Big test of DIP dual op-amps for audio: LM4562, NE5532, OPA2134, TL072 and LM1458

[EERecordist]

@pma what software are you using for your tests?

The TI OPA1612 is what is used in the Cosmos E1DA.That one would be interesting to test.

I was looking on the TI website at that. TI has small text fixtures, AMP-PDK-EVM to hold some of the surface mount packages for testing. They aren't cheap with a main board, then daughter cards - about $200 for both. Several other companies have inexpensive test sockets for bare parts, no soldering.

TI has great developer documentation and an engineer-to-engineer support forum E2E, for instance https://www.ti.com/product/OPA1612#support-training and a larger https://e2e.ti.com/support/audio-group/audio/f/audio-forum.

 

[pma]

Yeah, I know. I also have some evaluation boards from them, namely LME49720 and OPA1622 EVM. But I still needed to design my own ones. There are too many different applications and results from one scheme are not simply transferable to another ones - like different gain, filters, I/V, high/low CMV etc. One may be surprised easily. It is necessary to build and check.

 

[wwenze]

The most valuable takeaway is not the relative performance of the op-amps, but

1) Finding out most op-amps perform pretty much practically identical to at least 0.001% and low output voltage, despite being dropped into the same circuit without modification

2) Showing that different circuits have different "best op-amp", thus demonstrating that technical-performance questions asked without knowing what they are asking is pointless

 

[pma]

@morillon : Every little thing matters in wiring, grounding etc. You sent me an example of HF spuriae from one of the ASR measurements, in a PM, from another thread. I can show you something similar.

When I take my test rig "as is" for the Gain 11x measurements, I get this:



- and I do not like it. You can see that mess - 19kHz pilot from FM, and mains multiples, though small. I know this is all wrong and I cannot digest it.

Something had to be done. See the photo below:



The red 100pF WIMA polypropylene capacitor marked in the yellow oval. It connects input ground with output ground of the RCA connectors. One of them is isolated to prevent the ground loop. But - the isolation is wrong regarding HF EMI. So, put the cap there. And the result? Below:

Quite nice spectrum, isn't it? Mains spuriae are gone, 19kHz pilot is gone. So it goes. Audio is 20Hz - 20kHz, but is affected by whatever EMI interferences.

THD vs. frequency measured at 7.5V is also very nice. What a nice part:

 

[pma]

Samuel Groner on LME49860 (LME49720, LM4562 equivalent):

This is exactly what we have revealed here. The test circuit from post #1, with +6dB gain, has common mode voltage of 1/2 Vout output voltage. Thus the common mode distortion is visible and LM4562 failed vs. NE5532. But, in the test circuit with gain +20.8dB (post #64 and #46), the common voltage is only of 1/11 Vout, so the CM distortion is suppressed.

 

[pma]

One more comparative, THD vs. frequency at 7.5V, load 2kohm, gain 11x. LME4562 is the clear winner in this test.

 

[dfuller]

So, the much-maligned NE5532 redeems itself.

Sonically the 5532/5534 is totally fine. But it does present some limitations:

1. Its input impedance is pretty low, owing to its bipolar inputs. For mic level this isn't a huge issue, but it can present problems at line or especially instrument level.
2. It idles quite hot. A few of them it isn't a problem but with something like a large format console channel strip, each strip can easily dump 30w of heat at idle.

 

[pma]

2. It idles quite hot.

It is an interesting observation, to me. Supply current is 4 - 8 mA according to the datasheet. For OPA2134, it is 4-5 mA per one amplifier. LM4562 says 10-12mA total quiescent current. I have not noticed any special heating. Are you sure the parts were not oscillating?

 

[Cbdb2]

a large format console channel strip, each strip can easily dump 30w of heat at idle.

Not at idle. A 72 input strip SSL 4000 series power supply puts out 1500watts max, thats 20w per strip, if you dont count the mix busses and output section, with every EQ, compressor etc in the console running a +3vu signal.

 

[DualTriode]

To follow your comment and my previous explanation (DAC output stage rated SE voltage is 2Vrms), I have modified my test circuit from post #1, changing R2 from 10k to 1k

The gain is now 11x, or +20.83dB, as you like it. Load is 2kohm now. Please note that R5 creates a voltage divider with 2k load impedance. Output voltage is measured behind the R5. Right at the op-amp output the voltage is 1.1 x higher.

Now I stand corrected and cannot fully support this previous post of mine:

Though the NE5532P was a better performer in the +6dB test circuit (with 10k/10k feedback) than LM4562, it is not so in case of the test in Topping D10s socket and also not the case with gain 11x and output up to clipping.

The result with gain 11x (+20.83dB) for NE5532P, LM4562 and Burr-Brown OPA2134 can be seen below:

The LM4562 dominates above NE5532 and OPA2134 at higher output voltages and gain 11x. NE5532 is still very good performer, but LM4562 is excellent here.

THD vs. output voltage at 10kHz, LM4562 dominates again.

Note: This E1DA Cosmos ADC is an excellent thing, thank you @IVX for making it available to us engineering audio hobbyists. I wished I had something like this 20 years ago, when it was not possible to reveal such differences between the op-amps without having at least AP2 or SYS-2722.

Thanks for the additional higher voltage plots, a skillful demonstration of what can be done with E1DA Cosmos ADC and ARTA STEPS.

I am pondering of similar acoustic measurements with a OP-Amp pre-amp into a powered speaker and measured with a microphone. I am curious to see if the differences in the op-amp measurements will still be evident in the speaker output.

Thanks DT

 

[dfuller]

It is an interesting observation, to me. Supply current is 4 - 8 mA according to the datasheet. For OPA2134, it is 4-5 mA per one amplifier. LM4562 says 10-12mA total quiescent current. I have not noticed any special heating. Are you sure the parts were not oscillating?

Yes,

Not at idle. A 72 input strip SSL 4000 series power supply puts out 1500watts max, thats 20w per strip, if you dont count the mix busses and output section, with every EQ, compressor etc in the console running a +3vu signal.

A 60 channel Neve 88RS dumps 1800w of heat just sitting there. AFAIK, they're mostly 5532s/5534s.

 

[pma]

I am pondering of similar acoustic measurements with a OP-Amp pre-amp into a powered speaker and measured with a microphone. I am curious to see if the differences in the op-amp measurements will still be evident in the speaker output

It will not be measurable in the acoustical output, by microphone. If we exclude 1458 or 709. It may be measurable on speaker output of the amplifier, electrically.

 

[Cbdb2]

Yes,

A 60 channel Neve 88RS dumps 1800w of heat just sitting there. AFAIK, they're mostly 5532s/5534s.

The idle current of 5532/34s is not that high. How many opamps in one of those, and how much power gets used elsewhere?

 

[pma]

uA709 in a hybrid op-amp

This might be fun and also a kind of surprise. We had very limited access to western IC's back in the seventies and eighties of the past century. One of the ways was to develop hybrid circuits. This is one of them, it is based on old uA709, Tesla had a technology to produce it here as well. We all know limitations of this op-amp. So, when we needed a fast opamp with very high slew rate, a hybrid circuit based on uA709 was developed. It looked like this:

It had slew rate of 250V/us, compared to some 0.7V/us of the uA709. But it was never intended for audio. However, it has been a temptation to check it. Below is THD vs. frequency plot at 2V and @BW45kHz. Gain -1, inverting, 5k/5k.

we have higher noise, however very low distortion ...

19+20kHz IMD:

... and 100kHz (one hundred), 21.8Vp-p square:

 

[abrcdbra]

I can't understand why research ancient and bad-sounding OPAs?
Nowadays there are a very large number of good-sounding OPAs in the DIP case.
Which are quite inexpensive. It is even more profitable to use OPAs in modern cases and install them on DIP adapters.

 

[Sokel]

.
Which are quite inexpensive. It is even more profitable to use OPAs in modern cases and install them on DIP adapters.

Which can totally mess them up as the distances to their caps are more than crucial.
Adding the adapters translates to miles for some of them.

 

[pma]

I can't understand why research ancient and bad-sounding OPAs?
Nowadays there are a very large number of good-sounding OPAs in the DIP case.
Which are quite inexpensive. It is even more profitable to use OPAs in modern cases and install them on DIP adapters.

To show that op-amps usable for analog audio path have not improved much in the past 40 years.
DIP adapters are the worst way to use op-amps in SMD packages. Advantages of bypass capacitors close to the pins are lost and some of them may oscillate due to added inductances with the adapters.

 

[pma]

There is a very good article in AudioXpress by Douglas Self with op-amp measurements:

Modern Op-Amp Distortion Tests: Part 2 — A Few More BJT Types and JFET Devices

Many new op-amp types have appeared in the last decade, and Douglas Self thought it was time to evaluate those for distortion performance. In Part 1 of his article, he measured BJT-input devices using a similar protocol for five different loads and testing for basic op-amp distortion in...

audioxpress.com

 

[DualTriode]

Went looking at non-inverting op-amp circuits, just like the one in the Douglas Self link above.

Picked out 2.15 kR and 1 kR resistors for 3.15 times gain ~ 10dB gain.

Built a Pete Millett op-amp experimenters PCB and found that capacitors block my fat digits from accessing the op-amp socket on the pcb. Move the capacitors to the backside of the PCB.

I will post some test results.

DT

 

[SirPaulGerman]

There are questions if it matters which type of op-amp is used in the audio amplifier and my answer is that it depends on circuit schematics and op-amp parameters. Some of those parameters are obvious from datasheets, but others may be quite hidden and not clear from the first view. For the test I have chosen well known dual op-amps LM4562, NE5532, LM1458 (bipolar input) and OPA2134, TL 072 (JFET input).
LM1458 is one of the first op-amps usable for audio and it is dated at sixties of the previous century. NE5532 was then a big big improvement in audio op-amps, introduced by Signetics company in 1979. LM4562 appeared in 2006 as an improvement in noise and supposed to have lower distortion than NE5532.
OPA2134 was introduced by Burr-Brown in the nineties as a lower noise and lower distortion improvement of TL072.
As far the datasheet and popular info.

When making a choice of the op-amp, we always need to know the application circuit. For the test I have chosen quite common non-inverting amplifier circuit with +6dB gain and 10k feedback resistors.

View attachment 438490

The test rig is driven from a DAC with low, 20 ohm output impedance, and loaded with 1 kohm ADC input impedance. This is quite difficult load, but as the output voltage did not exceed 3V, all the op-amps are deep in the area of allowed output current.

Measurements

For the start, I have chosen measurements of THD vs. output voltage at 1kHz and 10kHz and measurements of 13+14kHz CCIF IMD vs. output voltage. The plots are shown below:

View attachment 438493

View attachment 438494

View attachment 438495



We can see that the LM1458 is simply unusable for audio, even at low output voltage and that we can assume it will have its own sound signature. Both JFET opamps, OPA2134 and TL072 have higher distortion than LM4562 and NE5532, in all 3 tests. The result of the old NE5532, that is slightly better than LM4562, is a small surprise to me, I have expected the opposite result.

In all cases, I have tested at least two samples of each op-amp type, to prevent random results. All the samples of the same types behaved identically within the measurement repeatability.

More tests can be done, suggestions welcome. But I will wait for the forum feedback.

All the measurements done with 96kHz sampling and 45kHz effective bandwidth.

Note: for loopback test please go to post #33

Can you test the TLE2072 which is a upgrade over the tl072?