^ At least below the knee frequency for 1/f. Something like the OPA165X series has an E_n somewhere in the low 5's, but that's above 1 kHz give or take. Don't know how much it grows down low.
There's the OPA209 as well, that I forgot about, has a superbeta pair up front which gives it a low overall E_n/I_n combo for moderate impedances.
They're other options, the one's you mention are inarguably good as well.
There's the OPA209 as well, that I forgot about, has a superbeta pair up front which gives it a low overall E_n/I_n combo for moderate impedances.
They're other options, the one's you mention are inarguably good as well.
CMOS input noise specs are getting better all the time, for instance https://d1d2qsbl8m0m72.cloudfront.n...tasheet/ic/amp_linear/opamp/lmr1802g-lb-e.pdf
knee at 100Hz or so, 2.9nV/√Hz at 1kHz, 7.8nV/√Hz at 10Hz
And the device is 5µV offset (typ)....
But its 5V supply max, and the distortion isn't in the best class, and only the output is fully rail-to-rail,
knee at 100Hz or so, 2.9nV/√Hz at 1kHz, 7.8nV/√Hz at 10Hz
And the device is 5µV offset (typ)....
But its 5V supply max, and the distortion isn't in the best class, and only the output is fully rail-to-rail,
opa op amps
Excellent for upgrading any system.
Hard to find in sa so when there is plenty of scrap lying around i take and strip for such.
I was a nec 5532 fan until i changed them and heard the sound.
Old school cant be rivaled, digital is to clean and sounds synthetic, do yourself a favor and listen to the tone of an old hk pm series or such..blow your mind!
Many people are digging their old systems out and WOW...
Excellent for upgrading any system.
Hard to find in sa so when there is plenty of scrap lying around i take and strip for such.
I was a nec 5532 fan until i changed them and heard the sound.
Old school cant be rivaled, digital is to clean and sounds synthetic, do yourself a favor and listen to the tone of an old hk pm series or such..blow your mind!
Many people are digging their old systems out and WOW...
That's pretty hilarious, along with this later in the lecture:
Bonsai -- I hope I made it clear I was talking about 1/f noise WRT *FET-input gain blocks! Repetition isn't a bad thing, nonetheless!
changed....to what, specifically? And, what was the sound improvement that you heard? The old NE5532 (40+ years later) is still among the very best, and they are a bargain these days!
Opa 189 - jfet input
0.00006%
total thd plus noise at 3.5 volts
– Zero-drift: 0.005 μV/°C
– Ultra-low offset voltage: 3 μV max
– CMRR: 168 dB
– Open-loop gain: 170 dB
• Low noise:
– en at 1 kHz: 5.2 nV/√Hz
– 0.1-Hz to 10-Hz noise: 0.1uv p to p
– Gain bandwidth: 14 MHz
– Slew rate: 20 V/μs
– Fast settling: 10-V step, 0.01% in 1usec
0.00006%
total thd plus noise at 3.5 volts
– Zero-drift: 0.005 μV/°C
– Ultra-low offset voltage: 3 μV max
– CMRR: 168 dB
– Open-loop gain: 170 dB
• Low noise:
– en at 1 kHz: 5.2 nV/√Hz
– 0.1-Hz to 10-Hz noise: 0.1uv p to p
– Gain bandwidth: 14 MHz
– Slew rate: 20 V/μs
– Fast settling: 10-V step, 0.01% in 1usec
Last edited:
But not above 1kHz, and certainly not into 2k load or worse above 1kHz
Very good, current noise is fine (but high for a JFET actually).
Sounds great, but the large signal gain bandwidth seems to be very low, it maxes out at 10kHz for a full-swing output at +/-18V rails, see figure 36 in the datasheet http://www.ti.com/lit/ds/symlink/opa2189.pdf
Obviously, the circuit topology will make a difference... however, I've found that that ADA4627-1 sounds quite a bit better than the LM4562, as does the NLA LME49990. OPA2211A/OPA1612 also sound a bit better than LM4562, IMO. ADA4625-1/2 are supposed to be even better than the ADA4627-1, though I've yet to try them. IMO, the ADA4627-1 has a fuller, more "musical", sound than the somewhat thin/sterile LM4562, whilst still retaining nicely detailed highs, and remains my favorite sounding op amp to date.
What does swapping IC's achieve
I am using LM4562 as a line stage amplifier with a gain of 2 in combination with a beefy power supply of 100VA out of an old Cyrus 1 amplifier housed in a separate box.
The regulators in the preamp section are based on a dual op.amp LM833P the variety claimed to have no dead-band distortion.
Silicon Chip an Australian publication claimed to have performed tests on LM833N the standard version in comparison with LM4562 using Audio Precision test gear from which it was claimed the LM833N returned slightly better distortion results.
In recent audio projects this magazine has included ferrite beads on input and output leads for preamplifier circuits and on power amplifier inputs. I have always held doubts about doing this and have been reluctant to try this myself.
Anyway having adopted this practice more recently I am inclined to think that perceived differences between op.amps has much to do with emi in the environment - computers, florescent lighting, digital sources, cell phones, etc etc which can penetrate through RC input filters.
I don't know of any test procedure that can quantify these effects which would vary where-ever you live but some solutions I tried nearly 20 years ago were to glue bits of IC anti-static foam or bits of cork tile with bath sealant to the top of an IC.
I also remember some tips through email contact with Guido Tent who at the time worked as an emi specialist for the Phillips company in Holland.
He reckoned that stability capacitors inside CD players should connect to earth as close as possible to the Chip pin for that purpose and any connecting lengths should be to supply voltage points. This made a significant difference to the sound from my Phillips CD player I had at the time.
I have cut out an article of Guido's that was published in a British audio magazine with more detail on dealing with emi.
A couple of years ago I acquired what I regard as a Classic CD player - the Conrad Johnson Sonographe SD-1. This had a couple of faults that needed sorting. I had heard an example owned by a audio nut friend when this player was new and the experience made a lasting impression.
The player itself was based on a Phillips CD player sold in this country as model CD350 which happened to be my first player acquisition. The analogue stages in this were LM833N dual op.amps one per channel.
The modifications in the Sonographe player were to replace these chips with LF353N one per channel. One section of these chips was used in the implementation which was the replacement by a J232 FET in place of the redundant second filter function of the LF353N plus the inclusion of military grade components along with a power supply using plastic capacitors which was routed back from the additional modification pcb to the Phillips board for the LF353 half sections.
The LF353N has been rated as a device to avoid from a distortion viewpoint, this being due to the harmonic structure of the output signal. With a FET performing the second analogue stages the resultant output signal is much more analogue sounding in my view.
I fitted turn pin sockets in my line and RIAA preamp stages when I built this and used to change IC types all the time. To an extent how one reacts must depend on circumstances - mood, and how alert or stressed one can be etc resulting from everyday living.
I think this sort of process distracts attention away from other causes of dissonance and plugging new IC's in is no longer my thing.
Maybe some keen enough might be interested in trying some of these alternative approaches for themselves.
I am using LM4562 as a line stage amplifier with a gain of 2 in combination with a beefy power supply of 100VA out of an old Cyrus 1 amplifier housed in a separate box.
The regulators in the preamp section are based on a dual op.amp LM833P the variety claimed to have no dead-band distortion.
Silicon Chip an Australian publication claimed to have performed tests on LM833N the standard version in comparison with LM4562 using Audio Precision test gear from which it was claimed the LM833N returned slightly better distortion results.
In recent audio projects this magazine has included ferrite beads on input and output leads for preamplifier circuits and on power amplifier inputs. I have always held doubts about doing this and have been reluctant to try this myself.
Anyway having adopted this practice more recently I am inclined to think that perceived differences between op.amps has much to do with emi in the environment - computers, florescent lighting, digital sources, cell phones, etc etc which can penetrate through RC input filters.
I don't know of any test procedure that can quantify these effects which would vary where-ever you live but some solutions I tried nearly 20 years ago were to glue bits of IC anti-static foam or bits of cork tile with bath sealant to the top of an IC.
I also remember some tips through email contact with Guido Tent who at the time worked as an emi specialist for the Phillips company in Holland.
He reckoned that stability capacitors inside CD players should connect to earth as close as possible to the Chip pin for that purpose and any connecting lengths should be to supply voltage points. This made a significant difference to the sound from my Phillips CD player I had at the time.
I have cut out an article of Guido's that was published in a British audio magazine with more detail on dealing with emi.
A couple of years ago I acquired what I regard as a Classic CD player - the Conrad Johnson Sonographe SD-1. This had a couple of faults that needed sorting. I had heard an example owned by a audio nut friend when this player was new and the experience made a lasting impression.
The player itself was based on a Phillips CD player sold in this country as model CD350 which happened to be my first player acquisition. The analogue stages in this were LM833N dual op.amps one per channel.
The modifications in the Sonographe player were to replace these chips with LF353N one per channel. One section of these chips was used in the implementation which was the replacement by a J232 FET in place of the redundant second filter function of the LF353N plus the inclusion of military grade components along with a power supply using plastic capacitors which was routed back from the additional modification pcb to the Phillips board for the LF353 half sections.
The LF353N has been rated as a device to avoid from a distortion viewpoint, this being due to the harmonic structure of the output signal. With a FET performing the second analogue stages the resultant output signal is much more analogue sounding in my view.
I fitted turn pin sockets in my line and RIAA preamp stages when I built this and used to change IC types all the time. To an extent how one reacts must depend on circumstances - mood, and how alert or stressed one can be etc resulting from everyday living.
I think this sort of process distracts attention away from other causes of dissonance and plugging new IC's in is no longer my thing.
Maybe some keen enough might be interested in trying some of these alternative approaches for themselves.
