Series/parallel loop - feedback needed

[greiswig]

Hi,

I'm trying to design an FX loop that will be switchable between series and parallel. The idea would be that the B+ for this would have another node on a Dumble dropping string, and this would be integrated into the amp. So the input and output jacks in this would just be hard-wired in.

Please give me some feedback if you have a chance...this is not much like a Dumbleator, but then I can't find any circuits that have a parallel option for an FX loop.

I'm especially wondering about the feedback circuit on V1b, since the gain is going to be affected by that. I'm not sure how to make a "recovery" control that keeps things pretty transparent, but I was wondering if a large pot there would work.

Thanks!

[Bob-I]

You'll need a better mixer stage to make that work really well. The dry signal part will degrade pretty badly the way you have it drawn.

I did a lot of experimenting with a parallel loop and decided that a feedback amp worked best for the mixer. By feedback amp I mean a true feedback amp stage, not just a common cathode with local feedback. Check on aiken's site for more information.

The send part is is sound but the return doesn't look correct to me. What I ended up with coming off of the return jack is a 10K into the return level pot, then into a .033uF > 100K into the grid of feedback amp. The dry part comes off of the same location as yours but goes into another .033uF > 100K into the grid as well. The feedback is off the plate into a 2.2uF into a 220K into the grid.

There's NO GROUND reference on the grid as the junction of the feedback/dry/wet signals acts as a ground reference. This is important to this is design as the grid reference resistor makes this a common cathode amp, not a feedback amp.

I've built 3 of these, one in a 1U rack and 2 internal and they all sound great.

I'd recomend that you bread board this out beofre committing into the amp.

[drz400]

Hi,

I'm trying to design an FX loop that will be switchable between series and parallel. The idea would be that the B+ for this would have another node on a Dumble dropping string, and this would be integrated into the amp. So the input and output jacks in this would just be hard-wired in.

Please give me some feedback if you have a chance...this is not much like a Dumbleator, but then I can't find any circuits that have a parallel option for an FX loop.

I'm especially wondering about the feedback circuit on V1b, since the gain is going to be affected by that. I'm not sure how to make a "recovery" control that keeps things pretty transparent, but I was wondering if a large pot there would work.

Thanks!

You wont find it done very often since tubes do not posses the proper characteristics to be good performers when it comes to negative feedback summing amp mixers.
Look for an Egnater TOL50 schematic
That works OK and is very similar to the Kevin Oconor version, As much as I have tried Mix loops with tubes next to an Opamp for a mix they are very poor performers, both in tone and isolation if you are not looking for a tone change.
It is difficult to get enough return gain without getting the feedback and input resistors too wacked out

One that does work well with tubes is the old 5E8A input, mix after the plates, however it isnt practical using 2 tubes, one for the return and 1/2 for the send

[greiswig]

Thanks, all. Well, I'm not married to the idea of using a tube. And I'm not necessarily interested in a DIY solution. What I am interested in is something that allows parallel operation, variable mix of wet and dry, proper impedances at both ends, variable send and return.

The other thing is that, ideally, I'd like the equivalent of a parallel Dumbleator. I've not used one, but I gather that there's more to what it does for the tone of the amp than merely rolling off some highs due to the cable capacitance. So perfect transparency isn't necessarily the goal. If I read correctly, part of the magic of the Dumbleator is the added harmonic content, which presumably *is* better using a tube than the alternative.

But I'd settle for a simple, transparent parallel loop of some kind. Suhr has a MiniMix, but it only is supposed to work with line level signals. Is there a circuit diagram or aftermarket solution out there for people who want an Ironsounds-type loop, but want a parallel loop instead of serial?

[phalanges]

You've probably seen this:
It's the public file Kleinulator that was posted a couple of years ago at BN.
It works well for me.

[Structo]

Here is the schematic.

[greiswig]

Thanks, guys! Nope, I had only seen the serial version. And thanks, Mr. Klein!

I may give this a go. Is this design on the "transparent" side of things, or on the "Dumbleator coloration" side?

Maybe another way to ask it...is the coloration in a D'lator due to the circuit itself, or is it because of the way it drives the PI, cable capacitance, etc.?

[phalanges]

Seems pretty transparent.
The return becomes your master volume when it's in the passive loop and as long as I keep the amp's master at 9 o'clock or under, there's plenty of headroom.