Chip experiments

[Mark Hammer]

Some 28 years ago, when I was living in Hamilton Ontario, a potato chip company test marketed fruit-flavoured chips in the region (grape, strawberry, cherry).  They were simply awful and died a quick and deserved death.  This experiment in chips is a little more useful and informative.

I recorded an MP3 (128kbd) in which I subbed 6 different op-amp choices in my MXR Dist+ clone, and played the same riff with the exact same settings.  I can't absolutely account for the consistency of my picking but it was pretty damn close each time.  The riff used is long enough to have a sense of how each chip sounds, and brief enough to provide easy comparison of adjacent sounds.

Audible differences ARE heard, some of which I like more than others  The 741 is included, as are a number of Bifets, a single 5534 and a piggyback paralleled pair of 5534s.  The zipfile contains the MP3 file and the descriptive textfile.

Though we've all talked about the effects of chips, I do believe this is one of the first times that we've been able to hear a side-by-side comparison of op-amps.  Personally, I found it interesting, and encourage others to post similar experiments.

You can find it at the top of my site:  http://hammer.ampage.org  The file is verified to work.  Right click, and save.

[Ry]

Hey, that's neat!  Of the differences that I can hear on my laptop speakers, I definately like the TL071 and TL061 the best.  They both sound the same to me right now, but I will have to try with headphones.

Thanks!

Ry

[Mark Hammer]

Good.  I was kind of worried that converting from 44.1khz sample would lose something.  If I look closely at the waveform on the original wav-file, the dual-5534 has very rounded edges, while the 741 is more like a descending ramp for the same part of the signal.

Note that the Dist+ circuit, which assumes a 741 is used, provides a very low current bias.  Where it is common to use a pair of 10k to 47k resistors to divide down the V+ to get a Vref for the biasl, the Dist+ uses a pair of 1 Meg resistors, which undoubtedly changes the amount of bias current available.  Now I'm curious about the extent to which the dual-5534 arrangement is really a byproduct of the biasing, rather than the chip itself.

At the very least, the experiment shows that subjective impressions of circuit improvements with chip changes are well-founded.  Alternatively, the individual tones of the chips may well be an interaction between biasing arrangement and chip-choice rather than chip-choice alone.

[Fret Wire]

It would be interesting to hear the same clips again with a 100k/100k/10uf biasing, 22k/22k/10uf (DOD 250), or similar.

[bioroids]

Very nice indeed!

I'm listening thru my headphones here (Sony MDR 7506), and I think both TL0xx sound the same.

I like the TL's family the most, though probably on a band context a little more bite would be needed. For that I'd choose the NE5534 alone.

Regards

Miguel

[Samuel]

i blind taste-tasted and I was definitely drawn to the first and last run throughs based on on lack of harshness. NE5534 all the way. Now why do I have a bin *full* of TLxxxs, and have never even tried any NE chips?

[Ethan]

Very interesting and very subtle differences but I can definetly hear it.  I think I like the 741 the best but didn't sound much better than the low power tl061.  I wish you hadn't edited out the noise as that would've have been very informative too. Anyway, thanks for recording those samples.
-ethan

[MartyMart]

Mark, that was an interesting little test !

My vote goes for either the 741 or the CA3140 ( never seen one ! )
I do like 741's in general !
TL071/TL061 - almost identicle and a bit "darker"
Piggy back 5534 - a bit "dull" sounding
Single 5534 - OK but not much personality, if you get my meaning.

We need a TS-9 test along the same lines ........

Marty.[/list]

[Fret Wire]

We need a TS-9 test along the same lines ........

Been there, done that...to death! Six tubescreamers at once! I didn't do any sound clips though.
JRC4558
RC4558
RC4559
TL072
OPA2134
OPA2064
NE5532

I liked the NE5532 the best, followed by the RC4558. It's like regular sound clips, you get the basic idea of the character, but it doesn't mean your's will sound the same. Also, some chips will sound good on little practice type amps, and sound not good at real volume. Low/high sound pressure (volume), and amp break up/clean headroom affect the outcome too. Some chips sound good no matter what, while others shine or choke at different sound/breakup levels. Of course, the same applys to the complete pedal too.

The Dist.+ is often maligned, but when used at good volume with an amp already breaking up, it's got a nice sound. When it's played in little practice amps that are not breaking up, you are relying on mostly the pedals distortion. People always like to say Randy Rhodes used it for a dirty boost. That's so much rubbish. He basically just used it the proper way. You take a 50 or 100 watt Marshall that's dimed, and you don't need the gain pot turned up 3/4, like you would with a clean little amp. You up the gain from zero until you've reached the extra dirt/distortion needed, then adjust the volume pot for unity or higher. That way your sound is mostly the amp, with the eq voicing of the pedal, and the slightly higher drive of the pedal. The guitar's character also is retained that way. That's the proper way to use any od/dist. pedal. At least with non-master volume amps. There's a huge difference in sound between an amp on 4 and dist. gain at 4:00, and an amp on ten, and the dist. gain at 11:00. Enough of that rant.

Btw Mark, did you run a compensation cap (22-30pf) across pin 5 & 8 on the NE5534? I like the 741 with the Dist.+. If I want Hi-Fi, I go with the NE5534. I don't bother with the OPA134a, it's more price than performance. The OPA177 is a true high performance single OA, however. But I think the NE5534 would satisfy most hi-fi type needs.

[puretube]

here we go again...  :roll:

[Fret Wire]

here we go again...  :roll:

How's that? I always wonder why people read a thread, don't like it, but butt in anyways with comments that don't serve the thread one bit. :roll:  :roll:

[cjlectronics]

The TLOxx series chips are like Fritos Scoops, they have always been my favorite.  Add a hash mark to the TLO71/61 column for me.  

Is there an op amp that we could catageorize as one to stay away from, kinda like the Olestra potato chips that caused anal leakage? :lol:

[Mark Hammer]

here we go again...  :roll:

How's that? I always wonder why people read a thread, don't like it, but butt in anyways with comments that don't serve the thread one bit. :roll:  :roll:

Nah.  Ton would never be that sarcastic.  I think what he means is a certain fatigue with the op-amp mythology surrounding the TS-9/808.  It was not my intention to start any parallel trend with regard to the Dist+.  Rather, it's a simple circuit, I had the thing built and socket-ready, and was happy to find a reason to use my recently acquired Roland USB device.

As for noise, Ethan, you need to understand that I'm sitting in a room with flourescent lighting, playing a non-humbucker guitar, right in front of the monitor, with an exposed circuit board on the counter in front of the monitor.  Trust me, the little bits between the strummed portion would not have added to *anyone's* impressions of the chips in question.

Fretwire,
No compensation used at all.  The 741 *does* have a little more sizzle.  It is entirely possible I'd love it through a 12" speaker, but through an extended range 8" speaker I find the mushiness of the dual 5534 more to my tastes.  The TL0xx chips struck me as having a certain "woody" quality that was hard to put my finger on.  And yes, I'm also curious about the bias x chip interaction.

Martymart,
That CA3140 was purchased in 1978 or so.

Personally, I have found it very hard to verbally describe the sonic qualities ofthese various chips in this application.  What I do know is that different devices WILL have a different sound in the same circuit.  I encourage others who might socket key semiconductors in their projects to record similar kinds of soundfiles to illustrate the impact of component changes.  Rather than boosting the mythology surrounding certain components, I think such soundfiles would do much to deflate the mythology by demonstrating that other things besides the "magical" semiconductors rumoured to be all-powerful are able to produce pleasing sounds that can suit the user's tastes.

[Fret Wire]

Nah. Ton would never be that sarcastic. I think what he means is a certain fatigue with the op-amp mythology surrounding the TS-9/808. It was not my intention to start any parallel trend with regard to the Dist+. Rather, it's a simple circuit, I had the thing built and socket-ready, and was happy to find a reason to use my recently acquired Roland USB device.

You're probably right. I just say what I think so no one has to wonder.  I don't know about "fatigue", I hope no one here is mentally fatigued over words in a thread they can easily pass over. A TS is just a mid-heavy OD pedal, nothing more. It's not even my bag. But, at the time, to see for myself, and learn in the process, I went out and got 6 like new TS, socketed them and A/B'd the hell out of many IC combinations. Only after someone has spent as much time, effort, and money as I did, do they have something to say about it to me.

If I had posted sound clips of my 808 chip experiments, you would hear a difference as we did with yours, but the opinions would not necessarily be the same as mine.

The most important thing your clips demonstrate is that there is an audible difference in chips, no denying that. It's a great inspiration to experiment, when you can hear the results, not just read about them. I was adding the point that different amp volumes/breakup also are part of the equation.

The biasing arrangements are interesting. I built My original Tube Reamer perf (long since stripped) with sockets for different biasing values, but I never messed with it. The next D+ I whip up, I'll have to try it. I'm curious to see how the different biasing works with the 741, as well as other chips.

[Mark Hammer]

Before I start some unjustified rumours, let me emphasize that I do not KNOW if the audible differences are due to biasing.  Rather, the biasing arrangement on the Dist+ is so undeniably different than what one finds elsewhere, I can't help but wonder if what we hear MIGHT be a product of the bias x chip interaction.  If it were the difference between a pair of 22k vs a pair of 47k resistors, I wouldn't make anything of it.  But a pair of 1meg resistors (and remember that there is also the resistor from the noninverting pin to Vref to cntend with) is so "out there" in the world of op-amp biasing as to be very conspicuous and worth investigating.

If any of the true EE's out there could explain the role of Vref bias current in op-amps to us, I'd be much obliged.

[RLBJR65]

Great post Mark! I think these types of posts are very informative.

I'll 2nd Marty's descriptions, with exception to the piggy backed 5534; I'm with Mark there I kind of like that mushy sound too.

I tried this same thing with an ROG tube reamer. I really like my TR but with the 741 it was really noisy, hissed etc. TL0xx were dark and kind of blah. I was very pleased however with the CA3140(E), no noise and retained more of the high end. I got the 3140's by mistake at a surplus store, they were mixed in with some TL071's, Mouser has the CA3140E's in stock though.

Here are some Intersil distributors that have them as well.
http://www.intersil.com/support/distilist_stkchk.asp?pn=CA3140E

I would also like to here opinions about what the tonal effect biasing may or may not have, that could be an excuse to add more knobs!!

[puretube]

hey Friends:
no sarcasm at all - I`m just lacking time to elaborate currently,
and am a little afraid of a new Mojo-ical excursion by 3rd/4th/5th parties...
:wink:

[aron]

One of the problem is that you cannot broadly say xx chip always sounds better than yy chip; it depends on the circuit.

However, I have liked TL07x chips and 4558 chips a lot.

It's not only the tone though... More often than not it's the feel.

One more thing, I did play around with biasing and I didn't notice much difference at all to be worth it.

Now I didn't play around with the voltage divider resistor values (as in 100K vs. 10K). This might be a lot different.

[Fret Wire]

It would have been great if someone asked Keith Barr (MXR) about it during his two recent interviews (fx books). You'd think he had some reason to choose those divider resistor values. Something in the 741's data sheet must have caught his attention.

Maybe Tom Hughes could ask him, as sort of a informal follow-up. :wink:

[Tobias Karlsson]

Great stuff Mark!

Here's another by me, side by side op test in a socketed TS style circuit.

http://hem.fyristorg.com/tobiaskarlsson/opshootout.zip

6.7meg

Quoted from the zip file;

I made an ”OP Shoot Out” using a TS style of circuit.


You could easily convert them to wave, import them to Cubase, make a loop at a specific place and enable the “solo” button in Cubase (or whatever) and switch between the different op amps in the fly.

I’ve tried to play the same way in every recording and managed to get pretty close, but of course it’s not EXACTLY the same.
I made some of the most important sounds (at least for me) in the short sound samples, a power chord, double notes (a second, e and d, a third minor), single notes, some muting on the “A” bas string and some places were you can here the note fade.
The recordings are made one after the other, same setup, nothing is done in the computer EXCEPT for “normalizing”, so this test does not show if one is louder than an another.


In my ears there’s almost or NO difference between the Op-amps in sound other than that the 1458 have a little bit less treble and the LM358N sounds like it’s maybe a bit miss biased?
In some places there is a VERY small difference but in the next “lick” in the next part they sound the same again. This could be because I simply can’t play exactly the same picking attack, where I hit the string etc, but it’s pretty much the same. If there were the same difference all down the line I would think it is in the op-amps.

Now, IF it actually is a different in sound, it’s so SMALL that personally I won’t sweat over which of these OP’s to use anymore (except for the LM358N) and 1458)!

Another cool part is, if you loop the very first bit of the tracks you could compare the noise levels. I save the beginning for that purpose.

The 1458 got the highest noise level, after that it’s the TL068CP, after that the LM358N.

The TL072DP, TL082CP and all the 4558’s have almost the same noise level.

The OP200 is slightly better.

But the NE5532P is a very quite OP in dist circuits!
I’m impressed by that one so for me I think I’m going to use it when I care about the noise levels (it seems to sound the same way as 4558, TLxx anyway!)

This shoot out was REALLY interesting for me, I always wondered about the differences. It strikes me that there could be differences when using them in another type of circuit but I don’t know, I think that most of the time there wouldn’t.

I picked the TS because all the discussions about “upgrading” OP’s and so on. (The JRC4558D I used should be the real one)

PLEASE don’t feel offended those of you who have a different opinion as I know there been some discussions preferring 4558 before TL etc.

FUN!!

We could easily make a test, listen to the samples and then I could rename them and we could see if it could be recognized!

Another thing is what Aron say, the “feel”, personally I haven’t though so much about that so far, but he that could be something that I should think about!

Sidenote, bikers in Sweden like to think of them self as knowing what kind of beer they like, colours of the bottle is important and so on. We were at a biker club and had to do a blind test at a party, my wife was the only one that could pick out and name all the different beers that we tested, the rest of us were “way out” :-)

Best regards

Tobias Karlsson in Sweden




Best regards

Tobias Karlsson in Sweden