Sound differences from different dual op amps in SS amps

[Pedro Vecino]

Op amps TLO72, 1458, 4558, 5532, LF353, LM833 ... do not sound identical on the same circuit. The differences are subtle and also depending on the particular design may be irrelevant.
A Tube Screamer do not sound identical with TLO72 or LM833, for example. They are little nuances related to the harmonic content that can be seen playing with him.
But fifteen or twenty years ago I tried LF353īs (one of the most distinctive in my opinion) in a Peavey Renown 400 as an experiment and the result was irrelevant. Today my old Peavey maintains the original op amps.

[pdf64]

RG has some thoughts on why different types of op-amp may sound different in a TS but it might equally well apply to a guitar pre-amp, about 3/4 down the page Tube Screamer Frame Definition
To summarise/paraphrase, it may be primarily due to how they react when overdriven, and their recovery from overdrive.

[dai h.]

this could be something akin to someone asking (for a tube guitar amp) what brand of caps to change to to achieve an "upgrade" when what is really required is a change in cap value (though one might hear slight differences for better or worse swapping different brands due to tolerance differences, dielectric differences, and possibly subtle coupling). Maybe if a schematic of the Peavey in question could be supplied there would be something more definite (quantifiable elements to looks at) to go on and look at (and advise on).

another thought with op amps is that what it seems is that in the attempt to maximize linearity, the type of amp they are (topology elements like the differential input, lots of feedback and other elements I don't quite yet understand) helps get more and more linearity but make for nasty sounding clipping if overdriven (so it isn't that they are "bad" but more that they shouldn't be used in such fashion).

That being said, the scope of "art" can be very very wide and from hearing things like (Beatles) "Revolution" and some New Wave era guitar sounds (which sound like nasty clipping) sometimes "bad"(say hard clipping producing harsh upper harmonics?) can be "good" or situationally so (or was there an element of wanting to be different from the Marshall toting longhairs?). (Note that musicians do seem to notice overdriving "designed for clean" type amps can produce really nasty harshness and apparently empirically solved or ameliorated that by LPF(hi-cut)ing the signal (knowingly and perhaps unknowingly sometimes using high end limited cassette multi-track or other gear with limited highs) so it's not quite so nasty but still different and interesting (compared to "nicer" tube amps and overdrives, etc.?).)

[Leo_Gnardo]

A snip from the info page from the link quoted above

HA! Salesman's promises of audio nirvana. We all listen to recordings made on consoles chock full of ordinary everyday op amps and we love 'em. Signals have passed thru hundreds of op amps from the mic inputs thru all the rest of the process. Results: some hit records (LP, CD, download or whatever) and some not, but they get bought & listened to, and you don't hear people complaining "sounds like a cheap fuzz box."

There is a "sound-good" op amp I've run across and used, Jensen JE-990. Also made from individual componenents in a one-inch cube. Does not fit a DIP socket. Expensive? Last time I ordered, about $45 each. Probably a lot more now. But I'd never suggest sticking 'em in a guitar amp. Makes an astoundingly good mic preamp, but so do tubes.

[tedmich]

this could be something akin to someone asking (for a tube guitar amp) what brand of caps to change to to achieve an "upgrade" when what is really required is a change in cap value

EXACTLY! Its like at work when the boss comes in with some newly acquired tech and says "don't change anything, but make it BETTER" I suspect the real perceived difference when people install "hand rolled hemp oil and gold foil caps" is because the actual values are different, kind of a twofer for bad production QC...And none of these audio-fools are actually going to test out the component before installing because "numbers can't capture the nuances of sound" etc, etc. Maybe we should sell some hand rolled esoteric material caps in dark (+10%) and light (-10%) flavors? It makes it so much easier for people to say they hear a difference when there actually is one...though obviously its not required.

[tedmich]

There is a "sound-good" op amp I've run across and used, Jensen JE-990. Also made from individual componenents in a one-inch cube. Does not fit a DIP socket. Expensive? Last time I ordered, about $45 each. Probably a lot more now. But I'd never suggest sticking 'em in a guitar amp. Makes an astoundingly good mic preamp, but so do tubes.

Steve Hogan and John Hardy still sell versions of this for about $49 (JH) Their newer versions use SMD components and some believe the sound is less "penetrating" (kidding)
see:
http://www.johnhardyco.com/pdf/990.pdf
The Sound Steward

[J M Fahey]

I stand by, strongly, to my "all Op Amps work/sound the same when used within their limitations"
You can call that design parameters, datasheet specs, "as God/Designer intended", "properly", it's the same.

I have been using Op Amps exclusively in my amps, self designs of course, far longer than Peavey, Fender or Marshall , who still used discrete transistor preamps far into the 70's , very easy to check by looking at old schematics.
Since February '69 , to be more precise.

I guess by then only Kustom used Op amps, but am not so sure, should check some schematics.

I mean popular amp makers, maybe (probably) there were other obscure makers around (such as myself )

For so long, that when I went to the only advanced SS parts supplier in Buenos Aires to buy LM709 , what I had been using so far, the store owner came personally to the counter when he recognized the nerdy kid who wanted the latest novelties and offered: "here, test this, it's the new one, same as 709 but needs no compensation"

And why was I instantly hooked?

because of:

Characteristics of a circuit using an op-amp are set by external components with little dependence on temperature changes or manufacturing variations in the op-amp itself, which makes op-amps popular building blocks for circuit design.

I suggest reading: Operational amplifier - Wikipedia, the free encyclopedia

As I always repeat, the Op Amp is meant to be used as it's meant to be used ... if beyond its open loop gain (which is huge, much higher than any standard tube anyway) it's not an Op Amp anymore but a bunch of transistors.

In an old style distortion pedal, which as I said before are used at or beyond the limit of old Op Amps :
* Tube Screamer gain 500k/4k7=100X
* MXR Distortion + 1M/4k7=200X
* RAT 100k/47=2000X <-- this impossible gain has an explanation
yes, there will be a difference, a bandwidth difference within the audible area.

But do not generalize from 3 "wrong designed" examples to millions of properly designed ones ... even in Guitar amps of course.

In properly designed stages, such as in countless Hi Fi or Pro Audio (including Studer and similar quality mixers, HH amps used by BBC as monitoring amps, etc.) even old Op Amps met product (very high) specs; substitution by modern ones' main advantage is lower noise, which is important, of course, and better rejection in balanced circuits, which is important specially in transformerless input mic stages.
Also NE553x Op Amps have higher current driving capability, which is useful when needed, at the price of doubled or tripled idle current consumption.

As I had said earlier, an MXR Dist+ with a 741 @ 200X gain runs out of gain at 5kHz (so it's still somewhat working "properly" at about half that, some 2500Hz , very audible) so using a TL071 there will have 3X that bandwidth , triple those values to 15000 and 7500Hz respectively.
That I agree can be heard, of course .... unfortunately in that particular use it's not exactly considered an improvement, because sound can be called audibly harsher/buzzier ... not a mystery if we are talking about distortion.

I don't much like it when anecdotic or not-the-normal-case is repeated over and over around the Internet, mostly by those who have not made the experiment themselves and simply repeat what they read somewhere.

In a way, it's a bonus , competitors around me are dismayed when they can't get specific , near unobtanium parts, and either have to stop or pay a fortune.

One had to stop 6 months because he couldn't get one specific Japanese FET on which he bases his preamps and Customs were closed for that length of time (don't ask) , I suggested he uses another and , worst case, hand picks and/or rebiases ... I was stared at in horror.

While I can buy Op Amps, any op amp, by the bucketload, for peanuts and do anything.

EDIT: I forgot the bit about the RAT.

As you can see, external elements order gain=2000X which is impossible well within the audio band.

This is the various compensations graph for LM308 , the Op Amp which has to be used there (yes, the RAT is designed, or to be more precise, it was a happy accident, around it).
In fact, it comes from a "RAT analysis page" so it shows exactly what I need:



As you see, at 2000X (66dB) and with that compensation, LM308 runs out of steam at amazingly low 500Hz ... and is probably already suffering at half that .

The designer itself tells it in interviews, something like (quoting from memory) "I was protoboarding it, and to get just a tad more gain, reached for a 470 ohms resistor ... by mistake took a 47 ohms one.
Didn't notice the mistake at first but the pedal started distorting in a weird, wonderful way, different from plain diode clipping which I had been getting so far ... after noticing the resistor error started wonderin why and then noticed the Op Amp was slew rate limiting and at the same time being driven at those frequencies, the result being a different kind of distortion

FWIW I have tested that (of course ) and I get, instead of the classic flat topped square wave, a triangle wave, which afterwards gets re-clipped by the diodes.

Also replaced LM308 with LF351 or TL072 and it became both buzzier and reedier ... although of course that can not be generalized to 99.9999% of other Op Amps.

The other circuit which uses such "clipping without diodes" is the Apex Exciter .

Lots of people look at the original schematic and wonder how harmonics are created (to fake lost highs) without visible clippers ... the answer is hidden in plain sight: slew rate limiting/distortion.

[Jazz P Bass]

Peavey Special 112 Schematic

[QUOTE=dai h.;377458] Maybe if a schematic of the Peavey in question could be supplied there would be something more definite (quantifiable elements to looks at) to go on and look at (and advise on).[QUOTE]

Here is the schematic.

[tedmich]

4558s??? Throw it out!
JRC4558, worst op amp EVER.... - diyAudio

[J M Fahey]

Just guessing of course, but probably JRC4558 is NOT an RC4558 photocopy but "something" JRC self developed, without paying Raytheon any royalties of course, whch could be used in audio applications and had same pinout.

Being much cheaper, massively produced and domestic Japan produced, in an era where Japan consumer electronics ruled the waves, (think Kenwood/Yamaha/Sansui/Pioneer/Panasonic/etc.) it obviously got massively used.

I very much doubt any of those Factories would have searched and used the American original, harder to source in quantity and at 3X or 4X the price.

That it was not an exact clone probably didn't matter much.

FWIW many Japanese ICs of that era, think tape head NAB preamps, magnetic pickup RIAA front ends and such, were incredibly un-sophisticated , just 2 or 3 transistors (no kidding) encapsulated in a DIP or SIP package for convenience, needing tons of external components to work at all.

I did repair Hi Fi stuff way back then, and sometimes even pulled dead ICs and put TO92 transistors in the proper holes, go figure.

So JRC4558 being just a close "functional equivalent" of an RC4558 would not surprise me in the least.

In that case, not being able to meet even minimum "official" requirements means it would definitely "sound different".

[dai h.]

IMO, while I could agree with the goofiness of woshipping some specific op amp, the idea that JRC4558 (or other 4558s) aren't capable of decent (or good, even) audio is bogus. He calls them "scratchy", so there must be something very wrong with the gear (or the implemenation). My impression is that is a case of ignorant bluster and grandstanding (and "piling on").

(From what I understand) there were pro multitrack tape recorders and outboard gear that used them (4558s--not sure if the same manf.).) I also had a Nakamichi cassette deck which had (one? NJM/JRC4558 my memory is a bit hazy), and that tape deck seemed to noticably sound better than other cassette decks (better general sound quality even with "normal" (as opposed to "chrome", "metal") cassettes and capable of recording a hotter signal. This was a 3 head deck where switching the monitoring from source to tape was possible, and (at least to my tin ears) I found it hard to distinguish between the original and tape esp. with higher quality metal cassettes).

Also (IIRC) there were pieces of pro gear that used (what might be considered the even lowlier) 741, so I think again more about how well understood and used than automatically being discarded for being presumably inferior.

[Enzo]

Pedro, I think you made the point some of us have been making. You said the different op amps sound different in a Tube Screamer, fine, but we are not talking about a Tube Screamer, we are talking about a Peavey combo amp. You also report you tried different op amps in your Peavey amp and they didn't make a difference. Exactly.

People should realize that even if different op amps make a difference SOMEWHERE, that doesn't mean they make a difference everywhere.

[Slobrain]

Peavey Special amp 1981 schematic.pdfPeavey Special board layout.pdf

Here is the schematic and board layout from the old Special amp

[J M Fahey]

[ATTACH]33044[/ATTACH][ATTACH]33045[/ATTACH]

Here is the schematic and board layout from the old Special amp

Excellent and thanks for posting.

Let's analyze a commercial amplifier, a classic and sold by the truckload.

By the way, it's amazingly close to a Bandit (a good feature in my book) but higher power, 4 output transistors instead of 2 and the unusual bonus of having an autoformer on the output which lets it drive high power (100/120W RMS) into 8 ohms, very unusual in an amp of that era.

Ok, let's see all Op Amps one by one, and how well (or wrong) did Peavy Engineers use the datasheets.

1) U1a , Normal channel: 33k/330r = 100X at max gain, active gain control.
So on 10 stage will have 100 gain, will be crunching with any reasonable guitar, will run out of gain at 30kHz and be quite flat up to 15kHz ... nothing (audible) to gain here.

2) U1b , saturation/distortion channel.
Here we surprisingly start with less gain than the clean channel (33k/680r=50X) , but Saturation is a dual pot: one section gradually introduces clipping diodes (interesting, instead of plain On/Off) and the other adds 10X extra gain which goes beyond what a standard Op Amp can give ...... if volume is already on 10 ..... but no problem if volume is, say, on 5.
Personally (may be wrong) , I think that this was added to get more distortion at lower volume, a problem which plagues Guitar players even today, 33 years later

3) U2a , channel Normal or saturation selector, gain (33/10)+1=4.3X ... piece of cake.
Bandwidth around 200kHz so I won't even mention it.

In general 100X is a practical barrier, 50X easy, less than that incredibly easy.

Of course, unless an Op Amp is used very close to the limit, a newer one will improve noise but not much else audible.

4) U3a , tone stack recovery and active presence.
Standard gain, as before (33/10)+1=4.3X .
With presence on 10, high frequency gain around 12X .

5) U3b , mixes dry signal and reverb recovery, gain 2*(68/22)+1=7X .

So we see that except in the input stages, which are expected to distort, everywhere else Peavey Designers were very conservative in their use of Op Amps (in this case 4558) , and they behave practically like textbook examples of "Ideal Op Amps" .

Nowhere they have to drive difficult loads, except in the reverb driver ... which is not in the direct signal path and which by the way will pass through a horrible mechanical spring delay so its sins become very minor by comparison.

As a side note, it's incredible how GOOD this Peavey design is, 33 years later .

If it weren't turned down by snobbery and low price, it would still be a very good competitor today.

[Slobrain]

Thanks for the input J.A.,

This particular amp sounds pretty good but I have a Peavey Renown that was built in 1984 that sounds much better and uses the auto transformer also. This amp is somewhat lacking and not as loud as I would have thought as the Bandit 65 seems to sound somewhat better and gets almost as loud even though the Bandit is 65 watts and the Special is 120 watts. I'm wondering if I might have a faulty opamp in it somewhere as it doesn't have the oomph the Bandit 65 has. Over all the renown when plugged into a Marshall 4x12 can get close to my Marshall non master 50 watt head in sound surprisingly. Not as articulate as the Marshall but still really good.

If you see anywhere to improve this design please post. I would like to make it better if possible. Peavey engineers were good at what they have done. I started to get my electronics engineering degree in 1989 but wrecked my car in the first semester and had to quit to go back to work... otherwise it would have been nice to look at such an old design and improve it to sound better. Yes, these old Peaveys just can't be beat for the build quality at the time. I just don't see that build quality in newer amps and its a shame. Peavey was really ahead of his time in business approach for the working class musicians. No one else was really doing that back in the day... Hartley Peavey was looking out for the poor musicians when he was building and starting his company. I think without him music might have not taken its course the way it did in the late 60s thru the 90s... but that's my personnel opinion.

BTW, the picture of my amps with the peavey add was something I did one day, just had too much time that day...lol...

Cheers