape-heads pre-amps
where the source can start out at maybe 180 ohms at low freq
and exceed 10 kOhms at high but audible frequencies).
Now, in *series* with the source signal generator we find:
330 ohms in the phono in line, then the contribution
from the opamp (+ to -) [assuming it's much smaller
than the load resistor 47k so we neglect the 47k].
We haven't closed the loop yet: we have another
330 between the op-amp inverting input and ground.
So I find the source impedance as _seen_ by the op-amp to be
(Z_cartridge + 330 + 330).
I say "seen" because the op-amp will add it's own contribution
to the loop considered above because the loop includes the path
that links the two inputs of the op-amp. This is important for
noise considerations because we want the noise contribution from
the op-amp to be negligible relative that of the source.
The best bjt op-amps are down in the 50 ohms or so range
of equivalent input referred noise.
Observe that the opamp has no way of knowing if noise generated
in the two 330 ohms resistors is signal or noise.
This circuit would be death of an MC phono stage where
the Z_source is much lower than 660 ohms and the noise
added by the two 330 ohmers would dominate the noise generated
in the signal source itself.
A bjt input stage in the phono preamp has the often
nasty habit of acting as a rectifier for RF that might leak in.
So in addition to picking your op-amp carefully I'd also
suggest you play with lowering the noise here by
replacing the 330 phono-in res. with a few ferrite beads
or maybe a value down in the tens if you can get away with it.
I'd also experiemnt with
lowering the 330 ohms res. in the feedback ground leg but
this of course will affect other circuit parameters.
>
>
> | \
> PHONO IN ---- 330 ----------| + \
> | | \
> | | -------- OUT
> | | / |
> ---| - /
> 47K | | / 1K8
> | | |
> | ---------------
> | |
> | 330
> | |
> 0V 0V
>
>
john
Monte P McGuire
john 015 <CC01...@brownvm.brown.edu> wrote:
>In article <3v86og$e...@pheidippides.axion.bt.co.uk>,
>jeffre...@bt-web.bt.co.uk (Michael Jeffreys) said:
>>
>>Further to my requests for knowledge on 'super' opamps, it is
>>obvious that I must know the source resistance. Can anyone
>>work it out for me from the diagram below showing the initial
>>stage of the MM phono input to my ARCAM Delta 290.
>
>be written R_dc + jwL with L rather big. I have no numbers
>for MM cartridges but speculate maybe 500 ohms + huge L
>terms which could bring you up towards 10 kOhms at 20 kHz.
>
>(This is a bit of a dilemma in especially tape-heads pre-amps
>where the source can start out at maybe 180 ohms at low freq
>and exceed 10 kOhms at high but audible frequencies).
I'd like to further add that only resistance actually generates noise;
inductive or capacitive reactance itself is noiseless. However, this
reactance, when driven by the op amp's input noise current, will
result in a noise voltage proportional to its magnitude.
Why make this distinction? Because different types of op amp input
stages offer different amounts of input voltage noise and input
current noise. High current bipolar input stages of amps like the
LT1028 provide low (1nV/root Hz) input voltage noise density, but also
have high (1pA/root Hz) input current noise density.
Large geometry JFET input stages of amps like the OPA627/637 offer
higher voltage noise (5nV/root Hz) but they offer extremely low
current noise (1.6fA/root Hz) and may perform better in a circuit that
has a large reactive input impedance.
As an example, let's add up the noise from a mythical cartridge and an
op amp. Assume a 500 ohm cartridge with enough inductance to bring it
to 5K at 1KHz. We'll include the 330 ohm resistor and use an LT1028
amplifier in a gain of 7 noninverting circuit.
So, we have the following noise sources, all to be added in an RMS
fashion:
Real component of impedance of cartridge plus 330 ohms = sqrt(4KTRB)
where K = 1.38E-23, T = 300 degrees K, R = 830 ohms, and B = 20KHz
bandwidth => .524uV
Real component of inverting input source impedance (278 ohms) => .303uV
Voltage noise of op amp: 1nV/root Hz * sqrt(20KHz) => .141uV
Current noise into noninverting source impedance: 1pA/root Hz *
sqrt(20KHz) * 5K ohms => .707uV
Current noise into noninverting source impedance (278 ohms) => .0393uV
so, sqrt(.524uV^2 + .303uV^2 .141uV^2 + .707uV^2 + .0393uV) => .940uV
* gain of 7 => 6.60uV total output noise for the LT1028 circuit.
If we use an OPA627, we get the following numbers: .524uV +RMS .303uV
+RMS .707uV +RMS 1.13nV +RMS 62.9pV => 6.52uV total output noise,
which is actually lower than the LT1028, but not by that much.
I must also point out that the above analysis is incorrect in that the
frequency dependent effects of the inductive reactance are not
properly taken into account over the entire frequency band: I
calculated the noise at 1KHz only and used that number for the
broadband noise. However, the results are not that far off.
The other point worth making is that in the case of the LT1028, the
dominant noise term depends upon the inductance of the source and
since that impedance rises with frequency, the noise related to that
impedance will be mostly concentrated in the HF region. In the case
of the OPA627, the noise contribution related to the input reactance
is low and so the output noise spectrum will be basically flat and
thus less annoying to the ear; it is dominated by white noise from
resistors and the 'close to white' noise from the op amp itself.
Brian Lenharth
: In article <3v8hlt$m...@cocoa.brown.edu>,
: john 015 <CC01...@brownvm.brown.edu> wrote:
: >In article <3v86og$e...@pheidippides.axion.bt.co.uk>,
: >jeffre...@bt-web.bt.co.uk (Michael Jeffreys) said:
: >>
: >>Further to my requests for knowledge on 'super' opamps, it is
: >>obvious that I must know the source resistance. Can anyone
: >>work it out for me from the diagram below showing the initial
: >>stage of the MM phono input to my ARCAM Delta 290.
: >
: >So your source impedance can probably
: >be written R_dc + jwL with L rather big. I have no numbers
: >for MM cartridges but speculate maybe 500 ohms + huge L
: >terms which could bring you up towards 10 kOhms at 20 kHz.
: >
: >(This is a bit of a dilemma in especially tape-heads pre-amps
: >where the source can start out at maybe 180 ohms at low freq
: >and exceed 10 kOhms at high but audible frequencies).
: I'd like to further add that only resistance actually generates noise;
: inductive or capacitive reactance itself is noiseless. However, this
: reactance, when driven by the op amp's input noise current, will
: result in a noise voltage proportional to its magnitude.
: Why make this distinction? Because different types of op amp input
: stages offer different amounts of input voltage noise and input
: current noise. High current bipolar input stages of amps like the
: LT1028 provide low (1nV/root Hz) input voltage noise density, but also
: have high (1pA/root Hz) input current noise density.
If you want to get a more detailed analysis of phono stage noise,
check out the section entitled "Magnetic Phono Cartridge Noise Analysis"
in the 1980 National Semiconductor _Audio Handbook_. See also
"On the Noise Performance of a Magnetic Phonograph Pickup" by B.I.
Hallgren in the JAES, Sept. 1975. Note that these were oriented to
MM cartridges, so you'll have to make adjustments if you're interested
in low Z MC cartridges.
bel
--
Brian Lenharth Tele: 206-335-2389
Lake Stevens Instrument Division/ MS10 FAX: 206-335-2828
Hewlett-Packard Co. bri...@lsid.hp.com
8600 Soper Hill Road; Everett, WA 98205-1298
Bruce Stewart
the cartridge impedance) + 278. (the 278 is the parallel combination of 330
and 1.8k)
The cartridge impedance is most likely going to dominate the 47K at low to
mid-band audio frequencies.
_______________________________________________________________
Bruce Stewart bste...@bix.com
>Further to my requests for knowledge on 'super' opamps, it is
>obvious that I must know the source resistance. Can anyone
>work it out for me from the diagram below showing the initial
>stage of the MM phono input to my ARCAM Delta 290.
>
>My guesses are either ~2K or ~47K with a gain of 7.
>
>If the source resistance is less than 2K I think I can use
>some of the ultra low noise opamps LT1028 etc.
>
>
>
> | \
> PHONO IN ---- 330 ----------| + \
> | | \
> | | -------- OUT
> | | / |
> ---| - /
> 47K | | / 1K8
> |
> | | |
> | ---------------
> | |
> |
> | 330
> |
> | |
> | |
> --- ---
> 0V 0V
Bruce Stewart
>The source resistance that the amplifier sees will be (47k in parallel with
>the cartridge impedance) + 278. (the 278 is the parallel combination of 330
>and 1.8k)
>
>The cartridge impedance is most likely going to dominate the 47K at low to
>mid-band audio frequencies.
>
Oops - I overlooked that 330 in series with the input. Make it
(47k in parallel with (cartridge impedance + 330)) + 278. If you're being
really anal, you will want to select that series resistor so that the first
term above is equal to 278 at DC. But the op-amps you are likely to select
aren't likely to require enough input current for this to be important.
_______________________________________________________________
Bruce Stewart bste...@bix.com
PETER LARSEN
TT> Remember, unless you are changing the rest of the circuit (scaling
TT> impeadances, optimizing for Vos, etc), the sound is likely better
TT> with the op-amp which more closely resembles the original part.
Obviously a caveat one should remember, sometimes replacing op-amps will
give funny results ....such results could include a small DC current
through a potentiometer that wasn't meant to have any, and a DBX 119
behaving like a guitar distortion pedal ..... [WHY? .... I'd rather not
put the LM301's back in, should I try with LF356 .... seems like the
'metal box' circuit in it gets confused by OP176's!)
While we are in this context, it happens to be so that the OP275/176 is
designed to be a plug in replacement with exactly the same circuit
requirements as the 553x's according to their Analog Devices application
note, so that point may not be valid as a reason to not replace them. A
point that may or may not be valid is whether it was the capacitor
replacement or the op-amp replacement that mattered most in those
devices with 553'x's replaced ......
TT> Also, the Vos vs Vcm is crucial to linearity. Without typical
TT> curves for this parameter, a single number for CMRR and Vos is
TT> almost meaningless.
Hmmm .... well yes, since you say so, but hey, I'm not a guy with a lot
of formal technical training, so I'll just take your word for it, OK?
TT> Also, to find out what is going on, you may want to look at the slew
TT> ra te vs the freq and signal level in your situation (you may like a
TT> little even order harmonic distortion). The OP-x series of devices
TT> tend to have a lower slew rate.
I did in fact recently say that I think they do have a 'kind of a
sound', and that they are 'sligthly valvy', at least if used in
multitude. There are worse colorations, mostly IC colouration, and I'll
leave it to someone else to say what the distortion kind involved really
is, seem to be in the form of an 'upper midrange, lower treble'
'cloudiness', 'sandiness', 'transient smear'.
TT> There is no perfect op-amp, all have trade-offs. The integration of
TT> th e device and the design are what make exceptional sound.
Of course, no contest, almost anything can be made to sound horrible
with bad enough conditions for it to work in, and most things only
perform to their max potential when correctly and intelligently used.
TT> -tINY