triode / pentode switch and removing UL in super twin reverb

[wagdog]

Greetings all,

I recently posted this over at the gear page, but there doesn't appear to be much action over there...


I've got a fender super twin reverb (Amp Guide » Fender Super Twin Reverb) I picked up a few months back. I know these aren't the most highly thought of amps in the twin family, but I really like it's clean sound.

I've been modding the amp a bit, moved the input to the reverb to the rail above the 5-band eq, that added a lot of warmness to the reverb. I've also picked up the gain a little in the pre-amp, and disabled the "distortion" triode in the 6C10 - will probably use that triode as a gain stage at some point. Also replaced the electrolytics in the entire amp, PS included.

It's a really loud amp; 180w of ultra-linear fender clean. I wanted to experiment a bit by removing the ultra-linear mode, either going with pentode or triode mode.

I put in a heavy duty DP/DT switch and wired it up.

Triode mode: no problem getting the power from the plates (which is run to the screens through the existing 470ohm resistors (2w btw - I replaced them when I bought the amp).

Pentode mode, I'm current getting the screen power direct from the B+ node that feeds the OT and then sending it through the existing 470ohm screen resistors. In other amps I've seen with pentode mode, there is usually a resistor (or a choke) and another filter cap (node) on the B+ to feed the screens.

Since the screens were previously being fed ultra-linear, they were getting the same voltage as the OT B+ tap (which is 500v, btw - yikes), and are getting that via triode mode as well.

My question is this; in pentode mode, is an additional node in the B+ necessary to feed the screens?

I will say this; there is a nice drop in volume in triode mode, however the amp sounds a little better in pentode mode. AND pentode mode sounds WAY better then it did in ultra-linear - not going back to that. Pentode mode is a bit quieter then ultra-linear.

Maybe some pics would help...

Omitting the switch, here is pentode mode without a node off the Center tap of the OT (B+) would look like:



And here is pentode mode with a node, dropping off some voltage from the B+ to the screen:



And triode mode:



And ultralinear (for completeness, taps are now taped off on my amp):



The question: is the first method ok? I've not seen a stock amp like that, that's why I ask. It seems to me that if it was ok to run ultralinear, and it's ok to run as a triode, that this should be fine as well.

Thanks for any help.

[tubeswell]

To my way of thinking - with the top method, the 470R acts not only as a screen grid resistor, but as an unbypassed screen supply resistor. This means that the gain of the stage will otherwise be less than if you had the resistor bypassed, because of the action of the screen current feedback in varying the screen voltage. If you don't like the effect of the 470R, then try upping it to 1k, 1k5 or even up to 4k7 which would make the attack softer and squish the grid curves together more. But there may not be enough filtering (at the B+ node) for the screens this way

But the next method down will have better filtering (but even better filtering still, if the 1k was replaced with a 3-5H 90mA choke) and more gain AFAICT, because more of the screen supply voltage is being clamped.

(Anyway JM2CW)

[wagdog]

@tubeswell - thanks for the reply.

I was thinking about getting a choke: Filter Choke for Fender 6L6 Amps - looks like it would do the trick.

I'm curious what it will do to the overall sound of the amp. It sounds WAY better in pentode mode then it did in ultralinear - it sounds like a twin now, and dirt pedals actually do something good to the tone.

[tubeswell]

I was thinking about getting a choke: Filter Choke for Fender 6L6 Amps - looks like it would do the trick.

Perfect

A choke provides superior ripple filtering in the power rail (for optimum ripple hum reduction) and (in a CLC Pii filter between the OT primary supply node and the screen node) allows the B+ to run a few volts higher throughout the screens and pre-amp, which in terms of tone, can add to the chimeyness and headroom.

[wagdog]

I ended up going back to the UL taps today to feed the screens.

I found the bottom end of the amp a bit flabby in pentode mode at my gig last night. Perhaps using a choke/filter cap would cure that and I may try it in the near future.

I did utilize mod the distortion switch to lift the NFB, and that, in "UL" mode actually sounds quite good.

Just thought I'd post a followup to this thread in case anyone was interested.

[Rhodesplyr]

I've always wondered why ultralinear was considered such a dirty word in musical instrument amps--as I sit here listening to music played through a modded Dynaco ultralinear amp.

Is pentode output stage sensitivity to speaker loads part of the essential sound of guitar amps?

You may already know this, but tube amps running their output sections in pentode mode are more sensitive to variations in the speaker impedance curve than ultralinear or triode. Speakers often have a big spikes in their impedance curve at their free air resonance point--in the bass region. This might explain the behavior of the bottom end. NFB helps correct for the load sensitivity issue. Triode is the least sensitive to load variations, but also the least efficient--no free lunch :-)

[Steve Conner]

Is pentode output stage sensitivity to speaker loads part of the essential sound of guitar amps?

Yes. What you said agrees with how I understand it, anyway.

Different people prefer different amounts of this "load sensitivity". It doesn't really seem to matter how it's achieved: either a pentode PA with NFB, or a UL PA without any NFB, seem to be about right for most players. (UL is just a kind of NFB, at the end of the day.)

You can think of a presence control as a "treble load sensitivity control", that raises the output impedance of the PA as it boosts treble. Because of voice coil inductance, you will find that the presence control gives a bigger boost with a speaker load than what you measure on the bench with a dummy load.

Some amps have a resonance control too, that allows you to have more or less damping of the cabinet's bass resonance.

Of course AC30 owners would disagree: they rock out to the sound of EL84s in pentode mode with no NFB. But maybe the EL84 has quite a low output impedance even without any NFB.