Phase Inverter question

[DaN0]

If I've got 2 channels that are out of phase with each other and I wanted to mix them post phase inverter (each channel has it's own inverter), would it be sufficient to ensure that output tube "A" is driven by the mixed signal consisting of the left PI plate from channel A and the right PI plate from channel B? Am I overlooking any issues?

[fperron_kt88]

Not sure of what you are trying to do.

Will there ever be signals flowing at the same time in both channels? From the same guitar? If yes to both, then you will have to deal with the phase relationship. Anything goes from total attenuation of certain freq to doubling-up depending on the settings, filtering and break-up of each channel. YMMV...

Just curious: how are you planning on summing-up the signals at the output tube grids ???

[DaN0]

I was thinking of using 220k mixing resistors into a coupling cap.

Here's the design:
Channel A has 3 gain stages into its own LTPI
Channel B has 2 gain stages into its own LTPI
The outputs of the PI are mixed and then sent to the output tubes

I'm wondering if it would work to do the following (so they're in phase):
Plate 1 of Channel A gets mixed with Plate 2 of Channel B and then goes to output tube 'A'
Plate 2 of Channel A gets mixed with Plate 1 of Channel B and then goes to output tube 'B'

[solderstain]

Not criticising, just questioning: Why are you using two PIs?

[fperron_kt88]

mixing into 2 - 220Ks: never tried it personally, but looking at how the frontend summing works on all these marshalls and bright+normal input configs out there, you should be ok to the onset of cutoff-clipping. Don't know if it would sound good or not when any of either PI's plate would clip though. I would double-check the pi operation taking into account the added loading of mixing resistor + the impedance looking into the other side... especially when it clips... The more I think about it, the less inclined I would be to try it...

I would also run the power tubes open-loop as well (or with cathode-bias), just to avoid any type of stability issues.

inverted phase: my take is that you only need to take care of this if the correlation is strong between the signals flowing into either of the preamp paths (ie. same guitar running to both channels at the same time). You won't get thermonuclear melt-down, either. I bet you *might* hear mild to strong canceling-doubling issues summing them back up. Think fixed-wah (or fixed-phaser) type sounds, if you know what I mean... You might find that either position (swapped or not-swapped) will give different tones as well as it is not quite clear what kind of modified phase relationship you will get from both preamp paths when merging them together just before the output tube grids.


Hope this can help... Hope it does not sound too negative... Still assuming here, but it sounds like the standard approach of using channel switching would yield a more robust design. If you really need clean and dirty at the same time (while having some funk from the PIs on both sounds), I would consider bi-amping way-way before merging the output section in a single amp. Again, just my thoughts...

I would certainly like to know more about the reasons that made you consider this approach...

[rooster]

DaNo - Take a look at an AC30TB schematic. You'll find your answer there, albeit with one phase inverter.

[solderstain]

DaNo - Take a look at an AC30TB schematic. You'll find your answer there, albeit with one phase inverter.

That's exactly what I was thinking. It doesn't even have to be a Top Boost version - it's the Trem channel that gets routed to the 'other' side of the PI, so even an AC30/4 would show how it's done.

Knowing it's been done before, it made me wonder if the OP has some other goal in mind.

[DaN0]

Thanks for all your quick replies!

My design goal is to end up with a flexible amp that can act as a channel switcher or for added texture have the 2 channels on at once. I really dig the idea of mixing channels ala the Komet Connie but in one of the channels the PI breaking up is a big part of the sound so I'd like to not have to compromise and come up with a generic PI for both channels. It seems like it should be possible to mix the channels post PI without too many issues.

I guess with regards to the phasing issue I could always use the extra triode and throw in a cathode follower on the one channel. Hopefully it wouldn't change the tone too much.

fperron_kt88: I've had some bad experienced with mixing preamps that are out of phase and being fed by the same guitar. Sounds aweful especially at band volumes so I'd like to get them to be in phase.

By the way, I got the idea to include 2 PIs from the Orange AD30TC. It has 2 basically identical channels with a PPIMV for each channel.

Is there another way to mix that's better than the 220ks?

[Jana]

The cathode follower stage is non-inverting.

[DaN0]

Thanks Jana... that would have been frustrating to try to figure out why they still weren't in phase. That makes sense though that it's not inverting... oops.

[jjman]

Sounds like a cool way to get clean+clipped sounds with one amp and no pedals. I’m seeing some bands on TV these days with 5 amps for one guitarist.

I agree you should keep the results in phase. I wouldn’t want to think about the NFB wiring ramifications of this. Have you planned that out?

I like PI clipping myself. You’ll need to ensure that the 3-stage’s PI clips on it’s channel-volume control before the 3rd triode clips. Otherwise the strategy of having 2 PIs won’t provide a benefit. (The Trainwreck Express does this PI clipping BTW.)

Do you intend to use a PPIMV on (at least) the 3-stage channel? If not, a clipping PI on the 3 channel side will likely be much louder than the non-clipping PI on the 2-stage side. This assumes you are using the same voltages/resistors on both LTPIs.

Or you could use lower voltages on the 3-stage channel’s PI, but getting the 2 channels “balanced” could take much experimenting there.

Do you have a link to the Orange AD30TC scheme ?

[DaN0]

I really like the idea of texturing the clean sound with an overdrive channel as opposed to changing the foundation of the tone altogether. Kinda like the idea behind a dumble.

That's a great question about the NFB. I was planning on running it back to both PIs but that might have interesting phase issues. Maybe I could run it back to the PI in one case and to the 3rd gain stage on the other channel. Any ideas?

I'm planning on doing a PPIMV for both channels just to make the balancing easier and it might allow me to get a bit of grit out of the clean channel too.