"Crush-n-Blend"

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

SSLtech

Well-known member
Joined
Jun 3, 2004
Messages
5,447
Location
Florida (Previously UK)
Try this out for size:

For all those who keep asking for a "blend" control for an SSL comp: this is a universal 2-channel (but if you build half, it can just as easily be mono!) variable blend control between a hard-crushed compressor and a direct, un-compressed signal.


(CLICK FOR FULL-SIZE IMAGE)

The benefit to a stand-alone design (to my way of thinking) is that you can use it on an 1176 pair, a GSSL, a couple of LA2a's, or whatever you like. it doesn't tie up 4 console channels, you can insert it clean across a rhythm section subgroup and dial-in the amount of crush.

The best directions for use: Set the blend control to "direct", set your gains. Set the blend control all the way to "crush" then adjust the makeup gain of the compressor for similar signal levels. Once this is done, you should be able to blend anywhere in between with the overall level remaning similar, just the amount of compression varied.

Keith
(Edited Dec. 12 '06, to shrink image thumbnail for readability)
 
Brilliant.. i might just build one real soon with all the comps i am starting to build.

Hell this might be interesting on Eq as well.... or anything really like wet to dry FX blends... because i dont always like how some Fx units do dry/wet blending.


any alternatives to the SSMs ?
 
Burr-Brown INA134s for the 2141s,
Burr-Brown OPA134s for the 2142s,
Some other offerings from THAT corp, from what I hear. Socket them if you like, just don't shove phantom power up their Jacksie, or they'll go on strike! :wink:

Keef
 
depending on which fx unit the Wet signal is coming from , especially if used for multieffects or even EQ.. maybe a comp or two) there will probably be a little bit of a phase issue. but there should not be if you were to plug the send directly to the return....( or am i wrong)
 
For analog compression, there shouldn't be. All the usual suspects should work perfectly.

If you try something with any A/D and D-back-to-A in the audio path, then there'll be a delay "phaser-stuck-at-half-sweep" sound.

Phasing with analog EQ... well, ya see here's where things are too often misunderstood: I'm generalising a little bit here, but most parametric EQs do their job by "sniffing" a little bit of the signal (in either polarity, depending on cut or boost mode, and at an amplitude proportional to the amount of cut or boost) and then feeding that to a tuned circuit, which is regenerated slightly. The center frequency of the tuned circuit is the only frequency which is at 0° phase shift. Above that might be progressively advanced, below might be progressively retarded, or vice versa. Anyhow, the frequency that therefore regenerates most efficiently is the center frequency...

Now, if all we heard was the regenerated filter, we'd only have a band-pass response. The band pass is added to the direct signal and so we get a boosted response in the region of the center frequency, or (if the polarity is flipped) a cancellation dip in the region of the center frequency.

Anyway, here's the upshot: There's no theoretical benefit to doing this, but there should be no "phase shift" problems preventing you from trying it out and proving or disproving the point if you wish. Potential disadvantages are that running at high levels of boost will significantly reduce your headroom before clipping.

The (rather over-simplified) description of how this notional parametric EQ works is applicable to Neve 81/82/V-series, SSL E/G/J series, Calrec UA8000, Amek M2500/M3000 and later parametrics. There might be other topoliogies which work differently, but I'd expect that a fundament is that for something to add to a signal and result in boost at a particular frequency -be that the center frequency or a proportionally related frequency- then it has to be fully (for center freq.) or partially (for other freq's) in-phase. If so, and no delay has occurred to the 'direct' portion of the signal, you should be golden.

In case your EQ/Compressor/gate effects a polarity reversal, a DPDT changeover for polarity switching might be a wise precaution.

For unbalanced use, the 2142s can be omitted, and the 2141s can be replaced with a simple unity gain inverting buffer.

Keith
 
Updated... realised that for unity gain, R5 and R10 needed to be 3.3333k instead of 4.99k... (unity gain at either end intead of at the center... not thinking straight! :roll: )

Keith
 
just don't shove phantom power up their Jacksie, or they'll go on strike!
What if you clamped the input to the SSM2141 it wouldn't be the perfect solution, but it could protect them? Correct me if I'm wrong.

Analag
 
The data sheet shows how to get it handling phantom, but this seems like it would spend most of its time used after the recording is all done anyways.

adam
 
[quote author="analag"]
just don't shove phantom power up their Jacksie, or they'll go on strike!
What if you clamped the input to the SSM2141 it wouldn't be the perfect solution[/quote]
That's the input.

Phantom is likely to appear at the output. ("Jacksie" or "Jaxie" is British slang for backside! :wink: ) It's the 2142 that will take the hit.

I've previously posted an output setup which will protect the 2142; it's just an electrolytic (with the positive plate towards the outgoing socket and the negative plate towards the 2142) followed by a 10k to ground in each leg. The 10k will load the 48V phantom down to a manageable 25V or so, and the cap can handle the rest.

Yes, it should be done after the recording, but that often times doesn't stop people from thinking "Balanced XLR out, huh? -Well, the only balanced XLR inputs that I have available are these mic inputs... I can put the pad in and it'll be fine!" ...forgetting of course that they left the phantom button in on one channel! :wink:

Keith
 
I was thinking about building something similar, based on the wet/dry tube drive circuit from my Drawmer 1962 but now I will build yours. Simple and efective.
Thanks Keith.
:thumb:

chrissugar
 
Keef, This seems like it would also work nicely as a simple 2x1 line mixer. I have to build one of these soon.

Thanks for the design,
Chris
 
In fact yes. You can use it to crossfade between two stereo sources. The crossfade law should be -3dB at the center if the sources are uncorrelated, -6dB if they're correlated, to produce equal apparent loudness throughout the rotation of the control... all you need to do is change the resistors feeding the potentiometers, and adjust the feedback resistors for unity gain at both extremes.

Keith
 
thank you...!! I always wanted to build something like this...


can the opa2134 be replaced by another opamp with comparable quality??
I can only find that type in the small smd version... no dip...


thanks,
matthias
 
Sure... TL07x series shouldn't be excessively taxed for output current in this application, and they should have similarly low DC offset (less scratchy pots without having to have capacitors everywhere...)

Or socket it and play at will, -as you see fit.

Keith
 
if I want to build that ciruit in an external box....

I'd like to use burr brown drv134 and ina 134 line driver/receiver, because I don't have a source for the ssm-chips...

do I have to add some coupling capacitors to the inputs and outputs ??

I looked at the datasheets and there's mentioned a 10µ bipol. cap at the output only ... and that's optional...

can anyone help me..??

thanks,
mat
 

Latest posts

Back
Top