Will a 2n5089 vs. a 2n5088 be quieter loess noisey in my Ross Comp

[matt239]

Cool, I didn't know that. Are those straight replacements for the CA3080 or is tweaking required for a circuit that would normally use a CA3080?

Right. Is there a drop-in replacement for CA3080 that is quieter?
I know some of these chips are no longer made, but there's some old stock still around.

[midwayfair]

Cool, I didn't know that. Are those straight replacements for the CA3080 or is tweaking required for a circuit that would normally use a CA3080?

Right. Is there a drop-in replacement for CA3080 that is quieter?
I know some of these chips are no longer made, but there's some old stock still around.

You can make a daughter board for the 13700 to run wires to the board ... But honestly, best to just build a better compressor. There's only so much you can do about the noise.

Gawd I hate the Ross comp.

[Thecomedian]

R.G. - respect !
The best explanation of the difference between the " input behaviour" of the usual OpAmp and the OTA I've ever read!

Yes, this is why I love to ask.  I learn so much.

yeah, R.G. knows his stuff, so it's trustable.

[Mustachio]

Aww really ? I love the dynacomp , I built one with the janglebox mod, had to change a few cap values to keep my tone correct but it was great! I gotta build another I have one old tin hat ca3080 left!

I'd say I agree with you about overall better comps out there but I'd definitely use a dynacomp often , its got that classic squeezed to much chicken picken sound!

Here's a easy daughter board to use a lm13700 or lm13600 in place of the ca3080

[bluebunny]

But honestly, best to just build a better compressor.

I would happily recommend building Jon's Bearhug compressor, or Merlin's Engineer's Thumb.  Both do the job very well indeed and are very quiet.

Edit: I should say that I've built the Dynacomp/Ross too.  It's not bad.  My tame guitarist is very happy with it.    He swapped it for his four-knob Keeley, which he no longer likes.   

[R O Tiree]

Edit: I should say that I've built the Dynacomp/Ross too.  It's not bad.  My tame guitarist is very happy with it.    He swapped it for his four-knob Keeley, which he no longer likes.   

... which is bizarre, because a Keeley is a Ross comp with a tweak or two. Hence your smiley, Marc?

This always makes me smile (from the Keeley comp blurb on their website)...

The Keeley Compressor effect has nothing but tone tested components and true bypass switching and the effect switching incorporates a triple pole-double throw switch! The user will marvel at the clean unaffected sound when in the bypass mode and the compressor has an LED for effect indication.

No, really? Who'd have thunk it, eh?

And this...

Hand matched transistors to less than 1% tolerance and ensures the musician receives a perfectly compressed signal with no chance for unwanted distortions.

<sigh>

[bluebunny]

Edit: I should say that I've built the Dynacomp/Ross too.  It's not bad.  My tame guitarist is very happy with it.    He swapped it for his four-knob Keeley, which he no longer likes.   

... which is bizarre, because a Keeley is a Ross comp with a tweak or two. Hence your smiley, Marc?

Yep, smiley was entirely intentional!      And his was a four-knob C4 too.  As for the b0ll0ck5 on the Keeley website, it does make me want to spit a little...   

[matt239]

Cool, I didn't know that. Are those straight replacements for the CA3080 or is tweaking required for a circuit that would normally use a CA3080?

Right. Is there a drop-in replacement for CA3080 that is quieter?
I know some of these chips are no longer made, but there's some old stock still around.

You can make a daughter board for the 13700 to run wires to the board ... But honestly, best to just build a better compressor. There's only so much you can do about the noise.

Gawd I hate the Ross comp.

"But honestly, best to just build a better compressor"

Oh I'd have to agree, the design of the dynacomp is pretty outdated, and doesn't allow for really low noise -"best practices.."

I have other favorites to build, I just happen to have an actual dynacomp in hand, and it's so noisy, I wouldn't even feel right about selling it, (or even giving it away..)
So I'd like to try to improve it at least a little!

[Paul Marossy]

Oh I'd have to agree, the design of the dynacomp is pretty outdated, and doesn't allow for really low noise -"best practices.."

I have other favorites to build, I just happen to have an actual dynacomp in hand, and it's so noisy, I wouldn't even feel right about selling it, (or even giving it away..)
So I'd like to try to improve it at least a little!

I have a newer dynacomp. I like it to a point (I mostly hate compressors, I never use them), but I can't use it because it's WAY too noisey with any kind of overdrive pedal - it amplifies all the noise from the compressor. 

[mistahead]

Oh I'd have to agree, the design of the dynacomp is pretty outdated, and doesn't allow for really low noise -"best practices.."

I have other favorites to build, I just happen to have an actual dynacomp in hand, and it's so noisy, I wouldn't even feel right about selling it, (or even giving it away..)
So I'd like to try to improve it at least a little!

I have a newer dynacomp. I like it to a point (I mostly hate compressors, I never use them), but I can't use it because it's WAY too noisey with any kind of overdrive pedal - it amplifies all the noise from the compressor. 

Bearing in mind we are talking about compressors - which by nature screw with the "Dynamics" of the tone...

I might be a little deaf, but "noise" from compressors that I've played with generally seems to be a hiss in a specific frequency that doesn't really change, oh but I am talking about lets call it IDLE AND EVER PRESENT NOISE... if that is the noise that people are unhappy with would it be possible to limit that freq. specifically? Or ghetto it out via a highpass and lowpass for either side of it and blend the two back together?

Or are we just meaning that compressing the hell out of our signals makes them noisy - for the sake of a name we'll make that one ACTIVE SIGNAL PROCESSING NOISE?

[Paul Marossy]

I might be a little deaf, but "noise" from compressors that I've played with generally seems to be a hiss in a specific frequency that doesn't really change, oh but I am talking about lets call it IDLE AND EVER PRESENT NOISE... if that is the noise that people are unhappy with would it be possible to limit that freq. specifically? Or ghetto it out via a highpass and lowpass for either side of it and blend the two back together?

Yeah, I've noticed that hiss is at a fixed frequency. In the case of the Dynacomp, seems to me that a lot of the noise comes from the OTAs (Operational Transconductance Amplifier) themselves. CA3080s seem to be pretty noisey devices in general when it comes to the hissing noise. I built Craig Anderton's "Volume Pedal DeSratcher" (which uses a CA3080) and it was very hissy, too much actually.

[PRR]

> compressing the hell out of our signals makes them noisy

Brings-up small hiss into large hisss. That would happen even with a hiss-less compressor.

> "noise" from compressors that I've played with generally seems to be a hiss in a specific frequency

Guitar pickup resonance. Pickup winding inductance resonates with winding and cable capacitance around 3KHz. Aside from jacking-up the pickup's inherent thermal hiss, it doesn't shunt any hiss-current in the input of the next amplifier stage. Present pickup systems are sure to have a hiss-peak around 3KHz.

This can easily be avoided: ask the winder for "weak and dull", fewer turns. But that's usually not the tone that guitarists want. That upper-end resonance puts the zing on the strings.

[mistahead]

So essentially I am right about the static nature of the hiss (in most compressors, maybe not in this exact case), but screwed in the solution because the hiss is overlapping the "mojo frequency" and reducing it will defeat the purpose?

[matt239]

Good quality "studio" or pro-audio compressors contribute very little noise of their own. They certainly will make the noise in a noisy signal more apparent..
At extreme settings they will bring up thermal noise, but those kinds of settings are really not necessary or useful.

Dynacomp/Ross style compressors use a noisy OTA chip, (some worse than others) and one that can't tolerate a high input voltage, so the signal has to be reduced to a low level, and re-amplified; a recipe for more noise.
It really is an outdated design. I can't believe they're still so popular.
- It's because a lot of famous guitarists are known to have used them, but that's because at one point, Dynacomps were one of about 2 available "guitar" level compressors, and the other one: Orange Squeezer, while quieter, didn't compress enough for some kinds of special effects compression, and wasn't adjustable.
Even since then, dynacomp had a response that was more pleasing to guitarists than some of the grossly over-simplified optical and "VCA" guitar pedal designs.

But none of this has to be so anymore; better optical and VCA designs exist from major pedal-makers, and in the D.I.Y. community.
Even some of the tech from mid-priced "rack-mount" compressors could probably be put into a $120 guitar pedal with 5 or so knobs to control basic functions.

I'd bet you could even get a Beringer Autocomm, to respond almost exactly like a Dynacomp, but with very little noise..

I'm just trying to decide what to do with this one poor little Dynacomp I happen to have..
I'm seriously considering scrapping the whole circuit and building something else in the box!

Still, if I could get some reasonable improvement from changing the chip, and two transistors, and a few caps and resistors, I could either use it, or pass it along..
For now, it sits on my bench, awaiting it's fate!

(Re; cutting the offending frequencies: I don't think there is only one band in which these things make noise, it's more like white noise, but is more apparent to us in the higher frequencies. You could cut a lot of treble, and reduce the apparent noisiness, but you'd also lose all your guitar treble, ending up with a muddy/squishy mess.. )
(There might be some benefit to pre- and de-emphasis; boost treble before compression, complimentary cut after.. )

- I suppose if someone was really addicted to Ross/Dynacomp compression, but couldn't abide the noise, the thing to do would be to put a nice limiter before the OTA to pre-manage the peaks, and then not divide down the level quite as severely before the OTA.. but that's a lot of trouble to go to to keep using a flawed design, and there would still be noise..

I'm not convinced even the compression curve of Ross/Dynas is all that special/amazing. I defy anyone to tell me a guitar would sound any "better" through a Dynacomp, than through a nice mic-pre and an LA-2A..  

[Mark Hammer]

For the most part, compressors are designed to hiss.  By "designed to", I don't mean intended to.  Rather, what they ARE intended to do inevitably results in audible hiss

A few things to consider:

Optimal S/N ratio will always happen when stage/device B is fed the highest possible clean level that can be fed from stage/device A.  This truism is gospel within analog audio engineering because ALL analog systems will generate some minimal amount of hiss, so the less gain any stage has to apply, the less advantage will be afforded any hiss added to the incoming signal.  If A has minimal hiss and a hot output, then feeding it to B in a way that requires less amplification means that not only will A's hiss not be amplified much, but whatever hiss B contributes won't be amplified much either.

Compressors are designed to provide maximum amplification when the input signal is at its lowest.  So when the S/N ratio of the input signal is worst, the compressor will amplify it even more.  Genius, sheer genius.  It is common enough that audio engineers have a term for this noise boost: "breathing" (because it sounds like a long and loud inhale).

If that was not bad enough, the gain element used in a great many guitar compressors has a very low tolerance for input level, distorting at very low amplitudes.  So, if it wasn't bad enough that max gan is applied when S/N ratio was at its worst, stompbox compressors defy yet another rule of noise-management in analog circuits: attenuating first, then amplifying second will yield worse noise characteristics.  So the signal being fed to the 3080 is severely attenuated to make sure the 3080 won't clip, and then it is boosted again to bring the level up to where it should be to have reasonable compress/bypass volume-balance.  The 3080 will generate hiss on its own.  If the signal fed to it was a hot one, then that hiss would be a very tiny component of the output.  But since the input has to be attenuated to prevent clipping, that internally-generated hiss becomes a much bigger component of the output.  And of course, when gain is added to the output, the hiss that arrived with the input to the pedal is once again boosted.

All in all, it is a recipe for max hiss.  The best way to avoid it is to either a) just keep playing (like that movie "Speed" where disaster will strike if you stop or slow down), or b) stick a suitable gate or other noise-control device before the compressor to prevent any breathing.

[matt239]

Isn't that what I just said? LOL
My primary experience with compressors/limiters has been as a sound man, rather than as a guitarist.
I'd still say; if you have a GOOD, FULLY ADJUSTABLE compressor/limiter, set up and USED PROPERLY, it doesn't have to add an unacceptable amount of noise.
For one thing, in pro audio, done well, they're almost never set up to compress at all times - to amplify the lowest signals (and the noise)
and the general rule is to use as little compression as possible.
-So we're kind of using them as limiters..  

On signals that required a lot of level management, one trick I used to use was to set one compressor with a lower threshold, but lower ratio, (like 1.5:1) (and still not set to be always acting..) and another as a harder limiter: 10:1 but with threshold set to only clamp the loudest peaks.
Of course none of this is possible with a one or two knob, antiquated OTA-based compressor.

Noise gates, or downward expansion is a handy tool to have, but in my mind it's a last resort for managing noise, you only apply them after you've done every other possible thing to reduce noise.
(Completely other consideration than using gates to tighten up a drum track etc.)

So a Ross/Dyna is just not up to my standards.
That said, I've heard recordings of them used on guitar, that weren't completely horrible in terms of noise.
And this one I have in front of me seems above average noisy, even for a typically noisy Dynacomp!
So just in the interest of not wasting things, I thought maybe I'd try to "fix" it, but at this point, I may well just build something else in the box!  

[PRR]

> right about the static nature of the hiss (in most compressors, maybe not in this exact case), but screwed in the solution because the hiss is overlapping the "mojo frequency"

Yeah. And worse than in many other audio problems.

A dynamic microphone can be treated as a near-pure 200 ohm resistor. This has a known hiss profile, White Noise, uniform at all frequencies.

Though because of the way we hear, the high frequencies generally count for more; and when a sound has few high frequencies, it is useful to high-cut the amplified signal and lose more hiss than sound.

Guitar is a bit different. The size of pole-pieces mean that highs are not picked-up well. Various other losses damp the harmonics. But incidentally, the inductor-capacitor network emphasizes some range of highs. Careful use of fine wire can give a pickup a LOT more ZING. Play it loud-- WOW!

But the same circuit also boosts the thermal hiss that is universal and intrinsic to useful electronics.

I was only 99% sure that this matters in guitar work, so I asked the idiot computer to apply the well-established laws of circuits and noise.



Top plot shows how the wiggly magnetic field in the pickup coil is leveraged to signal voltage at the amplifier input. The values are typical: different windings and different cables will shift the peak, but not a whole lot. Resistive loading will reduce the peak, and this sim does not include the Volume pot found in most guitars.

Bottom plot is what this circuit puts-out when strings are stationary, silent. At any reasonable temperature, electrons vibrate randomly. In a resistance, we get random thermal voltage/current. This energy is further leveraged by any reactances (inductance, capacitance).

BTW, the Noise plot is shifted-up to get it on the same plot as the nominal 1V response. The actual value averages -160dB per root Hz, but peaks *20dB* higher.