Output transformer for SVT-CL that doesn't limit low frequencies

[teemuk]

I think in many cases people are hearing something else than actual transformer saturation (e.g. tubes clipping, crossover distortion, dynamic modulation of bias points, non-flat frequency response, etc.), but simply think they are hearing transformer saturation because they have heard that catchphrase so often.

"I'm hearing something, must be that OT saturation I've always heard about..."

Kinda like the usual phrase about power tubes softly clipping and producing "musical" even order harmonics. Totally debunked by plugging amp to scope and then realizing it hard clips (with plenty of crossover distortion) and produces mainly odd order harmonic distortion.

It is easy to hear -something-. It is much harder to tell what that something actually is. I would think OT saturation is not that common because it would totally ruin a usable tone.

 

[JimmyM]

That's a possibility. But I have turned up amps all the way and totally dug it, so I'm thinking at that point, the transformer has to be getting juiced pretty heavily.

 

[teemuk]

I used to think that clean reproduction of the low B would be a good thing. But it is costly and involves compromises in size, weight and power. It is also not needed to hear the low B clearly. It is not just the harmonics of the low B that are being heard - rather it is also the intermodulation products of those harmonics. A non-linear mix of the 2nd and 3rd harmonics of the low B produces among other artifacts - a low B. Low B fund ~31 Hz. Low B 2nd ~62 Hz. Low B 3rd ~93 Hz. 93 - 62 = 31 (a low B!).

Psychoacoustics and "virtual fundamental". Our hearing process makes up the missing fundamental because we deduce its existence from other harmonics.

...Not to mention, in many cases the "missing" frequencies are patched by harmonic tones created by distortion.

The point is, since fundamental is not produced in -full amplitude- the chances of OT saturation, or its effects, are much smaller and intermodulation distortion caused by it has less effect to "mess" up the sound on every other frequency as well.

 

[Bassmec]

Well if you where thinking of extending available VLF and improving loudspeaker control in an SVT via the output transformer?.
Taking a leaf out of Fender and Marshall Simms Watts and others book, It would be perfectly possible to fit a five wire primary, Ultra Linear output transformer.
The modifications needed to convert the valve base UL wiring changes would not be impossible.
I would copy the Simms Watt 200 design for 4XKT88 and Just add another pair.
Fender tried doing this with 6L6's in the Bassman 135 but for some reason it's tone on overdrive suffered more than
amps using KT88's in UL like the Simms Watts AP 200 and the Marshall Major.
Mr Foxen used to have a Mattamp 300 slave. that would be a great 6XKT88 UL circuit to clone.

 

[Korladis]

I would agree with you, except for the butt-ton of bass players out there who like saturation. Some use cranked tube amps, some use SS amps and pedals designed to sound like cranked tube amps, but I hear distortion on bass all over the place. Might not be what you like, but lots of us do like it and don't judge it by what it looks like in a screen grab of a waveform.

Yep.

I play through a middling power tube amp at very loud volumes, setting the amp up for a lot of overdrive, pretty much all the time. With one band, every rehearsal and every show. So I'm sure my low B probably looks like that down at the fundamental if you graph it. But the sound is definitely not unpleasant, even though it is quite distorted. I mean, I'm sure some people wouldn't want their own tone to be like that, and would prefer cleaner sounds, but that's fine, too.

 

[Korladis]

That's a possibility. But I have turned up amps all the way and totally dug it, so I'm thinking at that point, the transformer has to be getting juiced pretty heavily.

Exactly. My favorite tone from my amp is with it turned all the way up. I don't get to do it that way very often, even in my doom metal band, though.

 

[Martin Beer]

With the dynamic nature of a bass signal (compared to using a function generator for testing), is it possible that the "nasty" distortion from transformer saturation could be more listenable than we might expect if it only happens briefly on the initial peak of the note? It might be interesting to view the output of some popular amps on a scope using an actual bass signal to figure out where the different types of clipping we hear from a cranked amp are coming from.

 

[nashvillebill]

Teemuk has done a ton of research on guitar amp distortion, I've read his work on other sites. Merlin Blencowe also has addressed the various forms of distortion that come from guitar (and bass) tube amps, in his classic book "Designing Preamps for Guitar and Bass".

Folks throw out a single word: "distortion" or "saturation" as if it's one single phenomenon, yet within a tube amp there are several ways distortion occurs. Within the preamp, the first gain stage or two: there's a couple of ways clipping can occur there, and they can sound different. Then there's interaction with the EQ, then perhaps more gain stages--then the signal hits the phase inverter. Different PI's distort differently, in different manners. Then there's power tube clipping...and then OT saturation. The amp design determines how the amp will go into distortion, or perhaps more specifically, where in the topology the amp clips. Even when dimed, the preamp section may be doing all of the clipping, with the PI, power tubes and OT still operating merrily within linear (unsaturated) limits. Or the PI may drive into clipping first, though the preamp and power section may still be clean (kind of an unlikely scenario, IMO). Or the preamp and PI still clean, but the power tubes are driven hard enough to clip them. These clipping/distortion/saturation effects, in those sections of the amp, are most likely to be occurring, rather than saturating the OT--even at full volume ("dimed")-- in a decently designed amp.

The scope waveform study is indeed very useful, contrary to some folks' opinion, because it can show exactly where and how distortion occurs as a signal goes through an amp. We typically use a sine wave because random bursts of signal from a bass don't give us a good reference point. And contrary to some folks' beliefs, sine waves are actually a quite difficult waveform for an amp to reproduce perfectly, so they work great to study distortion effects.

 

[agedhorse]

Teemuk and Nashvillebill both grasp both the complexity of this subject as well as the ease at which folks tend to attribute reasons to mechanisms that are not at all accurate.

This is in large part what much of the marketing in the industry depends on. The less a player actually knows, the easier it is to fabricate BS that then becomes internet truths.

 

[Martin Beer]

The scope waveform study is indeed very useful, contrary to some folks' opinion, because it can show exactly where and how distortion occurs as a signal goes through an amp. We typically use a sine wave because random bursts of signal from a bass don't give us a good reference point. And contrary to some folks' beliefs, sine waves are actually a quite difficult waveform for an amp to reproduce perfectly, so they work great to study distortion effects.

When I commented about whether looking at the response to a bass signal might be interesting, I intended it more to express curiosity (and invite elaboration) about how the behaviour observed under test conditions translates to what we hear in use. Of course, I would not dismiss the usefulness of testing using sine waves.
So while I'm quite happy to accept that the pleasing colouration we hear from valve amps is likely to be from several other sources than transformer saturation, I just have some curiousity about whether saturation always sounds bad or whether there is some level of it which might be acceptable to the listener when it occurs.

 

[Chris 'Wighat' Jordan]

What great dialogue! By thanking you all now (thank you all) I am certainly not suggesting that this discourse has concluded.

 

[nashvillebill]

I wasn't directing my waveform comment at you Martin, but rather at some of the Internet BS in audiophile land.

Teemuk would be far better at addressing the "musicality" of OT saturation than I, but I suspect it's rather like blocking distortion--when a capacitor-coupled tube reaches grid current limiting; not pleasant at all.

Certainly some amps are/were designed (or merely just built that way) so that their power amps will go into clipping--think the Deluxe 5E3 in "Cinnamon Girl". But it's my suspicion now that for larger, newer amps, manufacturers have tended to keep the power amp section clean and let all of the overdrive and distortion come from the preamp and PI. This has benefits: the power amp section can be designed with a better grasp of its requirements, and the amp's distortion is less affected by volume; more controlled, more predictable if you will. Neil Young purportedly has hundreds of 5E3's yet only one of them has "the" sound! Now though, think of something like a Peavey 5150 and its variants (XXX, 6505, etc) which have something like 5 gain stages in the OD channel! They are massaging the signal in the preamp yet keeping the power amp fairly clean, to the best of my somewhat limited knowledge. The power amp sections now are "known quantities", you can get up to 75 watts of clean power out of a pair of 6L6's! (the Musicman amps ran outrageous plate voltages. Peavey gets 60+ watts clean out of a pair of 6L6's with a more conservative 530 volts on the plates)

So now the amp designers can spec out a power section fairly reasonably, keeping the overall cost down, by merely keeping the power amp running clean and letting the preamp and PI handle the distortion/overdrive duties. They know a pair of 6550's can get 100 watts of clean power (2.5% THD) with 600 volts on the plates, into a 5K load, with 62 volts AF g1-to-g1 voltage (I'm taking these numbers straight from the RCA receiving tube manual). Now the amp designers can choose a desired frequency range and tell the transformer folks what they need for a transformer: 5K load, no more than -3dB down at 50Hz, etc etc, and the amp designers will set up the amp, even with a post-PI MV, so it delivers about 62 volts signal g1 to g1 maximum, and they'll design the circuits in the amp so that frequencies below the desired floor aren't going to the power amp. The distortion then is designed into the earlier stages of the amp.

The end result: folks hear the amp starting to "sing" at higher volumes, and they immediately assume it's the power amp section/OT saturating, but in reality it's now the earlier sections of the amp reaching their "happy place" as they start to clip and distort as intended and as the amp's power supply voltages reach their designed conditions.

 

[JimmyM]

OK, then I have to ask how come turning a tube amp up all the way invariably sounds better and fuller than turning the gain up all the way but not the master?

 

[ThisBass]

OK, then I have to ask how come turning a tube amp up all the way invariably sounds better and fuller than turning the gain up all the way but not the master?

good point.
I'm pretty sure that's a basic problem at any amp modellings. Sometimes time invariant properties turn into time variant behaviour at certain conditions like overload range. To understand all of the essential mechanism is engineering at it's best.
Although identical preamp and power amp should always sound the same, the sound is different if preamp and power amp come within a head instead of separated in pre amp and power amp units.

At the other hand every tube amp will emphasize impedance peak of the cab at the low band (more or less). I claim impdance peak of the cab is less audible at low volume, but indeed noticable at overload range where overload isn't quite audible as clipping.

 

[cfsporn]

It is threads like this that convinced me to end my long academic sabbatical and switch my major from mechanical engineering to electrical engineering.

I'll be watching from the sidelines, slightly confused but riveted and in intrigued.

 

[agedhorse]

OK, then I have to ask how come turning a tube amp up all the way invariably sounds better and fuller than turning the gain up all the way but not the master?

Good question Jimmy.

When you turn an amp's master up all the way, there are many mechanisms taking place simultaneously. These include such factors as phase inverter o/d, driver o/d, power tube o/d, any preamp stages involved with the master control o/d, any intentional dynamics management and power supply effects, and any tonal interjections that might be designed into the physical position of the control.

There's no way to know which mechanism(s) are responsible for what you are hearing without working with each mechanism (and various combinations) in isolation.

 

[Mr. Foxen]

Good question Jimmy.

When you turn an amp's master up all the way, there are many mechanisms taking place simultaneously. These include such factors as phase inverter o/d, driver o/d, power tube o/d, any preamp stages involved with the master control o/d, any intentional dynamics management and power supply effects, and any tonal interjections that might be designed into the physical position of the control.

There's no way to know which mechanism(s) are responsible for what you are hearing without working with each mechanism (and various combinations) in isolation.

Also speaker breakup, Fletcher-Munson curve. Also some master volumes are kind of badly implemented and mess with the sound when not fully open, usually post phase inverter ones.

 

[JimmyM]

Good question Jimmy.

When you turn an amp's master up all the way, there are many mechanisms taking place simultaneously. These include such factors as phase inverter o/d, driver o/d, power tube o/d, any preamp stages involved with the master control o/d, any intentional dynamics management and power supply effects, and any tonal interjections that might be designed into the physical position of the control.

There's no way to know which mechanism(s) are responsible for what you are hearing without working with each mechanism (and various combinations) in isolation.

Makes sense. So I guess we'll need to address specific amps specifically...what do we know about the SVT and its possibilities for tranny saturation? And while we're at it, how about the flagship tube amps from Mesa and Fender as well?

 

[Mr. Foxen]

Stuff with big giant transformers is going to be less susceptible to it. Grade of the iron in it and some construction probably has an effect, but bunch of it is mass.

 

[Korladis]

So my D-180, probably more susceptible to transformer saturation than an SVT?