It's worth mentioning that with a parallel loop it's important that no dry signal goes through the effect. It's very possible that the dry signal is out of phase with the dry signal in the parallel loop. I had reall difficulty getting a decent sound from a parallel loop until I figured this out.
My experience as well. I never use a serial loop. Ever. I find that the tone is always negatively impacted to a small or large degree. Even with high end equipment. I haven't used the 2290, but I understand it is essentially a parallel loop implementation.
As previously implied. You cannot add the number of gain stages and determine the phase of the signal. It depends on the type of each stage in the chain. Some stages are inverting and some are non-inverting.
martin manning wrote:I'm not exactly sure what Bob is talking about, but since the first stage in a D'lator is a non-inverting CF the "send" out of the D'lator is the same phase (grossly) as the "send" out of the amp, i.e., the dry signal. Now, each of the individual FX in the loop may or may not invert the signal, as well as shift its phase by some amount. Then, the recovery stage of the D'lator will definitely invert whatever comes in at its return jack. With all this going on, it's hard to say where the returning wet signal will be in relation to the dry, and mixing them together might have undesirable effects (or FX, if you like). From what I understand, this is not much of an issue for time based FX, but gain based FX might suffer from feedback issues and/or phase cancellation and loss of volume.
MPM
Yes, a time based effect, when both the dry and delayed signal come through the effect, may cause phase related problems. I had that experience when I first built the parallel loop. The Quadraverb simply sounded like garbage where the MIDI Verb sounded ok.
I'd never use a gain based effect in the loop. JMO
martin manning wrote:I'm not exactly sure what Bob is talking about, but since the first stage in a D'lator is a non-inverting CF the "send" out of the D'lator is the same phase (grossly) as the "send" out of the amp, i.e., the dry signal. Now, each of the individual FX in the loop may or may not invert the signal, as well as shift its phase by some amount. Then, the recovery stage of the D'lator will definitely invert whatever comes in at its return jack. With all this going on, it's hard to say where the returning wet signal will be in relation to the dry, and mixing them together might have undesirable effects (or FX, if you like). From what I understand, this is not much of an issue for time based FX, but gain based FX might suffer from feedback issues and/or phase cancellation and loss of volume.
MPM
Yes, a time based effect, when both the dry and delayed signal come through the effect, may cause phase related problems. I had that experience when I first built the parallel loop. The Quadraverb simply sounded like garbage where the MIDI Verb sounded ok.
I'd never use a gain based effect in the loop. JMO
Excuse me while I shout the obvious - NO MANUFACTURER OF EFFECTS CHANGES THE PHASE BETWEEN DRY AND WET. They may flip the phase of both but the phase relationship in and out is always maintained. A serial loop insures that you will not have out of phase problems because the effect has to do the mixing. All bets are off when you go parallel because you are at the mercy of the effects design and may very well result into a wet out of phase with the initial dry.
heisthl wrote:Excuse me while I shout the obvious - NO MANUFACTURER OF EFFECTS CHANGES THE PHASE BETWEEN DRY AND WET. They may flip the phase of both but the phase relationship in and out is always maintained. A serial loop insures that you will not have out of phase problems because the effect has to do the mixing. All bets are off when you go parallel because you are at the mercy of the effects design and may very well result into a wet out of phase with the initial dry.
I didn't say they'd change the phase between dry and wet. Calm down and don't shout at me.
What I said was that the phase can change through the effect unit, often it does. If you mix a dry signal with a reversed phase it will cause phase related issues. The delayed signal will not, even if the phase is reversed.
So you're ok in a parallel loop as long as you don't send a dry signal through the effect.
Now seriously, stop shouting, we're all here to build amps, there's no reason for nastyness.
I was kidding - My tongue was firmly planted in my cheek - because there are no real Dumbleators built as parallel loops meaning it never was or never will be an issue with a real one.
heisthl wrote:I was kidding - My tongue was firmly planted in my cheek - because there are no real Dumbleators built as parallel loops meaning it never was or never will be an issue with a real one.
The Dumbleator was designed in the late 70's or early 80's, at a time when every manufactorer used serial loops. Thats all the knew back then.
When the big 'rack revolution' hit the guitarist scene back in the mid 80's, every guitarist had to drag a large rack with the same studio effects they'd used on studio recordings in order to get the same sounds on stage. All these early systems used in essence serial loops, because the systems vere based around a preamp (e.g. soldano 88, CAE 3+ and Mesa Boogie) into effects into poweramps into two speakers for stereo.
Many people felt these huge racks 'sucked' tone.
My point being, that if Dumble would build Dumbleators today, he would probably chose parallel operation. The serial mode belongs to the 80s or to guitarists with one nice TC2290.
I was looking over the different D'lator schematics again today.
There seems to be some discrepancies about where the local feedback look connects. The 10M-.22uF-10M loop.
On some schematics it shows it connected between the plate and grid which is the usual way to do it.
But on others, it shows the 220K/500pf directly to the grid, then the LFB loop.
Which is the correct way?
I would think that you would want the loop to go between the plate and grid.
That is the way the amps are wired.
But the majority of the Dumbleator schems show the 220K/500pf to the grid and the loop after that.
I was going to wire in the parallel/ series mod when I noticed this.
First impressions is that when in parallel mode the Send/ Drive pot controls the effects volume and is interactive with the Return or Recovery pot.
There in lies another problem when talking about these loops.
There are several names people call the same things.
Such as the Recovery or Return pot and or the output pot.
So I will use the names that are on the official schematic.
That is Drive, Recovery and Return.
The Drive pot and Recovery pot are interactive.
THe problem I am having now is balancing the volume when the delay is on or off in the loop.
This is a EHX Stereo Memory Man in mono.
This delay has a blend knob for wet/dry mix.
So when I was trying it in the loop I maxed it all the way wet to see how well the parallel loop managed the mix.
I don't recall having this issue when the loop was strictly series.
Also when I did the mod I changed how the 220K/500pf was in the circuit.
In the parallel loop this is wired after the local feedback loop.
In the series model it is before the LNFB in relation to the grid.
I may try to swap that back and see if it helps the balance problem.
Anybody else tried this parallel mod and if so what were your results?
I wish a handful of us could get this loop to work good.
It is probably more down to the effects being used than any fault of the loop.
Pedals just don't have all the necessary adjustments to fine tune it to the rig.
If I could afford a TC Electronic 2290 I that may solve the problem.
