I'm looking into building myself a custom mixer matrix for my effects in my effects loop. In short so that I can split the signal, run some series, other paralell, that sort of thing.
I don't really want to add a level adjustment to my signal path, just a simple unity gain mixer.
I'm pretty clueless about electronics, but from what I've gathered an active op-amp circuit for summing signals is preferable, but I've no idea as to which op-amps are preferable for audio.
any ideas ?
cheers
LPS91
I don't really want to add a level adjustment to my signal path, just a simple unity gain mixer.
I'm pretty clueless about electronics, but from what I've gathered an active op-amp circuit for summing signals is preferable, but I've no idea as to which op-amps are preferable for audio.
any ideas ?
cheers
LPS91
LesPaulStandard said:
Hi,
As our friend Nigel Goodwin sad, TL071 (single) or TL072 (double) will work very fine... In my expirience most of j-fet based op-amps will sound great for musical equipment. Of course, you can use, and better, more expensive opamps like LF356, LF357, OPA2604 etc.
regards
Nigel Goodwin said:
Well there are people out there that BELIEVE it has some special sonic qualities. FX pedals that have the 4558 chip will consistantly bring more money then the same versions without. In fact people are buying 4558's and selling them at a substantial markup on ebay in the guitar FX section!
For example, the Tube screamer came in 2 versions. one with a 4558 and one without. the units with will bring near $100 or more and without maybe $25-30.00
Again weather or not it has some special sonic quality is up to the individual user.
Zc
The professional audio industry like Fender or Marshall uses 4558 because your lower prices when comparing with OPA2604 for example. In my experience OP2604 sounds much better than 4558 family. The 4558 family is OPERATIONAL amplifier basically. Today we have many others " operational " amplifiers FOR audio. The AD825 is very good for it. But you will very satisfied with OP2604 ( DUAL ) or OPA604 ( SINGLE ).
The major difference in op-amps I've noticed is noise. In most cases, replacing a noisy 4558 with something better makes sense.
Several years ago I purchased a Peavy mixer which had 1458/4558 throughout. Replacing them with NE5532's in the preamp circuits made a big improvement. Tl072's are great for low noise audio too, but will oscillate in some circuits. Adding a 100 ohm resistor in series with the output usually stops it.
I've replaced op-amps in effects pedals with good results too.
Several years ago I purchased a Peavy mixer which had 1458/4558 throughout. Replacing them with NE5532's in the preamp circuits made a big improvement. Tl072's are great for low noise audio too, but will oscillate in some circuits. Adding a 100 ohm resistor in series with the output usually stops it.
I've replaced op-amps in effects pedals with good results too.
Hello Brion55,
Yes! The noise of 4558 family is higher than most available opamps out there. But the 5532 is very old design from Philips Semiconductors. You can try to use the OPA2604. It has very low noise ( lower than 5532 in practice ) and brings high quality audio because it uses new topologies for audio ( Folded Cascode for example ). The sound is great !
Good lucky !!
Yes! The noise of 4558 family is higher than most available opamps out there. But the 5532 is very old design from Philips Semiconductors. You can try to use the OPA2604. It has very low noise ( lower than 5532 in practice ) and brings high quality audio because it uses new topologies for audio ( Folded Cascode for example ). The sound is great !
Good lucky !!
Brion55 said:
I had to arrange PA for a gig my daughter's band were doing last year - with VERY little notice. I'd recently finished building a pair of 1x12's with piezo tweeters, and I'd been given a couple of old PA amps.
In the event I used a 120W four channel amp of no particular make, unfortunately it provided 120W into 8 ohms - and I was using TWO 8 ohms in parallel.
So I doubled the output transistors (23772's with suitable emitter balancing resistors), and fitted a lower voltage transformer - which was probably still a bit too high?.
The mixer section used 741's, so I replaced the front end of two of the channels with TL071's (or which ever the single one is?). Despite the fact these preamps were being used to give a LOT of gain, from 600 ohm mikes!, I really couldn't tell any difference between the two upgraded channels and the old 741 channels.
Anyway, the gig went well, and the amp survived, although it did get quite hot
try opa2604
my opinion, opa2604 is more like tube sounding,i like it so much!! i don't agree using opa2134 coz it sound mellow, gets off the mids, it is better for hifi, i don't like the sound when the fx is distorted, when clean the sound is to deep but it is a sensitive opamp, still if you have more $$ try opa627/637 but it's a single device like ne5534... try socket and pick different opamp there you'll find which one is your taste.. goodluckwildswan said:
I can't help thinking that this question should really be irrelevant. The idea of an op-amp is to produce a standard black box circuit component which is free from subjective effects. If you design circuitry which allows them to clip then you have gone outside of their remit. They will sound different as they have different clipping characteristics but they should just NOT be used this way. Nor do they need to be!
Why not put in a very soft clipping circuit with gentle progressive limiting early in the proceedings, before any op-amp which would ever run the risk of clipping, and prevent the problem right from the start?
Build a ladder of about 5 or 6 diode levels with differing series resistors for each and you can produce real gradual soft clipping as they come into play one by one gradually shelving off the gain, not just slicing off the tops of any signal.
This will also produce a predictable maximum signal level which can be arranged to be just below clipping level for the following op-amps, (or better, the following simpler discrete BJT/FET stages!). These will never overload and will never be a part of the "sonic signature". (Said with pairs of fingers waggling in the air like Yuppy speech marks). Their contribution will be totally swamped by other effects, like the limiter or any power amp clipping, (not too nice), or power supply effects.
With a tiny bit more complexity, dividing the diode network into 2 opposing halves with different characteristics, you can even make it assymetric as it should be to model the transfer characteristics of those simple valve circuits we all love.
A bit of valve sound built in without ever bringing the op-amp overload characteristics into the frame, and you don't need use the op-amp to get your distortion!
I've only ever seen one real example of what must be one of the best design methods in guitar amp signal clipping. Who would be so amateurish as to clip the signal with one diode, or even 2 in series? It is always going to sound rough and gross and not even approaching a valve amp overloading gracefully. The one place I have seen this approach used is here, it's worth a look:
http://www.dogstar.dantimax.dk/tubestuf/graphics/dz_tone.gif
Don't be put off by the rest of the circuitry, you are only really interested in the diode based section on the right, none of the active circuitry is needed, we have nice clean neutral sounding op-amps to take care of that .
Why not put in a very soft clipping circuit with gentle progressive limiting early in the proceedings, before any op-amp which would ever run the risk of clipping, and prevent the problem right from the start?
Build a ladder of about 5 or 6 diode levels with differing series resistors for each and you can produce real gradual soft clipping as they come into play one by one gradually shelving off the gain, not just slicing off the tops of any signal.
This will also produce a predictable maximum signal level which can be arranged to be just below clipping level for the following op-amps, (or better, the following simpler discrete BJT/FET stages!). These will never overload and will never be a part of the "sonic signature". (Said with pairs of fingers waggling in the air like Yuppy speech marks). Their contribution will be totally swamped by other effects, like the limiter or any power amp clipping, (not too nice), or power supply effects.
With a tiny bit more complexity, dividing the diode network into 2 opposing halves with different characteristics, you can even make it assymetric as it should be to model the transfer characteristics of those simple valve circuits we all love.
A bit of valve sound built in without ever bringing the op-amp overload characteristics into the frame, and you don't need use the op-amp to get your distortion!
I've only ever seen one real example of what must be one of the best design methods in guitar amp signal clipping. Who would be so amateurish as to clip the signal with one diode, or even 2 in series? It is always going to sound rough and gross and not even approaching a valve amp overloading gracefully. The one place I have seen this approach used is here, it's worth a look:
http://www.dogstar.dantimax.dk/tubestuf/graphics/dz_tone.gif
Don't be put off by the rest of the circuitry, you are only really interested in the diode based section on the right, none of the active circuitry is needed, we have nice clean neutral sounding op-amps to take care of that
.I agree - the diode ladder is a good idea.
there's also this thread:
http://www.diyaudio.com/forums/showthread.php?threadid=38141&perpage=10&pagenumber=1
If you want to try something simple, you could wire some germanium diodes or LEDs to ground on the output of an op-amp. both have a softer breakover than silicon.
there's also this thread:
http://www.diyaudio.com/forums/showthread.php?threadid=38141&perpage=10&pagenumber=1
If you want to try something simple, you could wire some germanium diodes or LEDs to ground on the output of an op-amp. both have a softer breakover than silicon.
The germanium diode thing is certainly better than silicon but still less controllable than this. The ladder also has the advantage that it needs to work with larger signals.
It really is a great little gizmo. It used to be used a lot years ago to accurately shape triangular waveforms into sine waves for signal generators. Low distortion sine wave generation = hard, linear triangular wave generation = easy, (and a good square wave generally comes as part of the bundle). Then reshape the triangle with a diode ladder to give the required sine wave.
It is surprising how low you can get the harmonic content if you use enough diodes and work out your resistors carefully enough. But in our case we are adding controlled amounts of harmonics to end up with a distorted signal. 'Aint life strange?
It really is a great little gizmo. It used to be used a lot years ago to accurately shape triangular waveforms into sine waves for signal generators. Low distortion sine wave generation = hard, linear triangular wave generation = easy, (and a good square wave generally comes as part of the bundle). Then reshape the triangle with a diode ladder to give the required sine wave.
It is surprising how low you can get the harmonic content if you use enough diodes and work out your resistors carefully enough. But in our case we are adding controlled amounts of harmonics to end up with a distorted signal. 'Aint life strange?
Burr-Brown is Different
I agree with the posts that recommend the OPA numbers. I've used Burr-Brown op amps before. They have a nicer tone quality than many others.
The OPA227/228 series can be had in single, dual, etc.
The OPA627/637 series is only available in singles, and yes, they are expensive.
I use OPA627s and OPA637s exclusively in my guitar amps for the front end. After that point, the OPA227s and OPA228s are fine.
For an effects pedal driver, you want the absolute minimum self-noise that you can get, since whatever the op amp introduces will be amplified many times over from pedal to pedal. Also, I disagree with your idea that it should be a unity gain buffer. Even a small amount of gain from a top quality op amp at the front end of the pedal array can overcome a bunch of noise in the pedals. As long as none of the pedals goes into clipping, it is an advantage to drive them a little harder. Remember, the advertised signal to noise ratio of the pedals is based on them being driven to about 90% of their maximum voltage swing. They will seem much noisier if they are driven lower than that.
I agree with the posts that recommend the OPA numbers. I've used Burr-Brown op amps before. They have a nicer tone quality than many others.
The OPA227/228 series can be had in single, dual, etc.
The OPA627/637 series is only available in singles, and yes, they are expensive.
I use OPA627s and OPA637s exclusively in my guitar amps for the front end. After that point, the OPA227s and OPA228s are fine.
For an effects pedal driver, you want the absolute minimum self-noise that you can get, since whatever the op amp introduces will be amplified many times over from pedal to pedal. Also, I disagree with your idea that it should be a unity gain buffer. Even a small amount of gain from a top quality op amp at the front end of the pedal array can overcome a bunch of noise in the pedals. As long as none of the pedals goes into clipping, it is an advantage to drive them a little harder. Remember, the advertised signal to noise ratio of the pedals is based on them being driven to about 90% of their maximum voltage swing. They will seem much noisier if they are driven lower than that.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.