Using both grids for a LTPI - phase questions

[PharmRock]

Alright...another noob question here.  I understand, in general, how LTPI works and how the two outputs are out of phase with each other.  I've built some Plexi-inspired circuits where the two "channels" are mixed before the TS and then that is fed into a single grid on the PI.  I've also built a Marshall 18-watt where each channel feeds its own TS and then each has its own "grid entry" into the PI.
I searched through the forum and came across this thread:
https://el34world.com/Forum/index.php?topic=15954.0
In the first reply, Sluckey posted a link to a schematic in which an EF86 fed one grid of the PI, and a Lightning circuit fed the other grid.  I've attached it below.  If I am reading the schematic correctly, the EF86 channel feeds one grid with an inverted signal, and the Lightning channel feeds the other PI grid with a non-inverted signal.  It seems to me that if these channels were jumpered (either externally or internally), that the signals would cancel since they are out of phase.  However, Sluckey mentions in that same thread that it sounds perfectly fine (Unless I am misunderstanding). 
Can someone help me understand why feeding the grids of the PI with out of phase signals would not cause phase-cancellation?  I'm trying to get my head wrapped around it but am having a difficult time.

Thanks for your help.

[SEL49]

Can someone help me understand why feeding the grids of the PI with out of phase signals would not cause phase-cancellation?  I'm trying to get my head wrapped around it but am having a difficult time.

Sluckey explained it in reply #7.

The LTP is a differential amplifier. The output is the difference between the two inputs. The top triode inverts the signal but the lower triode does not invert the signal. With that in mind, suppose you put the same signal on both inputs. The output would be zero because the equal signals subtract. Now put a +1 signal on the top grid and a -1 signal on the bottom. The output would now be 2.

Probably easier to just view the differential operation as purely mathematical. IOW...

(1) - (1) = 0
(1) - (-1) = 2

[Latole]

Alright...another noob question here.  I understand, in general, how LTPI works and how the two outputs are out of phase with each other.  I've built some Plexi-inspired circuits where the two "channels" are mixed before the TS and then that is fed into a single grid on the PI.  I've also built a Marshall 18-watt where each channel feeds its own TS and then each has its own "grid entry" into the PI.
I searched through the forum and came across this thread:
https://el34world.com/Forum/index.php?topic=15954.0
In the first reply, Sluckey posted a link to a schematic in which an EF86 fed one grid of the PI, and a Lightning circuit fed the other grid.  I've attached it below.  If I am reading the schematic correctly, the EF86 channel feeds one grid with an inverted signal, and the Lightning channel feeds the other PI grid with a non-inverted signal.  It seems to me that if these channels were jumpered (either externally or internally), that the signals would cancel since they are out of phase.  However, Sluckey mentions in that same thread that it sounds perfectly fine (Unless I am misunderstanding). 
Can someone help me understand why feeding the grids of the PI with out of phase signals would not cause phase-cancellation?  I'm trying to get my head wrapped around it but am having a difficult time.

Thanks for your help.

Click to expend

[Merlin]

SEL's explanation is a good one. Personally I am not keen on using both the LTP's grids for mixing, because you can end up with poor AC balance depending on the source resistance feeding each grid.

[PharmRock]

Thanks for the replies. This makes more sense to me now.   

[tubeswell]

Vox did it


ac30_online_02.jpg

[glass54]

Is it my imagination or is the earth (OV) missing on Cathode bias of the EL84 and 2 volume pots in Tubeswells original 1960 schematic 
Regards
Mirek

[SEL49]

Is it my imagination or is the earth (OV) missing on Cathode bias of the EL84 and 2 volume pots in Tubeswells original 1960 schematic 
Regards
Mirek

It's there. Look closely.

[pdf64]

Follow the negative connection to C8 LTP supply node decoupler down.

[glass54]

Thanks PDF. I get it.
Not the way we draw schematics in Protel and Altium, etc or even by hand!
Even my  grandfather used a "Dot" to indicate connections. 
Thus the importance of understanding schematics.
Kind Regards
Mirek

[mresistor]

OK I see it now too..