Tube theory confusion

[nee]

[quote="gearhead"]Signal from anode and cathode are inverted relative to each other

Ah ha! [Light bulb goes on above head.]

Thanks!

[tubeswell]

Signal from anode and cathode are inverted relative to each other

Read: http://www.300guitars.com/articles/arti ... -inverter/

And in an LTP inverter the second ('followers') grid is A/C tied to ground, and the signal appearing at the 'followers' cathode (which is the same as wot appears on the 'first' stage's cathode) is the same as wot appears on the 'followers' anode (which is the opposite of what appears at the 'first' anode).

(Get a load of that!)

[nee]

And in an LTP inverter the second ('followers') grid is A/C tied to ground, and the signal appearing at the 'followers' cathode (which is the same as wot appears on the 'first' stage's cathode) is the same as wot appears on the 'followers' anode (which is the opposite of what appears at the 'first' anode).

(Get a load of that!)

Aw, now you guys are just making this up!

[nee]

The 300guitars article says re the split-load PI: "As long as the plate and cathode resistors are the same value, the amplitude of the two outputs will be similar, except for the flipped-phase."

In the 5E3 circuit the anode and cathode do indeed both have 56K resistors, but the cathode has an additional 1.5K. What gives? Is this to set the bias?

[tubeswell]

The 1k5 is to bias the grid a couple of volts lower than the cathode. Compared to the 56k, the tube doesn't really 'see' the 1k5 at all.

The cathode voltage should be about 1/4 of the HT voltage at that point in the circuit, and the plate voltage should be about 3/4 of the HT voltage (which makes your cathode about 1/3 of your plate voltage). If you use bigger resistors, you'll get more voltage swing. The cathode swings in unison with the grid and the plate is inverted. See Merlin's site for a fancier explanation.

http://www.freewebs.com/valvewizard/cathodyne.html

[Andy Le Blanc]

The tube flips the phase and it doesn't really matter where the load is placed.
And in this case its divided to get a equal signal. The 1.5k is for bias, you
can bypass it if you want, but the 1 MEG resistor provides feedback so you
wont see much voltage gain. The 1 MEG resistor sets feedback and sets the
input Impedance, many times the value of the resistor, 8 to 10 M, so the
preceding stage sees a high impedance. You can go from 220k to 1M in
practice without running in to issues. The high input impedance means you
can get away with a much smaller coupling cap at the input than you think
.001 or there abouts, but the tube see's the output caps in series so they
need to be x2 larger than you think.

There are a few variations on this type, direct coupling, and check out the
"Paul C Mod", a different way to get bias for the stage.

[Structo]

Aiken has some good stuff about the Long Tail Pair inverter.

http://www.aikenamps.com/

[gearhead]

Signal from anode and cathode are inverted relative to each other

Ah ha! [Light bulb goes on above head.]

Thanks!

Yep, but it also helps to understand -why- it's inverted. The tube itself does not do the inversion; it's the plate resistor that's crucial to that. Taking the signal from between the plate and the plate resistor is what does it.

Personally, I hadn't found ANYTHING that adequately explained the long tailpair. Aiken is okay, but he doesn't explain the mechanisms for WHY it does what it does.

Took a good fellow here, FunkyE9th, to get me on the right track, in this thread:

https://tubeamparchive.com/viewtopic.ph ... sc&start=0

If you want to get geeky and really understand what's going on, the books he recommended are excellent:

http://pentodepress.com/power-amp/contents.html