| Author |
 Topic  |
|
|
Laurie
Double Platinum Member
Canada
4854 Posts |
 Posted - 09/21/2008 : 18:05:50
|
I've wondering for a while why opamps with only slightly different specs have audibly different tone, even if it is only fractional (zentropa, I blame you for me thinking about this fractional stuff!). Particularly when the differences in specification should inaudible.
As I was looking at the graphs for the TL072 it occured to me that the one material difference is the phase-shift characteristic. I wonder if this accounts for the audible differences? I was lookng at the "LARGE-SIGNAL DIFFERENTIAL VOLTAGE AMPLIFICATION AND PHASE SHIFT vs FREQUENCY" graph.
The distortion, noise, slew rate (etc) difference between the devices should be inaudible, but differences in the phase-shift characteristic would certainly be noticable...
|
Edited by - Laurie on 09/21/2008 23:17:46 |
|
|
zentropa
Gold Member
USA
837 Posts |
Posted - 09/22/2008 : 07:59:20
|
i have no clue why they sound different but they definitely do.
welcome to the OCD-hell i live in every day :P |
 |
|
|
Laurie
Double Platinum Member
Canada
4854 Posts |
Posted - 09/22/2008 : 17:06:24
|
Oh <insert deity of choice>!! I'm stuck in purgatory with zentropa!!
Anyway... take a look at these. Top one is TL072, bottom is RC4558. This will definitely be an audible difference in the audio range. The RC4558 will probably sound smoother because the phase change doesn't plateau.
|
|
|
|
MoonWatcher
Bronze Member
USA
98 Posts |
Posted - 09/22/2008 : 19:33:45
|
I really like the TL072 and the way it distorts - I think the phase swing helps it to sound a little more crisp, although it could prove to be too much for some. It's a real rail bouncer when you drive it hard, too, so that phase swing has got to be audible.
The 4558 has to be one of the most flexible of all of them. I have a hard time getting a circuit with a 4558 to sound really bad due to it's "round" nature. Kinda funny that they're most common in mild OD pedals, since you can get massive round midrange out of one if you drive it really hard.
Funny that the term OCD was mentioned, since that pedal uses a TL082, which I think is the biggest contributor to it's bright aggressive tone.
I'm about to delve back into op amp hell myself - I ordered more chips than I realized. I'm particularly interested in some 4559's that I have, since they are supposed to have a lower noise floor than 4558's. |
|
|
|
nathanscribe
Silver Member
United Kingdom
376 Posts |
Posted - 09/22/2008 : 19:38:34
|
| I have an old Phonic mixer (PMC 1202B) and always thought it sounded a bit woolly... I recently replaced all the 4558s with 072s, and it sounds like I've taken out my earplugs. More sizzle, more clarity - it's opened up really. I've no doubt there better opamps for this kind of thing, but bulk 072/074 are my standard bulk-buy choice. |
|
|
|
ChristoMephisto
Platinum Member
Canada
1288 Posts |
Posted - 09/22/2008 : 23:58:33
|
TL072s sound good in an upgraded GE-7,
4559s were used in the GE-7 Sniper mod due to some special sound I can't remember. iirc shimmering...
Recently fixed a friends BMP (op amp version ) that suffered some damage from a 12v plug. He used a LM1458 opamp, found a JRC4558DD later on and swapped it out, the difference in tone is amazing!
Maybe there is some mojo to opamps |
|
|
|
MoonWatcher
Bronze Member
USA
98 Posts |
Posted - 09/23/2008 : 13:58:19
|
quote:
Originally posted by ChristoMephisto
Maybe there is some mojo to opamps
I hate to admit that there is!
If you think that there are variations with different chips, try stacking them. Better yet, stack two different chips!
The DS-1 seems very dependent on the chip that's in it. The latest NJM chip sounds very good, IMO.
I have a couple of 1458's that I use for reference, and they're really underrated, IMO. |
|
|
|
Racer X
Silver Member
USA
299 Posts |
Posted - 09/23/2008 : 15:14:49
|
Wanna start a flame war ? Go to a DIY HiFi forum and tell them how
much you love JRC4558's in your phase coherent ultra uber super excellent mondo mega hyper pre-amp.
I will admit that 4558's are some noisy critters but they do seem to be well suited to distortion applications.
The stacking trick is a good one. |
Edited by - Racer X on 09/23/2008 15:16:38 |
|
|
|
Dr. Bob
Moderator
Australia
6593 Posts |
Posted - 09/23/2008 : 15:37:54
|
quote:
Originally posted by Racer X
Wanna start a flame war ? Go to a DIY HiFi forum and tell them
how much you love JRC4558's in your phase coherent ultra uber super excellent mondo mega hyper pre-amp.
I will admit that 4558's are some noisy critters but they do seem to be well suited to distortion applications.
The stacking trick is a good one.
Hi Racer X
 You promised not to give away any info, about my next secret project.   |
|
|
|
RonNovy
Copper Member
USA
24 Posts |
Posted - 09/24/2008 : 15:50:12
|
It could be that the phase difference causes certain frequencies to cancel out other frequencies.
Hmm... Someday I'll have to write a program to graph out the effects of this phase shift. |
|
|
|
libra
Copper Member
USA
2 Posts |
Posted - 09/25/2008 : 01:06:25
|
The chips you are referring to are quite old. Chip technology has moved forward quantum leaps since the time when TL072' and 071'ss were the mainstay of many devices. I suggest you do more research and select some modern day chips to at least see if you notice any desirable improvement in what you get out of your device.
Check out some Linear Tech LT1351's. I recently added sockets to an Orban 622B and used the 1351's along with some new caps and the improvement is obvious. |
|
|
|
MoonWatcher
Bronze Member
USA
98 Posts |
Posted - 09/25/2008 : 04:49:59
|
quote:
Originally posted by libra
Chip technology has moved forward quantum leaps since the time when TL072' and 071'ss were the mainstay of many devices.
But those quantum leaps have been targeted at lower noise, a greater gain bandwidth product, lower slew rates, etc. - none of which are engineered with distortion in mind.
If anything, the engineering mindset with current chips is exactly the same as the old stuff - less distortion, less noise, basically less character IMO when it comes to OD and distortion boxes.
What I find interesting is that I can't predict how a chip will work in any given OD circuit simply based on specs. I loaded a 4580 (modern chip) into a DS-1 and BD-2, and it sounded horrible. I put the same thing in an SD-1 this morning, and it sounded fantastic.
In the DS-1 and BD-2, the op amp is driven at subtle levels, compared to the SD-1. So why did the 4580 fail to produce good results with the former two pedals? I can only surmise that it's input impedance is ill-equipped to be hit with a hot signal.
The hi fi crowd cringes when chips like the TL0XX or 4558 are mentioned - the former is a rail banger, and the latter is noisy and diffused sounding (to paraphrase what I read time and time again). But I'll take those two dinosaur chips over any of the Analog Devices or Burr Brown stuff any day.
The OCD uses a TL082 and sounds killer, IMO. |
|
|
|
Laurie
Double Platinum Member
Canada
4854 Posts |
Posted - 09/25/2008 : 06:20:53
|
quote:
Originally posted by MoonWatcher
But those quantum leaps have been targeted at lower noise, a greater gain bandwidth product, lower slew rates, etc. - none of which are engineered with distortion in mind.
If anything, the engineering mindset with current chips is exactly the same as the old stuff - less distortion, less noise, basically less character IMO when it comes to OD and distortion boxes.
What I find interesting is that I can't predict how a chip will work in any given OD circuit simply based on specs. I loaded a 4580 (modern chip) into a DS-1 and BD-2, and it sounded horrible. I put the same thing in an SD-1 this morning, and it sounded fantastic.
In the DS-1 and BD-2, the op amp is driven at subtle levels, compared to the SD-1. So why did the 4580 fail to produce good results with the former two pedals? I can only surmise that it's input impedance is ill-equipped to be hit with a hot signal.
The hi fi crowd cringes when chips like the TL0XX or 4558 are mentioned - the former is a rail banger, and the latter is noisy and diffused sounding (to paraphrase what I read time and time again). But I'll take those two dinosaur chips over any of the Analog Devices or Burr Brown stuff any day.
The OCD uses a TL082 and sounds killer, IMO.
Hear hear!!
Many of us here have experienced the same thing as moonwatcher - put a new hifi opamp into a distortion pedal and it sounds like crap (e.g. I tried an OPA2134PA in my love-pedal modded SD-1 and it was a definite step backwards).
Anyway, this thread is mainly about why the different opamps sound different when on paper, most of the differences in the audio frequency range should be well nigh inaudible. The question in my mind when I started the thread was not why old opamps are lower spec than newer designs. It was why does a 4558 sound "wooly" compared to a TL072, when in the audio frequency range, many of the on-paper specifications are similar - except for the phase response.....
|
Edited by - Laurie on 09/25/2008 06:29:49 |
|
|
| |
Topic |
|
Opamp differences - maybe an "aha" moment
