What's the current state of the art in IC OP Amps?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

dpdp

Member
Joined
Mar 11, 2016
Messages
8
I'm wanting to try swapping out op amps in a couple of cheap bits of audio gear. What's a good option for 8 pin socketed OpAmps? A lot of manufacturers and mod shops brag about using Burr Browns. I'm pretty green at modding, so I'm sure I'm leaving out pertinent information in my post.
 
Op amp rolling is rarely if ever the answer, especially if you dont have any issues in mind that you are trying to address. Most Op amps are chosen for specific reasons. Often times that reason is cost, but usually there are additional factors at play. Without understanding those factors, you are shooting blind, and at best you will end up with a piece of gear that likely sounds identical, and at worst with a piece of gear that simply does not function.

I understand the desire to mod things just to see if it can possibly be improved upon, but even in those cases its almost always better to pick one area where you feel the gear is lacking, and start researching how to address that, specifically. Its also worth noting that this approach comes with the added benefit of actually learning something when all is said and done!
 
Thanks for the reply. I'm mainly looking for clarity. I would think that's related to slew rate for transient response as well as thd and noise. When I changed out the op amps in my Bellari Preamp around 10 years ago, it made a tremendous difference. I chose that based on some other modder's suggestion who said it was lower noise and faster than the originals. So, lowest noise with highest slew rate is maybe what I'm looking for. I'm not sure what amps are in the Pro VLA II since I'm not near mine, but I know they are replaced in mods by Revive. I will have to put a socket on the pcb so I can try out different flavors. I was thinking about also upgrading the op amps in my Bellari while I was at it, right now it has the AD823ANZ dips I put in there a while back.

I really am just looking for suggestions on more recent op amps that I may try in these items. I know it's time consuming to burn in the op amps just to see if it sounds ok, but that's my plan. They're both 2 channel devices, so it will be easy to AB.

Is there going to be a different operating impedance for the different circuits, or is that a standardized thing? If not, I should be able to get the impedance specs of the existing amps, maybe? What other specs should I pay attention to as far as making sure it will even work in the circuit?

Edit: I thought I'd come back to say that you're saying that I should choose an area to improve upon, but in low cost prosumer gear there ARE general upgrades that can invariably improve the audio path like replacing the cheap caps with Panasonics and popping in higher quality op amps (if you know how to chose the right ones).
 
Last edited:
I do not advocate design by listening tests, or data sheets... Different op specs to optimize were pursued decades ago... I recall when the trend was for high slew rates (80s)... there were audio power amps that could double as RF radio transmitters but were still only so-so for audio.

There are objective bench test metrics for clarity... like S/N, frequency response, distortion, etc.

I have designed many SKUs over the decades and would only consider op amp upgrades for a handful of difficult sockets.

Good luck... measure twice unsolder once....

JR
 
but in low cost prosumer gear there ARE general upgrades that can invariably improve the audio path like replacing the cheap caps
Yeah there are some goofy caps in things sometimes. On the flip side, I've found myself scavenging some previously modded stuff for some nice polys that really didn't need to be there when working on something else..lol
 
If you mention the specific opamp you want to replace that would help with the suggestions. Generally you want to consider the voltage and current requirements and replace bipolar with bipolar and fet with fet.

For some general options OPA161x series and OPA164x series are well regarded newer opamps, though SMT only.
 
I'm guessing the Bellari sounded absolutely the same. Your brain was told what the differences would be by the other modder so that's what your brain heard.
Not necessarily true. No direct experience with that unit, but have modded other gear and often the difference is clearly audible. It won't turn low end into high end, but can be a cheap and easy way to make improvements.
 
I'll note that -
- if you're determined to stick with DIP8 ICs then you're choice will be limited given that many newer op amps are available in SMT only.
Adaptor pcbs SO8-DIP8 are the answer here.

-if you go popping in 'fast' opamps then you may get into trouble - instability / oscillations - if the circuit isn't up to it in terms of decoupling and layout.
And you may not hear the instability itself as it will likely be above audible frequencies but it is bad for anything downstream.

- as has been indicated you really need to take account of the impedances involved and basically stick with existing FET or Bipolar.

Personally for 'clarity' improvements I'd be looking at any ac coupling caps first.
 
Thanks for the reply. I'm mainly looking for clarity. I would think that's related to slew rate for transient response as well as thd and noise. When I changed out the op amps in my Bellari Preamp around 10 years ago, it made a tremendous difference. I chose that based on some other modder's suggestion who said it was lower noise and faster than the originals. So, lowest noise with highest slew rate is maybe what I'm looking for. I'm not sure what amps are in the Pro VLA II since I'm not near mine, but I know they are replaced in mods by Revive.
There is no justification for replacing a TL07x or a 5532/4 based on slew-rate only in linear audio circuits.
A TL072 swings from -15 to +15V in about 3 us. Put it in perspective with a digital signal sampled at 96kHz that takes about 10us to do the same, or a 20kHz sinewave that takes 25us.
Now for ADC's a higher speed is required because the ADC outputs very high frequency components and the opamp must behave linearly.
Is there going to be a different operating impedance for the different circuits, or is that a standardized thing?
Impedance of the circuit is a major factor in selecting opamps. Some circuits need to be paired with very low offset opamps; one must make sure the combination of bias and offset curents with the circuit's impedance does not result in drift from desired operating point.
Noise is also very dependant on impedance.
Bipolar opamps are more suitable for low-Z circuits, when BiFet opamps are more appropriate for Hi-Z. You must learn about OSI (Optimum Source Impedance).

If not, I should be able to get the impedance specs of the existing amps, maybe?
The commonly available impedance specs are not very useful for selection. The most relevant specs are bias current, offset current, noise voltage density and noise current density.
 
Last edited:
I'm guessing the Bellari sounded absolutely the same. Your brain was told what the differences would be by the other modder so that's what your brain heard.
And I'm guessing there are a bunch of cynics on this board, eh? The difference it imparted was like going from a ribbon trash mic to a C414. Huge. The transients were clearer and the low end was tighter. I think I still have the originals around, so maybe I can upload an ab test. The difference can be seen on an osciliscope.
 
In general, if not full agreement with 'Abbey Road' here, with the observation that 'cheap' gear may have more structural issues than simply the op amps themselves. Poor power rail 'hygiene' and grounding between sections of the unit can make the difference between an ancient 'low quality' op amp and the latest 'world beating wonder' almost impossible to tell apart. From this, each stage must be optimised individually but also keeping a close eye on the performance of the whole unit, and if it has multiple audio paths, does the 'improvement' on one path have a detrimental effect on other path(s). Dumping half wave rectified audio into a power rail or subcircuit ground because you fitted a new 'high power wide bandwidth chip may have unforeseen side effects.
Using a bipolar op amp when fed from a pot for example without input coupling capacitor can lead to scratchy noises especially if the gear gets hot in normal operation even though it was 'fine' when it was assembled as a single unit and was tested with the top off allowing it to remain near to 20 centigrade, whereas in eventual use it may run at 60 Centigrade plus which causes various parameters to drift.
 
And I'm guessing there are a bunch of cynics on this board, eh?
You may call them cynics, they think they are objectivists. Admitted, most of the members here tend to want measurements confirming subjective appreciation.
There are so many "gurus" that claim stellar improvements that just turn out to be unsubstantiated.
The difference it imparted was like going from a ribbon trash mic to a C414.
I've often heard that, many times, the magic that took off the blanket off the speakers.
Huge. The transients were clearer and the low end was tighter. I think I still have the originals around, so maybe I can upload an ab test. The difference can be seen on an osciliscope.
It just shows how poorly designed (or defective) the unit was from the start.
 
Personally for 'clarity' improvements I'd be looking at any ac coupling caps first.
Thanks. Your post is full of good information. How do you choose cap size? Are you sizing it to filter out ranges of frequencies similar to a low pass filter?
If you mention the specific opamp you want to replace that would help with the suggestions. Generally you want to consider the voltage and current requirements and replace bipolar with bipolar and fet with fet.

For some general options OPA161x series and OPA164x series are well regarded newer opamps, though SMT only.
Tons of great info in your post. Thanks. The Bellari currently has an AD823ANZ which appears to be a jfet type. I'm not 100% sure on the Pro VLA II, but the first rev. used NA5532 on the audio path. TL074s are used in the detector circuit, but I don't think I'll mess with them.
 
Thanks. Your post is full of good information. How do you choose cap size? Are you sizing it to filter out ranges of frequencies similar to a low pass filter?
With an existing circuit it's difficult to know if a coupling cap is being used as an intended filter component or just to block dc.
I'd focus on electrolytics in there. Hopefully these are being used simply to block dc with any intentional filtering done with caps that are better for that.
Having the pass frequency point associated with the electrolytic cap as low as practicable is advised. Taking an audible lower frequency of 20Hz, a decade below has often been suggested - so 2Hz. But I'd say go another decade to 0.2Hz. Low frequency minimises phase and level issues in the audio band (and bear in mind that any effects here cascade with serial stages) and the larger the cap the less voltage across it at any given frequency. Lower voltage across cap equates to lower distortion.
Now you may want to ask about actual values - but that depends on the values of associated components. Generally you would need to identify the relevant resistor(s) that form a RC product with the cap'.
Alternatively just use the largest value that will fit ! (checking that you have overrated the working voltage).
 
does the 'improvement' on one path have a detrimental effect on other path(s). Dumping half wave rectified audio into a power rail or subcircuit ground because you fitted a new 'high power wide bandwidth chip may have unforeseen side effects.
A beautiful bit of info. Thank you for that.
 
And I'm guessing there are a bunch of cynics on this board, eh? The difference it imparted was like going from a ribbon trash mic to a C414. Huge. The transients were clearer and the low end was tighter. I think I still have the originals around, so maybe I can upload an ab test. The difference can be seen on an osciliscope.
It's not cynycism. It's science. Every "audiophile" in the world swears their stereo sounds better with $1000 RCA cables and a similar opamp upgrade. But unless the listening tests were double blind the results are worthless.

Our ears are the weakest link in any test chain, and our brains lie to us.
 
If you're looking for what the major designers and vendors of current chips would call "state of the art" (TI, ADI, etc.), then good luck finding it in a DIP.
 

Latest posts

Back
Top