Flatline Optical Compressor - Blend Mod verification (not 4:1)

[LaloFP]

Hi!

  Im working on the Flatline and I added a Blend Knob looking at RG Keen´s Panning for Fun article: http://www.geofex.com/Article_Folders/panner.pdf .

  I know that the Madbean´s 4:1 has something similar, but its a Compressed volume and Clean volume. I want a real Blend, so I can adjust that, leave there and adjust master volume without having to move 2 knobs.

  Do you see something that can be wrong? One concern is about the coupling capacitors marked at blue. Can I remove C2? C5 is doing the same, doesnt it? I know I will have to test the resistor values around the Blend  knob to get the valances Ok, but Im waiting to receive the Vactrol.

  Its intended for Bass and Guitar.

  My mod:

  The Flatline sch:

  The 4:1 sch:

[rankot]

Following

[Kipper4]

I'd breadboard it and see.
Your mod looks nice and simple, less parts, similar achievements right.
You may find it difficult to find the sweet spot with it though.

While you have it there try the 4:1 method and see which you like.
I have been know to get it wrong.......

Rich

[slacker]

That looks like an interesting mod. You can get rid of all the caps marked in blue if you connect lug 2 of the blend pot to 4.5 volts instead of ground. You also then don't need R2 because IC1A will be biased by the 4.5 volts from the output of IC2A.

[PRR]

[LaloFP]

really? haha super

So... why does the 4:1 has the R1? Isnt that better in any way?

I´ll try it in brreadboard, but I wanted to get sure if I wasnt making any design/theory mistake

[Kipper4]

In the 4:1 schematic C1 blocks the dc voltage between ic2 A and ic1 A, so there's a need to bias the ic1 A at 4v5.

Remove the C1 as ppr drawers. And there's no longer a need to bias ic1 A wth a separate resistor as Slacker points out.
Get rid of C1 (bonus less parts, cheaper).
To keep all dc voltages in the parish tie the blend wiper to 4v5 as suggested.
Should be fine.
Let us know how it goes.

[merlinb]

Now remove the overcomplication:

[LaloFP]

In the 4:1 schematic C1 blocks the dc voltage between ic2 A and ic1 A, so there's a need to bias the ic1 A at 4v5.

Remove the C1 as ppr drawers. And there's no longer a need to bias ic1 A wth a separate resistor as Slacker points out.
Get rid of C1 (bonus less parts, cheaper).
To keep all dc voltages in the parish tie the blend wiper to 4v5 as suggested.
Should be fine.
Let us know how it goes.

Thank you for the explanation! I will investigate more on the topic now

Now remove the overcomplication:

Its getting simpler and simpler!

Is there any advantage in the other method? Obviously with more resistors I can tweak better the levels. But, other than that, is any advantage of having the pot wired the other way with VGND in the middle lug?

For what I red, and understood, your option works because for one Opamp the VGND is the other one, and then you are blocking DC with the 1uf.

But, in the theory, isnt there other differences or advantages between both methods? I think RGKeen would prefer to use less components too.

Thanks!!!

[merlinb]

Is there any advantage in the other method?

The other method is derived from audio pan pots where you're panning a stereo signal left/right. As you turn down one signal you want the other to turn up, so the overall loudness appears to stay the same. This doesn't apply to your (mono) situation.

The simpler method can blend from 100% A to 100% B, whereas the other method can never get fully 100% either way.

[PRR]

Now remove the overcomplication

You are correct, sir.

[LaloFP]

Is there any advantage in the other method?

The other method is derived from audio pan pots where you're panning a stereo signal left/right. As you turn down one signal you want the other to turn up, so the overall loudness appears to stay the same. This doesn't apply to your (mono) situation.

The simpler method can blend from 100% A to 100% B, whereas the other method can never get fully 100% either way.

Awesome! Thanks guys. I will try when I get the vactrols and let you know to verify the schem 

[R.G.]

The caps are there for two purposes. One is to limit bass response, and where you want more bass response, making them be infinitely big (i.e. a short circuit) works well IF you have limited your bass bandwidth elsewhere. The second purpose is to break any DC amplification chains that produce bad results.  For captive applications, where you know your gains can't amplify offset voltages up into being a problem, you can replace the caps with short circuits.

As a historical observation, where I put short cuts into published circuits, I get comments saying things like "hey, this doesn't work I used it in [something where the gotchas added up] and it is [horrible, non-functional, useless, whatever].  So yes, by taking into account special cases, you can sometimes simplify.

Now remove the overcomplication:

And we have a second item where there are simplifications that may or may not be good to use, depending on the requirements.

The simple pot-between-two signals does indeed do a blend. It is, as in UK usage, a cheap and cheerful way to get a blend.

What it doesn't do unless you're picky about exactly how you set it up is a constant power panner blend. The panner circuit I put up isn't original with me. I used a National Semiconductor application note as its basis, and there are a number of the asterisks and footnotes explained there. The picky resistor values are intended to approximate a much trickier and more demanding set of panner pots. What it has going for it is that it's predictable about panning and only uses one pot, while sidestepping some of the issues and footnotes with the pot-between-two-signal-sources blend. What it does demand that you do is to feed it from low impedance (i.e. less than 1/10 the first series resistances) and use some more resistors and feed the results into low impedance summing nodes to do the addition properly. That may be too much effort for someone who has a simple application and is happy with the results of the simpler circuit. 

I have personally used the panner as shown in a commercial production pedal specifically because the pot-strung-between-two-signals didn't produce a flat enough and predictable enough result for the application. But then applications vary, as do the levels of pickiness about the results. I'm an agnostic - my watchword is - if it works, use it; but KNOW what you wanted it to do first.

So once again, if the single-pot-strung-between-signals works well enough, use it. In my mind, "overcomplication" implies knowing what level of complication is acceptable or required. Quoting Lewis Carroll's Cheshire Cat in "Alice's Adventures in Wonderland",

"Alice asked the Cheshire Cat, who was sitting in a tree, "What road do I take?"

The cat asked, "Where do you want to go?"

"I don't know," Alice answered.

"Then," said the cat, "it really doesn't matter, does it?"

[PRR]

"....when two people disagree on a thing they generally argue the pros and cons of it and each one will come out better posted on the subject under discussion."
Power Plant Engineering, April 1, 1922

[R.G.]

I agree.

[LaloFP]

Thanks RG!!

I have an update: i dont have yet the real vactrols (still in transit), but made a DIY one with a photoresistor and a led to start to test the circuit.

Now remove the overcomplication:

This one didint work, I only got clean signal, nothing of compression.

But with the C1 and R2 added it worked perfectly.

The only "problem" is that the compressed signal is much louder than the clean signal, so the volumes are not the same in all the positions. And that is something to expect, but...

If Im just able to lower the output of the compressed sound, it would be awesome, to have always the same volume: the full compressed volume when you hit it hard should be the same perceived volume like the volume when the full clean is hit hard.

I tried several things, like voltage divider before the Blend pot and other stuff, but everything gives me less than 100% Compressed when turned all the way to that. I can get the volumes right, but the Blend pot stops to being 100%Clean - 100%Compressed and gets less% Compressed.

Is there a simple way Im missing? (Im going to try the "Panning for Fun" method to see how it behaves)

[samhay]

>The only "problem" is that the compressed signal is much louder than the clean signal,
Have been watching this tread waiting for you to work this one out.

You could put a resistor between the output of the compression op-amp and the blend pot. This will lower the wet volume, but will also blend in some/more dry signal when the blend is all the way to the wet side.

You could set up a trimmer as an volume control (to Vb rather than ground) on the output of the compression op-amp and connect the wiper to the blend control.

However, in either case, this probably won't be very satisfactory as you will only get a narrow-ish range of compression settings where your volume matches. Perhaps better to just use the volume control on the output in tandem with the blend?

[LaloFP]

>The only "problem" is that the compressed signal is much louder than the clean signal,
Have been watching this tread waiting for you to work this one out.

You could put a resistor between the output of the compression op-amp and the blend pot. This will lower the wet volume, but will also blend in some/more dry signal when the blend is all the way to the wet side.

You could set up a trimmer as an volume control (to Vb rather than ground) on the output of the compression op-amp and connect the wiper to the blend control.

However, in either case, this probably won't be very satisfactory as you will only get a narrow-ish range of compression settings where your volume matches. Perhaps better to just use the volume control on the output in tandem with the blend?

Yeah, I tried those options and both didnt give 100% wet 

NEWS: I just tested the "Panning for Fun" method (like its shown in the first post here) and works better... BUT in the middle position the signals add up and sounds louder. On the extrems I can get the same volume changing R12 or R16, but when I blend, it gets louder (logically... ).

If we just could get constant volume it would be awesome, supper easy and usefull.

We can:
1 - Improve the pot-between-two signals to reduce the Wet signal before Blend allowing 100% Wet signal mix
2 - Improve the Panning for Fun to not get louder at Half Blend (but I think its less likely)

What do you think??

[merlinb]

What do you think??

Adjust gain of IC2A until you have equality.



EDIT: I meant to say IC2A

[LaloFP]

What do you think??

Adjust gain of IC1A until you have equality.

I thought about that, but if I decrease the IC1A´s feedback resistor I will get less range of compression, because the minimum gain is fixed by the Vactrol. And, if I feed that opamp with less current, I will get less compression too 

The other think I thought of: feed IC1A(CompOpamp) with signal before IC2A(FirstBuffer) and raise the gain of IC2A. I think that will work, BUT it will increase noise 

haha

Thats why Im tryin to get the first idea to work:
1 - Improve the pot-between-two signals to reduce the Wet signal before Blend allowing 100% Wet signal mix
2 - Improve the Panning for Fun to not get louder at Half Blend (but I think its less likely)
3-  Feed IC1A(CompOpamp) with signal before IC2A(FirstBuffer) and raise the gain of IC2A. I think that will work, BUT it will increase noise