In phase or out of phase for jumping channels

[Dagan8]

I have a question about jumping the 2 channels on my amplifier. It is an 18watt clone with a normal and a TMB channel.

I'm trying to understand the whole in phase/out of phase thing.

The normal channel has a single triode gain stage in the pre. The TMB channel has a single triode gain stage followed by a cathode follower stage.

Wouldn't this effectively put the 2 pre amp outputs out of phase with each other since each stage inverts the signal?

I'm assuming that the PI stage in the typical 18watt Marshall design would actually negate this since both channels enter the PI stage at opposite "sides" anyway?

[TransLucid]

A triode gain stage inverts the signal, but a cathode follower does not. So, both of your channels will have the same phase.

[pdf64]

To be precise, the pre-amp channels are in phase until the power amp (assuming that I've understood your arrangement correctly). There, the channels are fed into seperate inputs, which are out of phase with each other.
If you were prepared to lose ~6dB but get them in phase, you could feed them both to one power amp input, mixed via 220k to 470k resistors, and ground the other power amp input. Pete.

[TransLucid]

pdf64 is right, and, looking at the Marshall 18W schematic (which I should have done before the last post) that is exactly what they do - the input channels have the same phase, but they are fed to the + and - inputs of the PI, so they will be mixed out of phase.

Pete's proposed solution is what Fender does - just make sure you AC ground the other input (not DC).

Bob

[Chuck H]

Just to add... If you do mix the two channels via mixing resistors and feed them into the same PI input the PI circuit can be changed to a bypassed inverted stage LTP and likely recover any gain lost through the mixing resistors.

[jimboyogi]

Just to add... If you do mix the two channels via mixing resistors and feed them into the same PI input the PI circuit can be changed to a bypassed inverted stage LTP and likely recover any gain lost through the mixing resistors.

Chuck H, can you please explain what you mean by this?

[pdf64]

Chuck may have been thinking of a cathodyne phase splitter, with a preceding gain stage. Pete.

[Dagan8]

TransLucid

A triode gain stage inverts the signal, but a cathode follower does not. So, both of your channels will have the same phase.

This is what must of confused me. The ValveWizard website I use showed the DC coupled cathode follower inverting the signal. The AC coupled did not. Maybe this is incorrect?

How to design valve guitar amplifiers

[pdf64]

What inverts the DC coupled cathode follower is the preceding triode stage (that it is DC coupled to). The 2 stages are tied together and so it's logical to analyse them as 1 'black box' unit.
Whereas the AC coupled cathode follower is independant and can be analysed as a 'black box' on it's own. Pete.

[Dagan8]

Ok. So now I'm more confused.

Let me break it down so I can get this cleared up.

18Watt "Normal" Channel Pre-amp
- Single triode gain stage inverts the signal

18Watt "TMB" Channel pre-Amp
- Single triode gain stage inverts the signal
- First stage of DC Coupled follower inverts the signal
-Second stage of the DC Coupled follower does not invert the signal

Therefore when the channels leave the Pre-amp section the are now out of phase but they are fed to the + and - of the PI so they are actually mixed in phase.

Right??

[pdf64]

Sorry, I got confused, should have checked a schematic.
'Therefore when the channels leave the Pre-amp section the are now out of phase but they are fed to the + and - of the PI so they are actually mixed in phase.'
You are correct and have got there in the end, despite my 'help'.
Sorry for any confusion - Pete.

[Dagan8]

df64

Sorry, I got confused, should have checked a schematic.
'Therefore when the channels leave the Pre-amp section the are now out of phase but they are fed to the + and - of the PI so they are actually mixed in phase.'
You are correct and have got there in the end, despite my 'help'.
Sorry for any confusion - Pete.

Probably more my fault for not clarifying which 18watt version. Thanks for the help!

[Chuck H]

I was talking about a LTP. But I was mistaken about my theory. I thought that the grid bypass cap on the non inverting stage increased the gain of the circuit. It actually corrects frequency non linearities caused by circuit impedance. My bad.

BUT, you could still increase PI output in other ways to partially make up for signal loss through mixing resistors and run your two channels parallel. The stock PI arrangement has a gain of about 28 (guestimate, depends on Vp). Increasing the plate resistors to 120k(on the inverting side) 130k(on the non inverting side), decreasing the tail to 22k and increasing the grid return resistors to 1M would give a gain of about 33. Any clipping in the PI lost by decreasing drive to the grid might be recouped by intnetionally biasing the PI closer to saturation (maybe 470R instead of 820R). I would also add the bypass cap (.1uf) to the non inverting stage grid (since it's no longer a channel input) just to maximize circuit efficiency.

[Dagan8]

Thanks for the info Chuck. Luckily my channels will mix in phase but if I ever run across an amp that isn't and I decide to use mixing resistors I will know how.