How to Bias SF Twin Reverb (135W)?

[johnny99]

Hey guys,
I just finished recapping and retubing a SF Twin Reverb that is the 135W version with the ultra-linear OT. There is no bias adjustment, instead just a balance control.

The schematic is here:

http://ampwares.com/schematics/twin_reverb_sf_135.pdf

Currently the JJ 6L6GC's are at 505V and 25mA, so about 42% idle dissipation. I know for other amps, this would be considered very cold. Should I be trying to change the bias on this then? Looking at the schematic, I'm unsure of what resistor to change. Is it the 2.2K resistor after the diode in the bias voltage section, or the 1.2K resistor right off the bias voltage tap, or something else?

[roberto]

IMHO old Fenders sounds better with cold bias, so don't worry.
Anyway, if you have different tubes, try two different quartets with different currents to see how you prefer it.

[jelle]

It is the 33K resistor off the Bias Balance control. If you lower that resistor, the negative bias voltage -that acts as a screen against bias current- will be reduced, and the bias current will go up.

Jelle

[SilverFox]

Does it need an adjustment? I don't know so I'm just wondering if it would improve things to change it?

Silverfox.

[Stevem]

If you have enough clean volume from the amp than bias wise that is not a issue, my only concern would be that the next filter node and its 20 uf @ 500 volt filter is running darn near right at its limit and in that regard dropping the bias voltage down to get yourself into 60% dissipation range will drop that voltage down to may be 480.
At 60% the outputs will still lead a good long life!

And by the way even though the amp is UL it could still make good use of screen resistors even if only 300 ohm!

[potatofarmer]

First off, calling these things "ultralinear" is like calling an AC30 "Class A." Close enough for guitarists, but technically not true.

Mine was idling noticeably over 500V on the plate with the original caps, so I went with the MEIC (or MIEC?) 600V rated electrolytics for all the filter caps. They've held up well the past year so, even though I'm pretty sure they have to be two caps in series, likely without ballast resistors.

Finding a way to "totem pole" all the filter caps is the best long-term solution, but I don't know how you could fit all those in the doghouse.

To bias hotter, replace the 33k to ground off the balance pot with a trim pot and a limiting resistor, like Jelle said. I just installed a pot temporarily, figured out a resistance that would work, and installed a fixed resistor. (no trim pots handy) Oh yeah, and the bias balance control is necessary with these as it's counting on balanced push-pull through the OT to cancel out the filter hum that winds up on the screens. Alternatively you can install a gigantic choke in front of the plate supply like Dynaco/Sunn did.

Stock, these come with 470R screen stoppers and while they indeed technically aren't necessary to limit screen dissipation due to the distributed load connection, they may be necessary to prevent oscillation. Increasing them is great; I put 2k/5W flameproofs on mine.

[johnny99]

It is the 33K resistor off the Bias Balance control. If you lower that resistor, the negative bias voltage -that acts as a screen against bias current- will be reduced, and the bias current will go up.

Jelle

Thanks for the info guys! In this case, the person really wanted to get as much clean headroom as possible out of this amp so I did move the bias up to 60% idle dissipation and it sounds pretty sweet. I CA glued a 10-turn trimmer onto the back of the balance pot to make it adjustable.

[matt h]

(deleted)

[Stevem]

That would be my question also, why are you calling this amp out to be not a UL output stage?

[potatofarmer]

First off, calling these things "ultralinear" is like calling an AC30 "Class A." Close enough for guitarists, but technically not true.

Ok, I'll bite. UL refers to a specific arrangement wherein the screens are fed by a dedicated pair of taps from the OT primary, typically around 40% but many examples otherwise.

In what way does that not apply to the amp in question?

"Distributed load" operation involves the screens being fed by taps on the OT. You can take those taps at anywhere along the OT, from pentode connection to triode.

"Ultralinear" is a specific case of "distributed load" operation where distortion is at a minimum.

This guy has compiled a bunch of information about it: http://www.oestex.com/tubes/ul.html

Here's a useful chart, this is a plot of output power, output impedance, and distortion for the KT88:

[img:665:706]http://www.oestex.com/tubes/Circuits/kt88b.jpg[/img]

Note:

the relationship between output impedance and ratio of screen tap turns.
the relationship between output power and ratio of screen taps.
the relationship between THD distortion and ratio of screen taps.
the effect of load impedance upon output power.
the power output for ultra-linear operation is greater than that for pentode or triode operation.
This graph shows that for ultra-linear operation the original GEC/MOV KT88 is unsurpassed!!

http://www.oestex.com/tubes/ul.html

The ideal point on that graph for "most linearity" and most power is clearly 40%, though really anything down to 25% would probably work well for our purposes too.

The tapping points on the Fender OT are much much closer to pentode mode. Here's a graph from MEF which bob p added a green box to, indicating Fender's choice of operating point with the 6L6GCs:

[img:797:279]http://music-electronics-forum.com/atta ... d-load.png[/img]
http://music-electronics-forum.com/t35295/

...which makes it even more obvious that Fender chose a point that was really really close to pentode mode. Why? Most likely because you can abuse the idle screen voltage a little more in a distributed load / UL output stage. With pentode the plate can plummet down to the 50V region leaving the screen grid (at close to its idle value) to deal with the lion's share of the current going through the tube. You connect that screen to the OT though, and now it follows the plate voltage. Not perfectly, but it never gets 400+V away from it either.

So you can push the screen voltage up another 50V without the tubes popping. And they saved on the cost of a choke.

...typically around 40% but many examples otherwise.

The "ideal" or "ultralinear" tapping point is different for every tube, though for a lot of tubes it falls at 20-ish% of the primary impedance / 40-ish% of the turns. 6V6GTs it's actually around 20-ish% of the turns, oddly enough.

I don't know, it seems like the only time DL operation is used is when people want UL performance, so it's understandable that when people see screen taps -> must be UL. This is actually a super clever (IMO) case of DL operation but the more I think about it I guess it is kind of pedantic to insist on not calling it UL.

PS - it breaks up like a mountain when you remove the global NFB loop.

[Firestorm]

What a great, informative post!

[matt h]

(deleted)

[xtian]

"Pedantry and pederasty--only one of them is illegal, but both are a pain in the ass."

Ooh, snap. That's funny and bad also.

[Phil_S]

I was afraid this was going over my head, but maybe I've got it now. Is this what it boils down to?

[pdf64]

I think that a distinction between DL and UL is worth making, eg regular UL OTs may not be like for like replacements for a bad Fender DL OT.
It may be seen to be incorrect to call the Fenders UL if they're don't meet the criteria that has come to represent, and as they may not have been intended to be ultra linear in performance.

these come with 470R screen stoppers and while they indeed technically aren't necessary to limit screen dissipation due to the distributed load connection, they may be necessary to prevent oscillation

The screen grids of these Fenders can draw a lot of current when the amp is pushed hard, even via 470 ohm resistors. It would surely be a bad idea to reduce the value?

Increasing them is great; I put 2k/5W flameproofs on mine

Definitely; however, this action seems to be in contradiction to the previous statement? eg why (if they aren't necessary) is it great to increase the value, what's the benefit?