Op Amp upgrade

[seamonkey]

I love reading these posts about swapping opamps to get a phatter/brighter/darker/wibblyer/whatever tone.

Anyone who has done even basic opamp theory will know that the frequency response curve (let's call it tone) comes from what you put round them not from the opamp itself. They are a pretty sterile and predictable device (as long as you operate them within their design specification).

Output offset - almost every stage is AC coupled so a few mV offset means nothing, you don't hear offset unless it is large enough for the output to hit one of the supplies (clip).

The only thing you will improve is the noise that the opamp is responsible for, but as has already been pointed out the other components probably contribute far more noise than the opamp.

I'm with you on this one. Big changes easily recorded on a PC or a small handheld recording device. Somehow this is never done. I just can't pull myself to believe in massive improvements. It's like on the cooking shows they always taste what they just cooked and say "Yumm". They'd never spit it out and say "Gawd that's awful".

I hope that people don't think they need to shop out and replace op-amps just because they are on sockets. Heck probably just pulling out and re-seating some op amps help with some cheap sockets with crusty contacts. It would be better to just remove the sockets and directly solder the op-amps.

 

[12bass]

... There's a middle ground between specs and listening tests, which is to use measurement instrumentation.

With modern parts like the LM4562, it can be very difficult to measure the vanishingly low distortions produced by the op amp... yet they still sound audibly different than, say, an AD797 or OPA211, which also have vanishingly low distortion. This is what leads me to suggest that specifications and measurements do not tell the whole story.

That is not to say that I do not think that audible differences are unmeasurable in principle, though... just that the connection between psychoacoustics and engineering may still have a ways to go in terms of sufficiently connecting auditory perception with measurements.

Some useful op amp testing done here, which shows that measurements in circuit are not what is suggested by specs alone: Link Removed

 

[lowfreqgeek]

That's cool. I've done rudimentary PIC and 8051 stuff, and have only recently converted to coding in C. I'm a physicist, but am also interested in electronics. At my day job, I get involved in "physics meets electronics" situations, such as noise analysis of optical detectors.

Are you at a company, or doing academic research? It seems to me that solid knowledge of both the analog and digital domain is a pretty useful combination.

I'm with FLIR Systems, formerly ICx Technologies, formerly Mesosystems Technology, and so on. Bio-threat detection is our area of expertise. I tend to do more of the critical analog stuff, but also dabble with FPGA/uC-based PWM controllers, motion control, various power supply topologies, batteries/charging, system design, algorithm development, etc. I wish I could enjoy programming more, but I'd much rather build build analog circuits as they are more mysterious and fun.

The one area of digital that has me really intrigued is digital audio. Of course, there's the analog side of getting maximum dynamic range out of the converters, but I'm also interested in the DSP aspects of hardware modeling. Combining the analog hardware characterization w/ the mathematics and DSP-realization is pretty interesting.

Back to my nightmare...

 

[iplay5strings]

One other thing I'm gonna go ahead and throw out there...when using swr (goliath) cabs (especially w/unmodified swr amps) people tend to describe the cabs themselves as sounding "scooped". What most people don't know is swr purposely wires their tweeters out of phase with the speakers, but not to remedy a phase problem in the cross-over. By wiring the tweeter out of phase with the speakers it creates a "scoop" in the sound as the speakers push out and the tweeter pulls in, thus tuning the cab to mimic the sound of the passive tone stack (aural enhacer) built into their preamps. They "color" their cabs with the same signature tone as they do their preamps!

 

[lowfreqgeek]

You can do more with high voltage and carbon composite resistors
http://www.geofex.com/article_folders/carbon_comp/carboncomp.htm

There's the Mojo

I like the Mojo!

That does not surprise me at all, though. Think about most "vintage" gear with mojo - carbon comp resistors, paper/oil caps, iron, high voltages, etc... All pretty "dirty" by today's standards.

Thanks for the link.

 

[12bass]

I'm with you on this one. Big changes easily recorded on a PC or a small handheld recording device. Somehow this is never done. I just can't pull myself to believe in massive improvements. It's like on the cooking shows they always taste what they just cooked and say "Yumm". They'd never spit it out and say "Gawd that's awful".

Straw man.

While making modifications to my Echo Gina24, I made RMAA analog loopback measurements both before and after. Among other measurements, both IMD and THD were significantly lower using LM4562 compared with the original JRC2068 op amps. In general, I try to use a combination of empirical testing and listening.

I hope that people don't think they need to shop out and replace op-amps just because they are on sockets. Heck probably just pulling out and re-seating some op amps help with some cheap sockets with crusty contacts. It would be better to just remove the sockets and directly solder the op-amps.

Not something I'd recommend. I use high quality sockets with gold contacts.


BTW, as noted above I've found that the quality of the analog path has a very noticeable (and measurable!) impact on A/D and D/A converters. Then there's the matter that not all converters sound the same.... even if the specs say they are essentially "perfect".

 

[delta7fred]

Atmel ATXMega in C. Pretty much a staple around here, but I'm a dyed-in-the-wool analog EE, so I much prefer the nuances of current sources and sine generators and transimpedance amps, etc.

OT I know.

I just bought a JYETECH capacitance meter (board) based on that chip, superb little piece of kit. 1pf - 500uf auto-ranging. Is there a protection circuit to stop you reading the code back (as there was on the PIC chips)?

It is something I fancy getting back into, I did a bit of PIC coding when they first came out. You can tell how long ago it is, my programmer would only run on a 386 as the 468s were too fast.

 

[rust_preacher]

..and I did my first programming on a PDP-8...

 

[aborgman]

Atmel ATXMega in C. Pretty much a staple around here, but I'm a dyed-in-the-wool analog EE, so I much prefer the nuances of current sources and sine generators and transimpedance amps, etc.

End app is particle concentration and biological aerosol collection/detection.

Arduino or rolling your own PCB? I just did a PCB for an ATMEGA system.

 

[lowfreqgeek]

Arduino or rolling your own PCB? I just did a PCB for an ATMEGA system.

Oh, I ALWAYS roll my own... [pcbs, that is].

I don't get to do as much layout as I used to, but I do design all my boards from scratch and work with a layout engineer to get the PCB done.

I wish I had the time to develop some of my own audio ideas (including AD/DA and current summing amps), but I'm not about to give up playing as much as I can (several times/week if I'm lucky).

 

[seamonkey]

Straw man.

While making modifications to my Echo Gina24, I made RMAA analog loopback measurements both before and after. Among other measurements, both IMD and THD were significantly lower using LM4562 compared with the original JRC2068 op amps. In general, I try to use a combination of empirical testing and listening.



BTW, as noted above I've found that the quality of the analog path has a very noticeable (and measurable!) impact on A/D and D/A converters. Then there's the matter that not all converters sound the same.... even if the specs say they are essentially "perfect".

Sure listen, but anything you can hear can be recorded. The original poster in this thread didn't record a before and after new op amps. It's just anecdotal if swapping op-amps helps with the SWR.

Loop back is a good way to test. I do that when testing pedals and such. I do two channels, one looped through the pedal, one looped through a cable. Measure, and can zero out any noise in the interface. I do a couple of measurements. Often there are surprises like how distortion is really present when it still sounds clean. I admit I don't have golden ears.

I haven't noticed much in the D/A direction. It all seems to sound the same on the way out. Using a DAW I keep everything digital until the last instance anyhow.

 

[iplay5strings]

Sure listen, but anything you can hear can be recorded. The original poster in this thread didn't record a before and after new op amps. It's just anecdotal if swapping op-amps helps with the SWR.

Loop back is a good way to test. I do that when testing pedals and such. I do two channels, one looped through the pedal, one looped through a cable. Measure, and can zero out any noise in the interface. I do a couple of measurements. Often there are surprises like how distortion is really present when it still sounds clean. I admit I don't have golden ears.

I haven't noticed much in the D/A direction. It all seems to sound the same on the way out. Using a DAW I keep everything digital until the last instance anyhow.

Like I said after doing the mods...the op amps mainly lowered the noise. As I stated I A/B'd the TL082's and TLE2072's in the sockets behind the inputs of the amp and the noise was OVERWHELMING with the TL082's...was SO happy with the noise level of the TLE2072's. I did notice a "cleaner" sound when using the EQ section, but mainly because the TLE2072's made it operate without so much inherent noise in the signal.

 

[seamonkey]

Alas, no recordings.

 

[iplay5strings]

Alas, no recordings.

What do I need recordings for? The op amps did what I wanted them to do, make my amp quieter (which could/should have been predicted based on the "all knowing" scientific specs). Whether or not they had any "real" effect on the tone is purely academic and unimportant to me.

 

[seamonkey]

Could be real, could be a placebo effect. If you're happy - no problem.
I would hope that anybody who would think they need to change op-amps should take a more objective view.

 

[gaffster787]

In a sense an op amp is at the mercy of its environment..
It's decoupling from it's own rail(s), the nature of
it's feedback loop, the range of its input voltage,
the circuits around it, and the load it must drive...

For audio..a slower amp may very well be quieter...but
In a perfect audio (practically dc) world, if applied correctly, it should not matter at all....

 

[David Jayne]

What most people don't know is swr purposely wires their tweeters out of phase with the speakers, but not to remedy a phase problem in the cross-over. By wiring the tweeter out of phase with the speakers it creates a "scoop" in the sound as the speakers push out and the tweeter pulls in, thus tuning the cab to mimic the sound of the passive tone stack (aural enhacer) built into their preamps. They "color" their cabs with the same signature tone as they do their preamps!

Actually, no. You can wire them either way(I have) and you won't hear any difference(I didn't). Someone smarter than me can explain why.

 

[iplay5strings]

Actually, no. You can wire them either way(I have) and you won't hear any difference(I didn't). Someone smarter than me can explain why.

I hear a quite a difference. With the the tweeter at the same volume for both, wired out-of-phase, popping is very loud and bright (almost brittle) while fingerstyle sounds a few dB quieter or "distant" (a common sound characteristic of speakers that are out-of-phase with each other). With the tweeter wired in-phase the edge is taken of the tweeter and finger vs slap/pop style sounds even volume wise or more "centered".

 

[iplay5strings]

Just thought I'd give an update after using the modded amp in a few "real life" live situations. I just recently played a larger venue with an unshielded drummer, 2 guitars, keys and around 5 vocals using only a single Goliath III 4x10 run in bridge mono for myself. I have to say that the tone and clarity in a live band setting, as well as studio recording, is PHENOMINAL after the AE bypass mod! Every note is clear, punchy and defined from top to bottom. As far as the op amp upgrade...the idle noise, from the cab and the direct out, is basically non-existent now even with the substantial amount of high end I have rolled up on the high mid knob of the semi parametric eq (I like the Andrew Gouche/Marcus Miller "bitey" tone and dialing in that kind of high end used to make the amp hiss/buzz like CRAZY). I'm so happy with this amp now and the now quiet operation of the amp/direct line makes the sound guy's job quite a bit easier! :-D

 

[will33]

What happened to that one line that used to be on some people's thing that said, " Every time seamonkey posts, somebody buys a tube amp."?