OPA4134 drop in for TL074?

I just swapped one of the TL074's in my Line 6 Toneport KB37 to a Burr Brown OPA 4134 and now the mic pre is super fuzzy sounding like a transistor out of bias.

So... either the two aren't a direct swap or my soldering didn't quite go as planned. I looked at it carefully and all looked good. I didn't use a magnifying glass so I should also check that out.

I thought a little more about it. The reason I did this was because I wasn't super thrilled with the mic pres in the Toneport (wanted to upgrade them a little and see if they'd be more useful). I have built a Neve EZ1290, so I have a better pre on hand but still, this is handy sometimes (the line 6 thing). Come to think about it, maybe the way this mod turned out was the best way it could have gone (basically made the mic pres disabled and unavailable to me).  I still like the guitar sound for quick tracking of an idea.

I am still curious on the OPA4134 and it's compatibility as a replacement for the TL074 chip if anyone knows.

When second guessing circuit design it helps to have some idea of what you are doing.

The 4134 looks like a slightly better part than a TL074 (30+ years old) but the long in the tooth TL074 is more than adequate when used correctly and neither is capable of being a complete mic preamp all by themselves.

Who knows what you actually changed, I won't speculate.

JR

PS I am not a big fan of blindly substituting newer parts to hopefully improve a design. It's more complicated than that.

I've seen it discussed before (DiyAudio, GS, etc) that the BB opamps sound better than the less expensive TL's. I've swapped them in other things (Digi001 for example) without issues; also the pinouts are the same. So I've done some homework and am not blindly swapping parts. I do have some idea of what I'm doing.

But, that's all I know. I asked here because I figured maybe someone here has a better feel for (i.e. more experience with) that particular opamp. I've been wrong before so I could be now, but I don't think it was a dumb question or an improper one to ask on this forum. Where else would I ask? 

And no it's not a mic preamp in a single chip. There's another opamp before this. I'm guessing it's a line driver portion of the mic pre in this instance (but just a guess without a schematic and since it's 2 sided SMD it's not exactly easy to trace without seeing the bottom side of the board - which still wouldn't be a simple task).

I do know that the OPA4134 was an improvement (to my ears) in the Digi001. That's why I wanted to try it in the Line6 piece.

There are so many other things to consider, such as, but not limited to:

Input/Output impedances being presented to the amp, the supply rails and their quality, the type of output being driven - capacitative, resistive?

Anyone suggesting blind swapping of parts without understanding the surrounding circuit and the likely outcomes of doing such a thing should be ignored wherever possible.

If someone can explain WHY a particular substitution should work and why that should sound better, then that would make some sense.

JohnRoberts knows what he's talking about, so listen to him. Blind substitution is pointless. Especially with opamps. Imagine blindly replacing the engine in your car. Not only might it not fit, but even if it does, your breaks, the wheels, the drive chain, etc. etc. etc. ain't designed to work with it, so it won't be optimal.

A schematic, if available, would clear up any questions we might have and allow us to understand why this substitution is doing what it's doing.

People who post on DiyAudio, GS, etc. IMO are not to be trusted. It's one of the main reasons why I only post here, and not there.

should work and will work are different. They are pin compatible so all things being equal they should work. but that doesn't mean they will work and even if they do does not mean it will be problem free.

Kingston had an excellent thread on what is REALLY important for noise & THD; opamps and local decoupling of rails, some questions

http://www.groupdiy.com/index.php?topic=37307.80

Many true gurus chime in.  It proves how OPA rolling takes a VERY poor second place to correct earthing, layout & decoupling.  It’s a long thread but read the whole thing from #41 to find pearls of wisdom.
_______________

Your OPA swap has almost certainly made the circuit unstable and what you are hearing is oscillation.  Check if it goes away if you swap the TL074s back.

777funk said:

I just swapped one of the TL074's in my Line 6 Toneport KB37 to a Burr Brown OPA 4134 and now the mic pre is super fuzzy sounding like a transistor out of bias.

So... either the two aren't a direct swap or my soldering didn't quite go as planned. I looked at it carefully and all looked good. I didn't use a magnifying glass so I should also check that out.

Click to expand...

Well, for starters, RFTDSs.

OPA4134 has higher bandwidth; is the output oscillating?

-a

Thanks for the replies. And 100% respect for John Roberts by the way. I've read his posts here and other places; I have no doubts about his credibility/opinion/statements that's for sure. He's a pro and has quite a background from what I've read. I was just pointing out what I have and haven't done (read parts of the data sheets, read some opinions, tried the chip in the past, etc). I definitely don't have the knowledge and especially experience of J.R. and many of the others here and that's the reason I asked. If I'm going to learn, it'll be by trying and when I fail, asking those I trust.

777funk said:

Thanks for the replies. And 100% respect for John Roberts by the way. I've read his posts here and other places; I have no doubts about his credibility/opinion/statements that's for sure. He's a pro and has quite a background from what I've read. I was just pointing out what I have and haven't done (read parts of the data sheets, read some opinions, tried the chip in the past, etc). I definitely don't have the knowledge and especially experience of J.R. and many of the others here and that's the reason I asked. If I'm going to learn, it'll be by trying and when I fail, asking those I trust.

Click to expand...

Both parts are unity gain stable and JFET input so "should" be compatible. Andy's comment about the newer part having more gain bandwidth is subtle but could explain being more sensitive to stray capacitance or phase shift in the NF network.

The general criteria for stability is that forward open loop gain drops below unity by the time that lag or delay, or phase shift in the Negative feedback = 180', and becomes positive. To keep the TL074 stable you only need to worry about a unity gain intercept of a few hundred KHz. The 4134 OTOH has a unity gain intercept up around 8 MHz, so it will take far less lag/phase shift to get that one unstable.

Of course also check for the obvious, that you did not inadvertently change something with you part swap.

Depending on the actual circuit and how used, this superior high frequency loop gain margin will deliver lower distortion and lower phase shift, while for a low closed loop gain application this will probably be inaudible.

Thus my suggestion to understand the circuit first, since a low closed loop gain will also be harder to keep from oscillating. (High closed loop gain attenuates the NF path increasing stability margin. Low closed loop gain has less NF attenuation). 

JR

777funk said:

I just swapped one of the TL074's in my Line 6 Toneport KB37 to a Burr Brown OPA 4134 and now the mic pre is super fuzzy sounding like a transistor out of bias.

Click to expand...

As your unit is USB powered, it might simply be running out of steam, considering the 4134 is drawing more than 3 times the quiescent current of a TL074.

I'll take it apart again and inspect carefully on the soldering first. I'd read somewhere that caps can help keep it stable when a substitution is made. I'm not sure if that was on the power rails or in the feedback path. I'd guess .1uF's if on the rails and I have no idea if in the negative feedback path (or if that was the case at all). Too bad I don't have a schematic of the Toneport. Not really worth the time to tear it down that much anyways and with this being a keyboard I'm always a little cautious to get to crazy with the disassembly. It's just one chip I swapped so I'd think they'd have allowed for enough current in the supply. But you never know how close they cut it with the power on this unit.

777funk said:

I'll take it apart again and inspect carefully on the soldering first. I'd read somewhere that caps can help keep it stable when a substitution is made. I'm not sure if that was on the power rails or in the feedback path. I'd guess .1uF's if on the rails and I have no idea if in the negative feedback path (or if that was the case at all).

Click to expand...

You definitely need the bypass caps on the rails.

-a

777funk said:

I'll take it apart again and inspect carefully on the soldering first. I'd read somewhere that caps can help keep it stable when a substitution is made. I'm not sure if that was on the power rails or in the feedback path. I'd guess .1uF's if on the rails and I have no idea if in the negative feedback path (or if that was the case at all). Too bad I don't have a schematic of the Toneport. Not really worth the time to tear it down that much anyways and with this being a keyboard I'm always a little cautious to get to crazy with the disassembly. It's just one chip I swapped so I'd think they'd have allowed for enough current in the supply. But you never know how close they cut it with the power on this unit.

Click to expand...

First, determine or confirm that it is oscillating. Throwing bandaids at it without understanding what is going on not the path to enlightenment, but design based on guesswork or superstition.

Regarding stability, both opamps are unity gain stable so they can tolerate a feedback cap, which forces 100% negative feedback at high enough frequency.  If there is no feedback cap, any stray capacitance present at the negative input (and connected PCB trace) will create a LPF introducing lag to the NF signal that could degrade stability. This feedback cap needs to be significant relative to the stray capacitance that should only be picofarads. (The data sheet specifies 5pF common mode, 2 pF differential input Z while pcb traces could add more).

Adding capacitance to PS rails is a different matter and I expect there already is some (?). The premise here is that the opamp's PSRR is not adequate to reject HF noise on the power supply from being injected into the input. While power supply noise is not literally injected into the input stage, that is how the specification is characterized and measured. The PSRR for the 4134 is specified as down 90dB minimum.

However if you add caps across the PS rails that are low impedance at HF, they may couple HF PS noise into an otherwise clean ground at more that -90dB, so take care that your remedy does no harm. 

I suspect I sound like a broken record by now but it is generally useful to understand any circuit before modifying it, and understand what your modifications are doing to hopefully get a proper result. 

JR

777funk said:

I'd read somewhere that caps can help keep it stable when a substitution is made. I'm not sure if that was on the power rails or in the feedback path. I'd guess .1uF's if on the rails and I have no idea if in the negative feedback path (or if that was the case at all).

Click to expand...

Read Kingston's thread I posted in #6.

The important caps on the rails are Electrolytics AT the OPAs..  They should be within 1".
___________________

Have you checked that the yucky sound goes away when you put the TL074 back?