tubescreamer tragedy

[MmmPedals]

How can the opamps clip before the diodes? Doesn't that mean that the signal levels don't exceed 1 volt peak to peak?

Not sure. Maybe its that they distort more before the diodes.
To clarify I mean to say the following. The higher you turn up the gain the more the diodes clip and the less apparent the "tone" of the op-amp. turn down the gain the less the diodes clip the more apparent the sound of the op-amp. I am not saying this from theory I am saying from experience.
For example I took an mc4558cn chip and had the gain about than 1/2 way up on a "high gain" tubescreamer type circuit. Sounded horrible even to the layman. Splaty, gated and harsh. Nothing subtle. Turned up the gain and although it still sounded bad it began to have a tubescreamer relted distortion. smoother more pleasant. Lifted the dioded (and turned down the volume) and the horrid sound of the op-amp was back in full power. Don't trust me; try it.

[petemoore]

  All kinds of headroom clipping amounts are increased with increased input voltage swings, the clipper doesn't move, so if the swing goes past it clips, the farther it goes the greater the portion of the signal is shunted.
  Rail clipping disrupts operation amplification steadiness to the degree that not only clipping but spiking of the supply enters signal and results in a 'unique' or dangerously horrible sound. Black Cat and OA clipper owners: caveats apply.
  Not that opamps don't distort the waveshape when amplifying within their linear range [when not otherwise disrupted or run outside the clean'-zone].
  Spec sheets and tryouts, many scope traces later [not]...it's just a cool thing until it's better understood as specific causes and effects, all except how the pdconsistant TS manages to still be one of the worlds most fussy opamp tasters. 
  Take away the clipping diodes, get hot enough input to clip the opamp [or even reduce supply voltage...rail out your favorite opamp clipper. This tends tward hash because the circuit-jerk supply voltage...expect at best: hashy harmonics, limited dynamics, rough...
  But a little bit of 'opamp contribution to distortion goes a long way.
  Much of what may make a circuit endearing is the complexity of the content, minding that the more complex the content the easier to get it all muddled into a mushpile. Opamp may just begin acting wierd about when the diodes hard or feedback clip in the signal waveform...the spot that if zeroed in seems to dynamic-focus, playing hard 'twists' the time constants of opamp/diode contributoins, also offering somewhat separate voicing for each example [the opamp is fed a different set of frequencies, the diodes/cap LP filters the diodes].
  Diodes shunt whatever they're fed above their threshold voltage.
  Diodes have no opamp-related worries such as trying to keep inputs steady and feeding output current as needed when their supply runs out, or holding or following 1 set of AC 'instructions' with basically ''2 inputs'' [the second input being related to the ''noticable if not terribly objectionable'' power supply upsets having heavy influence on output wavespikes.
  Like an overflow dam as part of the water tank, diodes dump anything more than X to Gnd. [or feedback in case of feedback clipper. Any voltage above X is simply shunted or fedback through the clipping element diode.   

[JDoyle]

The gain setting has no effect on whether or not the op-amp clips in a Tube Screamer circuit.

None.

The determining factor in the op-amp clipping in a TS is the relative size of the input voltage. If the input signal would have clipped the op-amp output if the op-amp had been set up as a noninverting buffer, it clips. Otherwise, no matter what the gain setting, even if the gain pot is a megaohm or more, it simply will not clip. Because once the diodes conduct, the op-amp effectively IS a non-inverting buffer.

Regards,

Jay Doyle

[petemoore]

Take away the clipping diodes
  Same point exactly J, the OA influence on the circuit Vs. how much the perception the 4558 [enter most any good TS type high performance opamp here] has on the circuit, never fails to amaze.
  Just 1 node move over to the diodes makes...depending on the gain and...a whole lot more difference than a nice Jfet input HPOA.
  Like these:
http://search.yahoo.com/search?p=JRC+4558&fr=chr-offrhap
or these:
  http://search.yahoo.com/search?p=tl072+datasheet&fr=chr-offrhap
  Perhaps some of these:
  http://search.yahoo.com/search;_ylt=A0oG7nA7ojNNDaAABRZXNyoA?ei=UTF-8&fr=chr-offrhap&p=jrc+4558+op+amp&rs=1&fr2=rs-top
 
  Fortunately I also had to swap opamps many times, I remeber actually being wrought out over whether I was making the right decision with this. Best thing to do is use a suitable component, then when the compulsion to swap is overwhelming, expect underwhelming satisfaction, once a good one is in there...well...can't say it doesn't make a difference...as depicted within the third link [open TSOA wormcans at your own risk].
 

[JDoyle]

If you take a look at the internal schematics of the 4558 vs. the TL072 there are but three differences. First, the input transistor type, obviously. Second and third are in the way the current mirror on the differential amplifier is arranged. The 4558 has a cap bypassing the resistor on the diode connected half of the CM, and the TL072 has an extra resistor from the Wilson emitter/CM bases to ground. The resistors on the CM emitters to ground are most likely to swamp/balance out differences in Vbe between the two, resulting in a better balanced differential amplifier and lower second harmonic distortion. What advantage bypassing one with a cap would offer is beyond me - a current mirror loaded differential amplifier's output is a current, not a voltage. The extra resistor in the TL072 serves to stabilize the Wilson buffer which itself serves to improve the mirror function by eliminating the base current error.

How any of this would influence the behavior in a TS circuit, I can't say. But it seems to me that it would be minimal. As R.G. has pointed out before, any differences must be due to the circuit's ability to alternate between high gain (below diode threshold) to low gain (above diode threshold).

Regards,

Jay Doyle

[edvard]

so all the hype about the vintage '80's 4558 japan chips being the only ones for that truly authentic tubescreamer tone is a huge pile of hogwash then?  

Emphatically, yes.  There is NO difference whatsoever, except for maybe a slight variation in the ink on the stamp, from older batches to newer.   People are probably more likely to hear a slight difference between the TI version and the RC/JRC version, then from RC to JRC.  It's silly what people get off on. 

I just recently built a TS for a friend, and I must say I found a VERY noticeable difference between a newer Texas Instruments RC4558P and a Raytheon RC4558NB I pulled a bunch of from an ancient burnt-out mixer board.
In fact, I tried a LM353, a MC1458 and a TL072 and I will admit there wasn't a whole lot of difference between any of those.
But the difference between the TI and the RA chip was definitely noticeable, and in the end I used the Raytheon chip.
I even swapped in a few of the same chips just to be sure I wasn't getting a bad one.
All the TI chips sounded much weaker, the others were a little better but not by a ton, and all the Raytheons (that weren't burnt) sounded much "livelier" (subjective, I know...).
To be fair, I still have not heard how a JRC4558 sounds, so maybe I just got lucky and folks are gonna start seeking out the RA chips now.

I do agree that ~90% of the time, a part is a part as long as the numbers match, with subtle differences here and there.
But sometimes, and ONLY some times, something is clearly different and a simple chip-swap is a small yet worthy thing to do even if all it does is dispel a myth or confirm a hype.

$0.02

[Philippe]

Amazingly & somewhat surprisingly, no one mentioned the Burr-Brown OPA2134...quite possibly the best one of the lot in terms of low noise & fidelity.

A good replacement/alternative to the traditional 4558's used in Tubescreamers & Orange Squeezers. They're a tad more expensive than 4558's & TL072's but what the heck.

[DougH]

I've done op amp shootouts with my Pentode Driver. TI4558 vs. JRC4558 sounds different, which sounds different than TL072 and LF412. How different can be subtle or not, but it's all noticeable. Don't know how this applies to a Tube Screamer circuit as I've never tried it. My stock TS9 sounds just fine with the non-vintage, non-mojo op amp it has in it.

[askwho69]

+1to opa2134! most wonderful opamp for dirt pedals

[Al Heeley]

from what i read, the anal purists mock the ts9 as being inferior! Takes better ears than i have remaining to detect differences in real life during a pub gig playing green day covers

[jasperoosthoek]

I think the biggest differences can be heard either playing alone or on a high quality studio recording. I have trouble hearing differences between the stock TS9 and modded versions on youtube videos.

But you do get used to a cirtain sound. Something only sounds different when you compare between two not on its own

[Philippe]

I think the biggest differences can be heard either playing alone or on a high quality studio recording. I have trouble hearing differences between the stock TS9 and modded versions on youtube videos.

Or at a large venue where the guitar signal is being miked, run through a soundboard & fed into a massive PA system.

[petemoore]

  Another mysterious case of A5 4 5 resulting in 1 aught and three of the other.
  Though it really is not necessary and recommended only after other recommendations are considered, you will have to swap opamps back and forth to determine in greater accuracy and detail, what some of the myriad varieties of the infinite possibilities of answers to:
  What differences could I expect ?  None until it is plugged in/plugged out from and to some things. Changing any of the things [amp/guitar/settings/etc] like speaker will completely dwarf good OA Vs. good OA challenges.
  If all the things other than the oa have been fine tooth combed, the differences in oa's may be more discernable making it easier to make a personal preference choice.
  Soon as it is plugged in, any less than clairvoiant expectations are answered in the kind of detail only the perception of the pressure changes in our 'unique to ourselves' ears feed info and conjure up as images in our minds as the search for words is repeatedly precluded by searching for more highly refined words to allude to the missing content, as the familiarity increases and the perception begins to change. 
  Plug in MY amp with vintage and modern supergoodies through supersweet blue speaker [or any other amp/PA/other that isn't like yours], and the perceptions visaged during the previous affirmation and/or debunking of expectations or preconcieved notions. This is then presented with a whole new set of circumstances again, altering preception and perhaps providing a totally mind blowing experience [if there's a huge difference].
  Any of the other things mentioned make better mileage [as far as 'difference] for time as well as conversional conversation.
Get the HQ OA that you have/we recommended [TLO72, or any other dual] or an unavailable one in there and mess with the diodes, they're like a 'wide parameter/coarse' adjustment, the OA swaps would be like very-fine tuning, but as long as they bias, no further tuning is needed.
  Although it's no fun desoldering, I have found some super vintage 'extra sweat' duals with 4558 stamped on 'em in a few salvaged items. Good new 072 anyday of the week.
  Pretty much need two identical as can be platform racers side by side at the starting line [balanced out settings to accomodate for R/C tolerances and testing anomolies], then the opamp engines go in A and B, B and A respectively...to accomodate for circuit differences and to determine the absolute difference makes no difference to wording here best possible Opamp type/individual/candidate match to your choice of nearly identical TS circuits.
  ...then we''ll plug it in...uh....heeere, she doesn't mind, he never liked TS's but this one is different.

[Al Heeley]

A post bordering on almost poetic enlightenment in parts from the revered Mr Moore. Back down to earth, I suspect my question has been answered many times over now, for which I am grateful; the reality of the build centering on getting a unit working with bits already purchased and in my possession rather than having to put in another order to replicate components I do not actually need, to give a working unit to my guitarist in our rock covers band for a gig in a local pub this saturday. The nuances of tone in a blind hearing will be lost on him with his pointy dean guitar and EMG pickups. For electronics hobbyists the prosaic postings about minute tonal nuances is no more than indulgant  technical masturbation when you get back to the reality of putting together a home made overdrive stompbox.