I got yo big ol' Ultra Linear right here.

[LeftyStrat]

Well actually, I just want to spark some discussion about Ultra Linear. I know some think they have no place in a guitar amp, yet here are some successful amps using UL:

- Richie Blackmore's 200 watt Marshall Major.
- Dr Z's Route 66.
- The Selmer Zodiac Twin 30.

In my research one article by Richard Kuehnel suggests that using 20 percent taps instead of the usual 43 percent gives maximum power, as well as:

"... much more pentode and less audiophile triode. Perhaps a better choice for a distortion-salivating stadium amp."

I'm interested in this, but I figure I'd to have a transformer custom wound.

So does anyone have any positive experiences with UL?

[Suzanne]

Just flapping my keys because no one else is

[Reeltarded]

hahaha Suz!

Have you ever played a Major? No breakup to speak of, only clean sustain and you can see the Moon shedding dust.

Say no to UL!

[Blackburn]

Sooze knows tone!

I guess we're not talking about Underwriters Laboratory... They don't play Majors over there.

[matt h]

UL is one of my favorite topics for discussion... so, I'm breaking my vow of silence to comment.

I played around with ghetto-UL on a single-ended build (lowish voltage EL34) about a decade ago. ("ghetto-UL"= my term for using a push/pull transformer single ended and using the CT as the screen tap.) Starting point was an ax84 HO circuit with octal option. Loved it. Distortion generated in the preamp and kept the power amp from caving in under br00talness.

I designed a p1 variant for acoustic-electric use for a friend of mine back when. (He built it for his son.) If memory serves, the amp sounded great. He built a ported cab for it. It featured a 6L6GC at higher voltages for about 15ish watts out. (If memory serves the projected output was 14ish watts. I can't remember if he ever scoped it to confirm.)

I've never played with a "real" UL transformer, or anything in the p/p department. That being said, I think I've come to the same conclusion that many have: UL gets a bad rep in guitar amps because of the tragedies that were 1970s amp design. Namely, UL is a form of NFB, albeit local-ish. Fender fubar'd by running UL *and* throwing in a global nfb ("presence") circuit. Ever over-apply a global nfb loop in a non-UL amp? You know how it sounded turdish? Yup...

I'd wager that half the reason the Rt66 isn't turdish is the lack of global-NFB. The other half being the tube selection, kinkless tetrodes lend themselves well to UL operation. My semi-educated guess there has to do with the lack of frequency control in transformer coupled UL (whereas with a global nfb loop, you throw a tone control on that badboy and call it a "presence"/"resonance")

Anyway, if you remember that UL is nothing more than a form of NFB, it doesn't need to be a scary unknown. A power amp running UL doesn't have to be "completely" sterile. It will, however, be cleaner, the application of nfb will be more evenly spread across the frequency response. This lends itself to a decently "clean" (which is to say "balanced" rather than "devoid of distortion") power amp.

The Rt66 is really just a ripoff of that Leak circuit (tl25 was it?). I mention this because it not sounding turdish is seemingly accidental (albeit happily). That being said, it didn't make the mistake of trying to improve anything by adding a presence/global nfb loop.


I've done some reading about the "12w" class of power tubes (el84, 6v6) being happier with 20% taps and the "25w" class (El34, kt66, 6l6/g/gb/gc) happier with the "standard" 43's. I've heard some theories here and there as to why it is (and if it is). Can't say I necessarily put too much stock in any of them. I suspect more of it has more to do with Pg2 ratings with respect to nominal reflected impedance than it does with, say, straight up plate/screen voltage ratios.

Anyway, UL is great. In moderation. Which is to say, "nfb is great... in moderation... sometimes... when it achieves what you want it to do."

[JazzGuitarGimp]

Hey Matt, I know you've been here for seven years, but WELCOME! . I do have a question: when using a PP OT for a single-ended UL tranny, you wind up with a 50% tap, yes?

Cheers,
Lou

[matt h]

Lou, thanks for the welcome... As for the winding, it's a "50%"ish tap. The "ish" for two reasons... ever seen a CT that was actually centered? me neither!

the second has to do with remembering that the percentage you see the taps listed at is for the "physical" winding position rather than the impedance. So yeah, 20 vs 43 vs 50ish... but we may as well just call them X, Y, and Z or Fred, Bob, and Carl. I know people have reported not being able to hear the difference between 40 and 43% taps, my guess is that "50ish%" is largely in the same category. You could calculate a difference, you'd notice it in the numbers, but you wouldn't so much notice it in the listening.

[Jackie Treehorn]

Demeter tga3 and Alessandro amps.

I like UL quite a bit. My d amp is UL.

[Reeltarded]

UL makes me cry. You guys all like way too clean a breakup. It's very bold by the time the volume is high enough to make up for the missing clipping.

I do have a question and with my level of understanding an answer may even make sense now: what must I do to remove the UL from a Major besides taping back the taps and the other obvious bits? (ieeieio; wire sockets like a plain old 100) The 1kV screen resistors are freaking tremendous, like a barrel full of monkeys.

The only thing sadder than UL is NOL if you feel me, yo.

[Structo]

Me thinks ol' Suz is a spambot.

[Phil_S]

.. ever seen a CT that was actually centered?

I'm not meaning to be argumentative, but I think it's worth discussing this. I think it depends on what you mean by centered -- a question of precision and tolerance to spec.

I am fairly certain that a transformer wound on a machine with an accurate counter is going to be properly centered. (Maybe a one-off hand wound transformer is more difficult to precisely center.) The center is a direct function of the number of turns on each side of the tap. There is going to be an inner winding and an outer winding. Due to the distance around the bobbin, the inner winding will be shorter than the outer winding, hence the ohms reading on each side of the CT will be slightly different. This doesn't mean the CT isn't centered! You can prove this by running a modest amount of AC voltage into the primary and metering the voltage out on each half of the secondary. I don't expect it will be perfectly symmetrical, but just how much of a variance are you willing to allow before you say the CT isn't centered? I'd certainly be willing to give 1% or 2% and call it quits. I imagine the typical 43% tap doesn't stand up to tolerance any better or worse.

[David Root]

Regarding the 200W Major, IIRC Dumble didn't use UL on his 150W 4x6550 amps, and I don't think on the 300W 6x6550 bass amps either (could be wrong on the 300W amps as I don't have any gutshots of those). However I think he copped the Ampeg SVT power end on his 300W amps. But he did use global NFB on all of them.

You can make a case for UL on high powered bass amps I suppose but even Ampeg didn't use UL on the original SVT.

Maybe UL or GNFB but definitely not both in a guitar/bass amp.

The "Ghetto UL" with a SE pentode or tetrode, OTOH, sounds very interesting for a guitar amp! Thanx matth for posting that!

The audio guys would have a conniption because it's all wrong for audio. Using a push pull OT for SE with the CT as a screen tap guarantees distortion from saturating the core when you crank it, because of the DC on the primary. What that might do to the screen supply DC might imply the possibility of screen current overload and/or voltage spiking so the power tube life could be compromised. Don't put an EL37 in there!

PP OTs have a very narrow air gap in order to accomodate minor amounts of DC from output pair imbalance. The airgap on an OT designed for SE use is much wider, so that the core does not saturate. This lowers the power output obtainable so SE OTs have to be a lot bigger than PP OTs.

I'm going to have to look at the Ghetto UL SE some more! Maybe with a KT120, should be able to get 20W or so without blowing it up, and they're much cheaper than a good 845 or 211.

matth what Pd did you run the power tube at?

[matt h]

.. ever seen a CT that was actually centered?

I'd certainly be willing to give 1% or 2% and call it quits. I imagine the typical 43% tap doesn't stand up to tolerance any better or worse.

And that's not being argumentative, that's exactly what I was saying. haha I'm not convinced getting caught up on 40 vs 43% taps matters a whole lot when "center" is still somewhere in a nebulous "off by as much as 1-2%". So then your faux-UL "50%" tap might actually be a 48% tap on your SE... Or your 43% might actually be 45%... This all assumes the machine-counted winder application where all taps are done at the correct number of turns. So there's the turns. What does that mean for DC resistance? What does that mean for AC impedance?

then... what if you're using a PP UL OT? And your 43% is actually 45%... and your CT is actually 48% (on wrt one side). So one of your taps is now greater than the magic 43. And what of the other side? Is it the same proportion from the CT? Is it the same "within 1 or 2% tolerance" of... oh- right- are these things being counted out from the CT to the plate tap? or is the entire winding from plate to plate? You see where this is going...

So yeah, I'm not offended, you're not being argumentative... but it becomes a hair splitting matter when you start looking at tolerances- because 1-2% tolerance for five individual taps on the primary starts to (potentially) add up.

You also mentioned the difference between DC resistance and AC impedance- awesome. Thanks for bringing that up (I realize that sounds potentially sarcastic- it's not... damn internetz). If you think about what the UL taps are doing to the screens, you can see how the differences in DC resistance might end up mattering in a way that never particularly mattered with the plate taps (or mattered much less).

I'll still call CT's "the center tap." But when running a CT as a ghetto-UL screen tap, I'm sticking with my "50ish%" tap... 50/50, 49/51, 48/52. Which end of the primary went to plate? which to B+? That just went and skewed (in one direction or the other) the plate to screen relationship. (again, if we care about such things as precision and getting rid of the -ish-iciousness)

Bringing the Marshall Major into a discussion of UL application is little different from Fender's failed UL experiments of the 70s. The goal was nothing more than "more watts, less dollars." UL does that well. At least on a large production scale.... A regulated screen supply and additional filtering(caps, choke?) VS two new wires out of the OT.

A reimagined Major-ette would rule. Something like a JTM45/bassman pre or maybe even the 2466 Vintage Modern preamp into a 12ax7 PI, no fizz cap but a limited "cut" control labelled "presence" as not to scare people with new controls on marshalls into KT66's running UL, no global nfb loop. That'd be pretty damned glorious.

[matt h]

David, hi- I didn't see your message when I started composing the last. I honestly don't remember what the plate diss. was in my high gain build with the EL34. It wasn't 100%, I think I went for about 90% and lowered it to 85%ish, plate voltage was shy of just shy of 275V. I actually preferred it biased cooler with that preamp.

The acoustic-electric amp (i bet it'd be magic with a hollowbody jazz box) was north of 350V. I think the p1-Extreme was the desired platform/starting point. I want to say we increased the first filter tap by combining the plate and screen taps (maybe minus the power resistor between them). It would've been biased at whatever the (then?) posted 6L6GC bias resistor was suggested for an Extreme variant at the time. It was right at 30W (though that'd be combined plate and screen).

[David Root]

Thanx matt h. Rather than doing the ghetto UL on a KT120, I think I'll look at a pair of 6L6GC in parallel SE, or maybe three of 'em, WTF! Those Hammond 125 PP OTs can be wired to a wide range of primary impedances. I have the 6L6s and I think a H-125 in stock, so that would be cheaper.

For added zing, how about a GA-40 type flavor with a 5879 pentode preamp? I have some of those too.