When is it okay to DC couple?

[AZJimC]

I've been looking at the split load phase inverter, and noted that some schematics show a coupling cap between the gain triode and splitter triode, while others do not. Why?

Also, on the ones that are dc coupled there is no biasing resistor between the load resistor and cathode. I think I understand why that is, but is one preferable?

I also see this on gain stages sometimes, sometimes not?

[HotBluePlates]

I've been looking at the split load phase inverter, and noted that some schematics show a coupling cap between the gain triode and splitter triode, while others do not. Why?

Also, on the ones that are dc coupled there is no biasing resistor between the load resistor and cathode.

Bias is bias.

Tubes with a cathode bias resistor use the current through that resistor to define the voltage between tube grid to tube cathode, which is the tube's bias. The tube's cathode is positive of the grid, or the grid is negative of the cathode (2 ways to say the same thing).

The current through a cathode load resistor creates a voltage drop, or a positive voltage on the cathode relative to ground. The preceding stage has a voltage on its plate. If that plate were directly connected to the grid of the following stage, it would define that grid's voltage relative to ground.

If that grid voltage is negative of the cathode voltage, and by enough difference to result in the tube current which would create the original cathode voltage, then the cathode-loaded stage can be direct coupled.  Which is to say the difference between the planned cathode voltage (due to tube current & load resistance) and the plate voltage of the previous stage is the right proportion so that the tube feels the same bias as would have been created if a cathode bias resistor and coupling cap had been used. This arrangement then saves 2 parts.

[tubeswell]

Also important to remember that the h-k voltage of the following stage needs to be within spec, and the higher the plate voltage in the first stage, the higher the h-k voltage of the 2nd stage will need to be (unless you knock the grid voltage of the 2nd stage down - with a voltage divider, but then you have to watch the HF rolloff in the 2nd stage). Also the higher the plate voltage in the 1st stage, the less 'headroom' there will be for getting gain out of the 2nd stage (assuming the B+ supply is the same for both stages).

[PRR]

> split load phase inverter, and noted that some schematics show a coupling cap between the gain triode and splitter triode, while others do not. Why?

I've wondered that for decades.

No-cap nominally has one less low-cut pole to twist bass phase and reduce possible NFB. However by bootstrap action this pole can conveniently be placed about as low as you might ever want. (Also guitar amps don't use a ton of NFB.)

No-cap puts a *severe* design restriction on the stage before it: plate must be roughly 1/4 of supply voltage. We more often aim for plates to be 1/2 to 2/3 of supply voltage. It can be hard to get it way down low; harder to ensure it stays there with tube and resistor variation.

With plate way down at 1/4, it can be hard to get a BIG swing. Generally need a low-Mu tube at low current and huge plate resistor.

No-cap overloads one way, with-cap overloads another way. I'm not at all sure which is better, for what, or with/without heavy NFB.

With-cap is simple re-use of standard design tables. No-cap requires original design and bench verification.

As always, plagiarize designs known to work good for your purposes. Life is too short to understand everything.

[AZJimC]

"As always, plagiarize designs known to work good for your purposes. Life is too short to understand everything."

Yep, I know that to be true, but, that's discounting the OCD effects. The list of things I don't know grows exponentially daily. Everything I learn just exposes more questions. Keeps me awake nights often :) It would seem a guy would have more sense, it's a loosing battle really.

I agree with plagiarism. It keeps things a known quantity. I have a couple of amps dedicated, or cursed, to my plagiarized section replacement. The other day I was thinking about one of these, and how I have a year or so of on and off labor in it, it's crazy. But, OTH, I've learned a great deal, owing no small part to this forum. Just think, IF I ever get it DONE, how great it will sound,