Hard Clipping

[ViperDoc]

Here are some better scope shots. Things behaved when I set my probes to AC coupling. Yellow is PI first triode plate, Pink is PI second triode plate.

BRIGHT 100 NORMAL 50 AT THE PI PLATES

5F8A PIP B100 N50.png

BRIGHT 100 NORMAL 100 AT THE PI PLATES

5f8A PIP B100 N100.png

My rudimentary understanding here suggests that the slope of the clip indicates phase shift, and the onset of clipping looks sharp, never smooth like I see on some scope shots. Even if I dial the gain up slightly, the onset of clipping is still sharp. How best to smooth it out?

[ViperDoc]

Sorry about the dual post, I admit, it's a tad annoying. I looked at the JTM45 schematic and there is twice the filtering available for V1 and 2 (16 uF vs. 8 uF on the 5F8A), so I tack soldered another 8 uF cap across the first (two 8 uF in parallel = 16 uF!) and it might not be coming off. Kinda nice, a little tamer. Maybe a touch smoother. I upped the gNFB to a 100K and it's nice, but maybe a bit too high. Anyone ever done "tuned" gNFB? like run more NFB against one side of a 1 nF cap? Like 8 R tap--> (1) 82K---> Presence pot; 8R Tap--> (2) 1 nF cap--> 100K --> Presence Pot?

[Reeltarded]

I swear.. try 33k on hi input's stage grid.. right on the socket. Don't remove anything, just add that one grid stop. You have no shielded wire with an open amp and it's driving you insane.

Try that, play only that channel.. then do it also to the normal channel and seal the amp up without adding and tweaking until the amp is ugly, you are old, and Elon Musk makes a gas powered version with a flame thrower.

[teemuk]

Even if I dial the gain up slightly, the onset of clipping is still sharp. How best to smooth it out?

Get rid of global negative feedback. Drive output tubes with a low Z circuit so they don't "load" and clamp the PI when grid clipping. Different bias on PI? Lower gain overall. Try soft limiting the signal before PI.
Point is, you will never get very "smooth" clipping with this type of power amp circuit. Most would still regard it as an example of revered and archetypal "power amp distortion" of guitar amps. Forget how it looks on scope.

[Reeltarded]

It's simpler though. He isn't getting Tweed Whine.

This amp needs input grid stoppers and a shielded cover when operated. That is all. It will function as an original. Bet.

[ViperDoc]

It's simpler though. He isn't getting Tweed Whine.

This amp needs input grid stoppers and a shielded cover when operated. That is all. It will function as an original. Bet.

So this amp has four 68K grid stoppers, one on each input jack. You would add a socket-mounted 33K on the V1 grids with 2 inches of shielded wire?

[ViperDoc]

Here's what the PI looks like on the scope now after correcting the 100K/82K PI plate resistor misplacement. BETTER!

5F8A BRIGHT 100 NORMAL 50

SDS00023.png

5F8A BRIGHT 100 NORMAL 100

SDS00024.png