Eliminating channel bleed through in dual channel preamp

[BiBi]

Hello,

I'm working on a dual channel preamp like the title says. Schematic is included.
The problem I'm having is that the channels are too interactive-- turn one channel up and the distortion comes through the second channel and vice versa.
Is there some way to isolate the channels more completely?
Thanks for any help and suggestions...

[MWJB]

First step would be to rearrange the inputs, Chan 1, Chan 2 & "Both" with the unused channel's input grounded when not used.

Avoid shared grounds & daisy chaining component grounds that might be shared between the channels, ultimately you may not be able to eliminate interaction completely.

[defaced]

The addition of shunting devices (optocuplers, relays, or others) may be needed to shunt the unused channel's signal to ground.

[Wilder Amplification]

I would do a shunting relay from the wiper of your gain pot (center terminal) to ground. With any type of dual channel amp where one channel is an overdrive channel you will have bleed through to the clean channel unless you shunt the output of the earliest gain stage.

On that shunting relay you could use a SPDT relay, ground the common and wire the other two ends of the switch to the wiper of the gain pot on each channel if you wanted to.

[BiBi]

Thanks for all the replies. Actually I realized last night after I wrote this that I forgot to mention that both channels will be heard at the same time. They will use separate speaker cabinets as well with the clean channel going to a solid state power amplifier.

The grounds are separated as much as possible other than sharing the 1M grid leak resistor. I tried to tie the input ground directly to the star ground but this actually increased noise, so this ground currently goes to the cap on the first stage of the dirty channel.

Given the situation about the only thing I can come up with would be to have some sort of transformer isolation? Or, running them out of phase as early as possible.

I certainly appreciate the help and sorry for not clearer earlier.

[Steve Conner]

I forgot to mention that both channels will be heard at the same time.

Well then, the bleed isn't a problem, because it'll be drowned out by the "acoustic bleed" of the dirty channel blasting out of its own speaker.

The reason for bleed is that the dirty channel generates very high signal voltages, with lots of high frequency components due to clipping, that couple very easily through stray capacitance. You may need shielded wires, tube screening cans, etc. Even the capacitance between two tube plates that share the same bottle will be enough to cause bleed.

[mooreamps]

Hello,

I'm working on a dual channel preamp like the title says. Schematic is included.
The problem I'm having is that the channels are too interactive-- turn one channel up and the distortion comes through the second channel and vice versa.
Is there some way to isolate the channels more completely?
Thanks for any help and suggestions...

It is because V4 is configured as a mixer.

-g

[cbarrow7625]

Bibi, just curious; what is that block you have labelled as "Digilog"? What is the intention of that channel with the mixer at the end? It's pretty unique.

[Wilder Amplification]

Bibi, just curious; what is that block you have labelled as "Digilog"? What is the intention of that channel with the mixer at the end? It's pretty unique.

If I were to venture a guess I'd say that's a digital FX unit in a parallel loop config with the 50K pot being the "wet/dry" pot. If that were my loop though I'd want the input to the FX unit to be fixed while controlling how much of the FX unit's OUTPUT gets to enter the mixer's grid but then again I'm anal about gain structure.

I'm also noticing one grid of that mixer is being driven out of phase relative to the other grid, which may/may not be an issue if he's returning a 100% wet signal OR if the FX unit itself returns its output signal 180* out from its input signal.

It is because V4 is configured as a mixer.

-g

Yep that's it...you nailed it.

[BiBi]

Well thanks everyone for the replies...
I think Steve is right about the stray capacitance. It's honestly something I don't really understand well enough. I hadn't thought about the fact that it could happen within the tube itself. Maybe I'll just see how it goes with the bleed-- my concern was when the channels are recorded separately.
In the other question, the Digi-log is a simple digital reverb effect made by Belton. I wasn't such a fan at first, but when I tried the medium length version I found it very attractive. It's really easy to install-- I left out the power section for the unit, but it only needs a 5V supply.
The 50k pot is just an input or dwell control if you like. The wet/dry mix comes at the end of the chain. The reason for an input control is to prevent distorting the effect but also to achieve better signal to noise ratio, or?
Wilder: would you explain why the output is better controlled from a gain structure point of view rather than the input? I would have thought the opposite somehow.

[cbarrow7625]

Thanks. That makes perfect sense. I had never heard of those reverb modules before. Sounds pretty good from their sound clips. I like your application of it; using V4 as a mixer instead of summing passively. I might have to get my hands on one or two of those digilog units to play with.

[Wilder Amplification]

Wilder: would you explain why the output is better controlled from a gain structure point of view rather than the input? I would have thought the opposite somehow.

Most certainly. If you think about how firefighters fiight fires, ever notice how they max out the hydrant (input) while controlling water flow at the end of the hose (output)? This completely fills the hose and make maximum pressure available at the output of the hose where they need it. If they did it the other way around, they wouldn't have very much pressure at all until they turned the hydrant up to a certain point, and even then they may have sufficient pressure, but they may not have sufficient flow rate to fight the fire. It's kind of the same concept.

Most people like to max out the output while controlling things at the input. Well the issue with this is that when you max out the output but then limit the input signal, you're letting all the gear generated noise flow freely on through to the power amp input while not being able to get enough input signal into the unit without adding too much of the effect. In this config, signal/noise ratio is very poor (high noise floor/low input signal level).

Now let's reverse it...let's maximize front end clean signal and control it at the output (i.e. set the input signal as high as you possibly can without distorting the device input).With clean input signal maximized, it is now putting out much more signal than the small amount of noise it generates (all audio gear generates SOME amount of noise). With clean input signal maximized you'll also notice you don't have to run the output control nearly as high to get the desired wet/dry mix, which further reduces the amount of noise from that FX unit that can proceed through to the power amp (i.e. lowers the noise floor). Thus you've just maximized signal/noise ratio.

This is exactly how recording and live sound engineers do it. They maximize clean gain from the mic preamp (input), then control how much of each channel's output signal gets to proceed through to the main mix bus (output), rather than maximize the channel fader then control it at the channel's mic preamp.

Another way to look at it is that when you control things at the input, you're actually controlling how much actual "wet" signal is even available. Whereas if you maximize clean input while controlling the output, the wet signal is already there and you're just controlling how much of that said wet signal gets to be mixed with dry signal.

[mooreamps]

If I were to venture a guess I'd say that's a digital FX unit in a parallel loop config with the 50K pot being the "wet/dry" pot. If that were my loop though I'd want the input to the FX unit to be fixed while controlling how much of the FX unit's OUTPUT gets to enter the mixer's grid but then again I'm anal about gain structure.

The 50K pot would be the dwell control.. The 10K pot is the "wet/dry" pot.

Yep that's it...you nailed it.




-g

[loudthud]

The addition of shunting devices (optocuplers, relays, or others) may be needed to shunt the unused channel's signal to ground.

In other words, mute the unused channel. You might not be able to reduce the signal to zero in the unused channel but with a little work you can get it low enough.

[Steve Conner]

Now let's reverse it...let's maximize front end clean signal and control it at the output

First of all, I think there's a misunderstanding in this thread. Mooreamps says the cause of channel bleed is that one of the tubes is configured as a mixer. That's true, but it's not mixing the dirty channel and the clean channel. It's mixing the reverb return with the dry signal, all in the context of a single channel. The bleed issue is separate.

Now, maxing the input and controlling the output. There are a number of issues here. First, the way it's set up, the volume control will only affect the dry signal. The reverb will always be full volume. If you turn down the 50k pot to play at low levels, you'll be using less bits of the reverb's internal processing, and that makes them sound kind of "digital" and grainy to me. Also, the reverb unit's internal noise will be amplified fully at all times.

That is how classic amps do it, because they have separate volume controls for their channels, no master volume, and one shared reverb tank. The signals have to be mixed before they can be reverbed (or reverberated or whatever ) so the volume controls have to come before the reverb tank.

That works because of the wonderful dynamic range of spring tanks, only limited by hum pickup in the coils and thermal noise in the reverb recovery amp. (Which again, is probably the main source of noise in classic amps.)

But with a digital reverb unit, you really need a level trim before it and another after it, so you can use the dynamic range while playing quietly. I don't know how to do that in the context of a channel-switching non-master volume amp.

Lastly, by using the "anode follower", a tube version of the op-amp summing amplifier, you can get proper summing with a single triode.