Good morning,
I have an opamp in my hands, the OP27, possibly to be used as a reference for a regulator circuit. I later realized from the datasheet that an example circuit is also a phono stage in a preamp.
I was wondering out of curiosity if anyone has ever used it like this, and if it sounds good despite being a rather obsolete component. Thank you
I have an opamp in my hands, the OP27, possibly to be used as a reference for a regulator circuit. I later realized from the datasheet that an example circuit is also a phono stage in a preamp.
I was wondering out of curiosity if anyone has ever used it like this, and if it sounds good despite being a rather obsolete component. Thank you
In my flat balanced input MM phono preamps I tried several different op amps comparing their noise floor using a real cart as the input termination.
I tried the dual OP27, the OP270, and it performed surprisingly well despite being an input bias current-compensated part.
Using noise as the selection criteria the NJM2068 measured the lowest in an MM preamp with cart input termination. There may be other parts that sim or perform better with a resistor as input termination.
If you use the OP27 I think you'll be happy with it.
https://www.proaudiodesignforum.com...OP27_OP37_Precision_Monolithics_June_1981.pdf
https://proaudiodesignforum.com/forum/php/viewtopic.php?t=753
I tried the dual OP27, the OP270, and it performed surprisingly well despite being an input bias current-compensated part.
Using noise as the selection criteria the NJM2068 measured the lowest in an MM preamp with cart input termination. There may be other parts that sim or perform better with a resistor as input termination.
If you use the OP27 I think you'll be happy with it.
https://www.proaudiodesignforum.com...OP27_OP37_Precision_Monolithics_June_1981.pdf
https://proaudiodesignforum.com/forum/php/viewtopic.php?t=753
Interesting what you say. BTW think I previousy used just the NJM2068 (it covers a quite good current demand) as output stage for a hybrid amp and I've been pleasantly stuck by it.
I'll take for sure a look to the 2nd link to have furher informations about.
I'll take for sure a look to the 2nd link to have furher informations about.
I was also surprised to find the NJM2068 had as much output current capability as it did. My sample measured 52 mA Ipeak.
I did an output current comparison of a number of duals here: https://proaudiodesignforum.com/forum/php/viewtopic.php?t=1289
I did an output current comparison of a number of duals here: https://proaudiodesignforum.com/forum/php/viewtopic.php?t=1289
Yes, from your test emerges that the NJM2068 is a very good chip indeed as me too I discovered on practice. I tried the NJM2114 before, and despite only slightly lower in I(out) I found not so musical and detailed at time. In my opinion a good mate for the former is the LM4562, but with half current drive
Coming with 2 years delay 😊.
I used OP27 as general preamp chip including MM phono preamp some 30 years ago, indeed nice forgotten opamp. Nobody notices that except very good voltage nose and DC precision, it also has better current noise than ne5534... Plus runs up to 22V PS rails (like 5534). Absolutely good to use IMHO.
For high current delivery, another forgotten one is Njm4580, that drives 100R load without sweat but with all other specs better than njm4560, many people criticize it's sound as it is "house" opamp of Behringer, but I couldn't find it sounds bad
I used OP27 as general preamp chip including MM phono preamp some 30 years ago, indeed nice forgotten opamp. Nobody notices that except very good voltage nose and DC precision, it also has better current noise than ne5534... Plus runs up to 22V PS rails (like 5534). Absolutely good to use IMHO.
For high current delivery, another forgotten one is Njm4580, that drives 100R load without sweat but with all other specs better than njm4560, many people criticize it's sound as it is "house" opamp of Behringer, but I couldn't find it sounds bad
0.4 pA/√Hz typical at 1 kHz according to the Analog Devices datasheet revision H, the same as the NE5534A (which is better than the NE5534 without the A). The OP27 has base current compensation, though, and the datasheet doesn't make it clear whether that 0.4 pA/√Hz applies with equal impedances driving the inputs (datasheet condition to cancel common-mode noise from the base current compensation circuit) or with unequal impedances (normal use).
The noise with cartridge is low, as expected and indicated by the datasheet:
There is still some usable loopgain reserve with the RIAA curve.
This was useful PMA, thanks!
Not that I will again use op27 since I moved away from MM carts long time ago, but it is nice confirmation of ancient and than not measured findings. Definitely useful for anyone into MM.
Not that I will again use op27 since I moved away from MM carts long time ago, but it is nice confirmation of ancient and than not measured findings. Definitely useful for anyone into MM.
Who of us ancient EEs would have expected this result? I am simply thrilled - and think back to the good old days when we were all still young and read thick data books and application notes in paper form (and above all understood them, and also believed that there was such a thing as trust back then).
Dieter Burmester made this family of OPs socially acceptable in the 1980s, he had the type plate /label removed ...
Nothing new under the sun, after my 38 years of audio electronics experience, rediscovering this precision operational amplifier is tiring. Don't you agree?
#
Of course, this OP is a first-class representative of his guild, no question, and hopefully I still have a few originals in my personal stock.
Dieter Burmester made this family of OPs socially acceptable in the 1980s, he had the type plate /label removed ...
Nothing new under the sun, after my 38 years of audio electronics experience, rediscovering this precision operational amplifier is tiring. Don't you agree?
#
Of course, this OP is a first-class representative of his guild, no question, and hopefully I still have a few originals in my personal stock.
You know, I may have a kind of know-how, and I do not intend to disclose everything.
I use E1DA Cosmos ADC:
https://www.audiophonics.fr/en/devi...-18348.html?search_query=e1da &fast_search=fs
I use Topping DAC (and USB isolation) as a signal source with output divider to reduce DAC output noise to negligible level.
Dear Pavel,
that is not the point (in no way am I questioning your reputation, quite the opposite).
The point is:
“where is the sketch of the used circuit, that you put in the DUT” ?
No hard feelings,
HBt.
that is not the point (in no way am I questioning your reputation, quite the opposite).
The point is:
“where is the sketch of the used circuit, that you put in the DUT” ?
No hard feelings,
HBt.
Some of the current noise data I have measured in the past for some OP-AMPs, including the OP-27 and some of those mentioned here, are shared for reference.
It may just be that the batch I have is accidental, but TI's NE5534 has a slightly higher current noise with the A suffix.
For this reason, I don't really trust TI's A-suffixes. The NJM5534 seems to have less current noise in that respect. (Unfortunately, I don't have the 5532 A suffix, so I couldn't compare.)
One of the questions raised in the data sheet has already been mentioned here, but another
The datasheet shows that the OP285 is a DC high-precision version of the OP275, which should be basically the same, but there is a big difference in DS values: 0.9 pA/√Hz for the OP285 and 1.5 pA/√Hz for the OP275. Moreover, this value is too large for an OP-AMP without bias current cancellation with a voltage noise density of 6 nV/√Hz. We wondered about this and measured the values, which were still the same for both and much better than the DS values.
Naturally, even with bias current cancellation, those with high beta input seem to have less current noise.
Voltage noise density around 1 nV/√Hz (not suitable for MM)
Of these, the AD797 has the lowest current noise, consistent with this data from PMA. The bias current is 0.25 µA (type) compared to the collector current of 0.9 mA, so it is unlikely that β = 3600, so there must be bias cancellation... but the correlated noise is also low.
The current noise of the LME49990 (discon) is the highest, but the voltage noise is the lowest of these. (My actual measurements were around 0.8 nV/√Hz)
Some other popular ones.
| Substantial current noise density of one-sided input only (A/√Hz) | Common-mode cancelled current noise density (A/√Hz) | |
| OP27 (AD) | 7.40E-13 | 5.32E-13 |
| OPA27 (BB) | 7.50E-13 | 5.27E-13 |
| NJM2068D | 3.59E-13 | 3.41E-13 |
| NJM4580DD | 2.38E-13 | 2.51E-13 |
| NE5532P(TI) | 3.90E-13 | 3.88E-13 |
| NJM5532DD | 3.74E-13 | 3.88E-13 |
| NJM2114D | 4.14E-13 | 4.26E-13 |
| NE5534P(TI) | 5.23E-13 | 5.32E-13 |
| NE5534ADR(TI) | 5.76E-13 | 5.91E-13 |
| NJM5534D | 3.44E-13 | 3.41E-13 |
| NE5534 (SIG) | 4.48E-13 | 4.17E-13 |
For this reason, I don't really trust TI's A-suffixes. The NJM5534 seems to have less current noise in that respect. (Unfortunately, I don't have the 5532 A suffix, so I couldn't compare.)
One of the questions raised in the data sheet has already been mentioned here, but another
| OP275 | 3.10E-13 | 3.01E-13 |
| OP285 | 3.10E-13 | 3.20E-13 |
| OPA209 | 5.32E-13 | 3.80E-13 |
| OPA2210 | 4.24E-13 | 3.13E-13 |
Voltage noise density around 1 nV/√Hz (not suitable for MM)
| LT1028 | 3.59E-12 | 1.82E-12 |
| LT1115 | 3.38E-12 | 1.72E-12 |
| AD797 | 2.53E-12 | 2.01E-12 |
| AD8597 | 3.34E-12 | 1.94E-12 |
| ADA4898-1 | 3.42E-12 | 1.94E-12 |
| LME49990 | 4.52E-12 | 1.68E-12 |
The current noise of the LME49990 (discon) is the highest, but the voltage noise is the lowest of these. (My actual measurements were around 0.8 nV/√Hz)
Some other popular ones.
| LM4562 | 1.25E-12 | 8.92E-13 |
| OPA1622 | 7.50E-13 | 6.84E-13 |
| LM833 | 3.75E-13 | 3.60E-13 |
| LME49723 | 2.72E-13 | 2.78E-13 |
First, great job!
So, looks like current noise wise Op27 is mediocre, 553x middle and jellybean Njm4580 a shining star?
Not exactly inline what PMV measured in his test where op27 is close to 6 times more expensive fet input 627🤔. Could be not all dut-s were representative specimens of their families, or?
So, looks like current noise wise Op27 is mediocre, 553x middle and jellybean Njm4580 a shining star?
Not exactly inline what PMV measured in his test where op27 is close to 6 times more expensive fet input 627🤔. Could be not all dut-s were representative specimens of their families, or?