the big paper v plastic OT debate

[centervolume]

I have been enjoying watching how heated the rhetoric gets on the right way to build an output transformer. Apparently it is a pretty lucrative market to have cornered, or else there wouldn't be too much to say I guess.

Much of the contention seems to revolve around the material used for the bobbin: paper or plastic?

The Mercury website is confusing in that has testimonials from industry professionals who simultaneously say, "musicians prefer the sound of paper" and then "it doesn't matter which is used because both are magnetically neutral" or "transparent".

A friend of mine is a scientific minded guy and likes to keep notebooks of measurements on OTs, reverse engineer them, etc. So his opinion was the first one I sought on the issue, primarily since he had nothing to sell me... was a seeker of truth, had no horse in the race... you get the picture.

his take was that there was an effect on the sound, and it has nothing to do with the electromagnetic properties of the bobbin material, but everything to do with its width. He said that one job of the bobbin is to separate primary from secondary windings, and that the gap of this separation was determined by the thickness of the bobbin material. The plastic used in the sample he reverse engineered being thicker than the paper he found in others (mostly older examples - he called it "fish paper?") In any case, he explained that a closer gap improves the inductance of the transformer, and that inductance was a key property (measured in henries is it?) In any case, the reduced inductance caused by the wider gap of the plastic bobbin he said would not be noticeable until the amp was being pushed to "near wide open" at which time the difference would become audible.

So I continue to wonder about this famous "paper v plastic" debate. Is this some weird science? the ramblings of a madman?? or does it sound plausible? and if so, is the prevalence of plastic bobbins due to increased profit (use cheap materials and then charge expensive prices)?

your 2 cents please

[cminor9]

My opinion? More snake oil.

Transformers used paper bobbins back in the day when manufacturers were building the classics, therefore paper is the only way to go for vintage tone, amirite? Ok, that'll be more money.

I dunno, Can anyone actually tell the difference? I say spend the extra money on tubes and speakers.

[Enzo]

I don't really care either way, but if someone thinks he hears a difference, more power to him. The only way to tell if it matters is to have two IDENTICAL transformers based upon the two materials. ANd that includes wire spacing and separation.

It is easy to look for inconsistencies in a presentation, but looking at what was presented, and thinking about it...

The Mercury website is confusing in that has testimonials from industry professionals who simultaneously say, "musicians prefer the sound of paper" and then "it doesn't matter which is used because both are magnetically neutral" or "transparent".

How about this take: the material doesn't matter because both are magnetically neutral means what it says. As far as the material itself directly affecting the magnetic flux, it is a wash. But:

the gap of this separation was determined by the thickness of the bobbin material. The plastic used in the sample he reverse engineered being thicker than the paper

And there it is, the plastic bobbin affects the spacing. SO when someone prefers paper over plastic, it doesn't necessarily follow that it is because the material is affecting the fields, just that the material is somehow involved. In this case, affecting the spacing.

Your pal says these differences will only show up when the amp is pushed hard. That is not news, many things only really show up when the amp is pushed hard. Power tube distortion ONLY occurs when the amp is overdriving the power tubes. And unless you plan to push the power amp into breakup, the differences in brand of power tube, and current ratings will likely be trivial. Even preamp tube, they mainly show their individuality when in overdrive. At least in my opinion.

Yes, "fish paper" is the term for that thin black cardboardy stuff often used as insulator. Not sure exactly where the term came from, but I have been using it for well over 50 years.

ANy manufacturer wants to reduce costs, and maximize profit. Most of them also want to produce a quality product. Plastic bobbins will be consistent in size and shape, and will be efficient to use in the manufacturing and assemby processes. machine can handle the plastic parts. Fish paper may be marginally cheaper, but will not be as simple to use, I'd think.

Is the debate real or nonsense? Oh I think it is real. That doesn't mean ther are not plenty guys out there who decided on one or the other because they think it is cool, and I am sure there are many who can;t hear the difference, or who hear differences betwen amps and attribute them to the bobbin material even when it is not the factor. But that doesn;t mean no one hears a difference.

[R.G.]

I did an article on this topic in the past. Here's the boiled-down pith of the issue.
1. Transformers are inherently magnetic devices. The first-order effects are all magnetic.
2. Second and third order effects come from the resistive nature of the wires and capacitive effects on the voltages inside them.
3. Transformers are not well understood
(a) by most EEs, because transformer design is a very old sub specialty, and most EEs avoid anything to do with them
(b) by all but the rarest of non-EEs, no matter how well experienced, because they have many complex interactions inside them; if you haven't mastered the theory and the math, no amount of winding and rewinding will help you understand what is going on
(c) by any salesman; no interest, no theory, no practice
(d) by no lay person at all; no interest, no theory, no clue, and lots of wrong information they've been fed by salesmen and true believers of various random stuff. Lay people are often actively MIS-informed on anything to do with transformers, because they are the money donors in the whole process

Given the order of effects, you have to first look at things inside transformers that are magnetic, then things that are conductive, and finally all the rest that are both transparent to magnetic fields and non conductive. That is the order of effects.

For magnetic materials, the core and any gaps overwhelmingly determine where the M-field goes. The number of turns, magnetic core area and material, and the inevitable gaps determine the primary inductance and the low frequency response to a given load. Bobbin material and insulating materials are completely invisible to this process. They may as well be air or vacuum. They are invisible to the M-field.

A much smaller effect is the spacing and placement of the current carrying wires, because they create a magnetic field that the core then sucks in (iron "conducts" a magnetic field nearly 10,000 times as well as free space) and uses in the conversions that happen. The interaction of the core and any gaps and the physical spacing of the wires in the winding allow some magnetic field to leak outside the core. This creates leakage inductance, and this is the overwhelming limitation on high frequency response. It's massively larger than any of the other stuff. Once again, non-conducting materials are completely transparent to the M-field, even to M-field leakage.

Going on to the nit-picking details, there is an effect of the wire insulating material and the interlayer insulation. These materials have a higher dielectric constant than free space, so the self-capacitance of the winding(s) is marginally higher than it would be if the wires could somehow be suspended in a vacuum without touching. This does have an effect on frequency response but is is trivially smaller than the effect of leakage inductance, and very much a secondary issue. In the Golden Age, they used the ratio of primary inductance to leakage inductance as the figure of merit for output transformers, ignoring capacitance and insulation nits entirely. If you pot the transformer in transformer varnish or beeswax as they did at one time, the capacitance increases again because the windings are full of stuff with a higher dielectric constant than air or vacuum. But you really have to work to measure the change this causes overall.

The material of the bobbin, as long as it's nonconductive and does not eat up a noticeably larger part of the window area for wires, would be very difficult to measure with sensitive laboratory instruments. Nylon versus cardboard/fiber/fish paper/etc. is a non-issue, **as long as everything else is the same**. If you use a nylon bobbin that restricts where you put the wires, or you use a nylon bobbin to let you scatter/random wind wires into the winding area, it makes a difference, but the material does not play in this game. It's where the wires are that matters. The magnetic field can't even see the bobbin material.

Similarly, if you are sloppy with stacking laminations, or use cheap iron or thicker laminations to cut cost, or butt-stack the core instead of interleaving, the frequency response changes. The M-field doesn't care if you do these things inside a nylon or fiber bobbin, because it can't "see" the bobbin. But it knows if the bobbin or core stack are sloppy.

IMHO, the whole issue of nylon bobbins versus cardboard/fiber bobbins is a red herring. Nylon bobbins can be used to lower the cost of winding. This does not have to be detrimental to winding quality at all, and can result in more space being available for wire, which has the possibility to make the transformer *better* than you can do with cardboard and fish paper, because you don't have to be so careful in stopping the wind 'way back from the edge of the bobbin. You get more wires in the window. But if you are out to make the cheapest possible transformer, you can (mis)use nylon bobbins to let you be sloppy with winding. And if you're out to be cheap, you probably load cheap, thick iron into the nylon bobbin and do not carefully pack and stack the laminations to make the gap small and consistent.

You'll hear stuff about nylon being thicker and spacing the wires further from the core. To a first order, this is the purest and most refined nonsense. The M-field wants to be inside the iron 10,000 times as much as it wants to be in the air outside it. A few thousandths of an inch further away from the center leg is hardly measurable in the M-field.

None of the cheapnesses show outside the transformer except the nylon bobbin. So someone who does not understand transformers - which is everyone without some specialized training of some kind - says "Hey! This nylon-bobbin transformer sounds worse than this cardboard-bobbin one. That must mean that nylon bobbins are inherently bad!" This reasoning is very human, but is the worst kind of naivety, approaching superstition. What's the song say? "When you believe in things you don't understand..."

Notice that this setup - like all superstitions - is tailor-made for hucksters to sell things. If someone has a pre-existing belief that nylon=bad, then you make more money reinforcing the superstition than educating the buyer. Huckster is probably the wrong word. Many people have come to believe the nylon=bad idea, and they believe it so firmly that they will actually hear nylon as inferior inside their heads if they know what they're listening to, their belief is so strong.

I won't get into the detailed "reasoning" that I've heard over time for nylon versus cardboard. There are too many, and the believers will get offended. Shoot, I'll probably get yelled at for this. But I would have a hard time coming up with a laboratory test where I got to use only a magnetic field and a sense coil to micro-sensitively detect whether the coil was sensing cardboard versus nylon. To M-fields, they are both nearly as clear as vacuum.

[big_teee]

Here's a link to a comparison test by Classic Tone in Chicago.
See what you think.
Paper-Vs.-Plastic
Terry

[Enzo]

I like RG's answer better than mine...

And the Classic TOne shootout looks to be exactly what RG mentioned - they are comparing different transformers, and making the sound distinctions by blaming the bobbin material, at the exclusion of all the other factors.

[big_teee]

I like RG's answer better than mine...

And the Classic TOne shootout looks to be exactly what RG mentioned - they are comparing different transformers, and making the sound distinctions by blaming the bobbin material, at the exclusion of all the other factors.

The Classic Tones still sound good for the buck.
I have that very OT and it sounds great. You can pay a lot more for the Mercury's.
B_T

[cminor9]

The Classic Tones still sound good for the buck.
I have that very OT and it sounds great. You can pay a lot more for the Mercury's.
B_T

Agreed, I like their transformers a lot. I have used them for three builds now. I don't spend the extra $ on paper bobbins.

[R.G.]

It is very much NOT in the best interest of a seller of anything to educate the buying public that some visible factor does not matter. As a seller of -anything- you want the buying public to think that everything matters so you can use any factor at all to discriminate your stuff from the other stuff.

Maybe especially if there is a visible, external factor that has no real effect on the costs of producing the stuff inside the package. Case (literally!) in point: packaging. When you see "New! Improved!" on a package of soap, crackers, canned soup, whatever at the grocery store, do you know what is new and/or improved? There have been legal cases tried and court precedence exists that allow "new" and "improved" and similar advertising to refer to the color of the product *or packaging* to skirt the truth in advertising laws. This is how similar stuff at the grocery store always has "new and improved" on the package, but contains the same old stuff inside.

Advertisers want the buying public to think that everything, no matter how small, matters in the quality of the product. If it does make a better product, that's fine, but if the public can be made to think that purple transformers are better than black ones, that's cool too.

[imaradiostar]

I really like Magnetic Components transformers. It makes me sad to see that they're advertising on the grounds of paper versus plastic bobbins. I have one of their plastic bobbin OT's in my main amp and I think it sounds fabulous!

Kudos to RG for pointing things out so clearly.

jamie

[big_teee]

I really like Magnetic Components transformers. It makes me sad to see that they're advertising on the grounds of paper versus plastic bobbins. I have one of their plastic bobbin OT's in my main amp and I think it sounds fabulous!

Kudos to RG for pointing things out so clearly.

jamie

I thought the Classic tones, were Paper Bobbins?
If you combine the traditon of Paper, Made in the U.S. Plus the Price, they are a great buy.
Just hope they can keep things priced as they are in the future.
Terry

[dai h.]

sort of like the PCB vs. point-to-point wired thing, where the buyer is supposed to draw an inference that one is better (or worse) than the other. I suppose in some highly specific cases it might work (that is differentiating desirability based on one factor), but the reality may be more complex and subjective. There might be some truth (by inference) which helps a layman/novice make a decision but that doesn't help makers of PCB amps made well and transformers using nylon bobbins which are made well (guilty by mis-association). Maybe in the general circumstances, "not enough time + too many decisions (leads to, or tends to lead to) = oversimplification".

[jrfrond]

I'm not a big believer in "snake-oil" claims, but my belief, over years of experience, is that output transformers account for a good part of the saturation mojo in vintage tube amps, and they were all made with fishpaper insulators back them, which affords a closer wind on the core than your typical plastic bobbin. After that, it's power tubes. Ditto for NOS power tubes. I haven't seen one case yet where NOS tubes didn't outshine current Eastern Bloc and (gasp!) Chinese offerings. They go a long way toward recapturing the vintage tone.

However, those paper bobbins, as well as the old-style varnished magnet wire, are also the reason that the transformers eventually fail, as they tend to absorb moisture and break down internally. If amps were still used as they were back then, it might not happen, but turning it up to 10 (or 12, as the case may be) and the resulting core saturation can tax the aged materials, especially if the laminations are rusted together (a byproduct of the humidity factor).

Having said that, I also firmly believe that tube amp tone is the cumulative effect of transformers, plus the chosen component types. Last year, I rebuilt two BRAND-NEW 5E3 reissues for the Broadway show "Spiderman". These reissues were built with carbon film resistors, Illinois caps and Russian tubes, your garden-variety components. They DID have factory Mercury iron though. They sounded pretty good out-of-the-box. However, when I replaced all of the caps and resistors with Sozo and carbon-comps, plus NOS tubes, the amps became different animals, closely matching a vintage 5E3. I stopped and evaluated at each step along the way to qualify each change and the sonic results. So to me, the arguments have basis.

OK, so there's thin paper and thick plastic bobbins. Paper has obvious problems with heat, but are any transformers, custom or otherwise, made with Mylar film insulators? Mylar is fairly heater-resistant and thin, plus it's sturdy. Don't forget, this is what drum heads are made of, which is typically 10mil thick, but as low as 2mil.

[GregS]

However, when I replaced all of the caps and resistors with Sozo and carbon-comps, plus NOS tubes, the amps became different animals, closely matching a vintage 5E3. I stopped and evaluated at each step along the way to qualify each change and the sonic results.

Just curious what were the sonic results at each step (i.e. how much of this sound change was affected by the caps, resistors, tubes).

Greg

[R.G.]

I'm not a big believer in "snake-oil" claims, but my belief, over years of experience, is that output transformers account for a good part of the saturation mojo in vintage tube amps,

An output transformer is a complex, nonlinear electronic filter that also changes characteristics with drive level. There is good reason to believe that it has a big effect on sound.

and they were all made with fishpaper insulators back them,

They were. It was the cheapest way to mass produce them with the available materials and labor.

which affords a closer wind on the core than your typical plastic bobbin.

This can be true in some cases. However, it is almost completely immaterial. It has much the same importance to sound as painting the shells black versus purple. I've read rationales that say that "just like a guitar pickup getting closer to a string, the wire has to be as close to the core as possible." If it were a pickup, where the vast majority of the field is in air, that's true. But it's surrounded by a complete magnetic path of iron. The magnetic field very, very, very strongly prefers to be in the iron, not the air, and will leap over into the iron as much as it can. That "as much as it can" is something for which there are equations that produce answers matching real-world measurements.

Here's another thing to think of: if getting the wire as close to the core as possible, even a few thousandths of an inch making a difference is a big, audible effect, doesn't that also mean that getting the wire as close to the ends of the winding window and the outside of the winding window is a big deal, too? Wouldn't the best transformers be wound with styrofoam spacers in the middle to force all the wire to the ends of the bobbin to get near the iron on the end, and to force the outer windings to be as near the outside of the window as possible?

In fact, the Golden Age transformers used healthy margins - the quit winding a layer well before it was at the end of the bobbin and reversed for the next layer. It was common practice to leave a margin of 1/8" to 1/4" depending on wire sizes. This resulted in a higher yield of good bobbins, which were wound many at a time on a 3' stick of bobbin and later cut apart.

What DOES matter about wire position is the height of the build per section. This causes a change in leakage inductance which is a big factor in high frequency response. This is calculable, and real world tests match the calculations to a good degree.

Really, I promise you - whether the wire is a few thousandths of an inch closer or further from the center leg is not what causes audible differences in output transformers. It's just something that's visible, and easy to blame.

...turning it up to 10 (or 12, as the case may be) and the resulting core saturation

I hear the term "saturation" tossed around a lot. Very few people who say it know what it really does. Iron saturates. That is, all the magnetic regions get aligned at some point, and it can't "absorb" any more field. When this happens, the primary inductance drops dramatically and the coupling from primary to secondary plummets. Very little more gets across.

However, saturation is a voltage times time issue. You saturate a transformer primary by forcing more volts*time into it. You can do this by increasing the voltage or by increasing the time; both work just as well. Conversely, the higher the frequency you put in, the more volts you need to get to saturation. So saturation is a bass phenomena. Where an output transformer saturates for low E, 82Hz, is 1/4 of where it saturates for the high open E string, two octaves up. This stuff is all nonlinear of course, but to a first approximation, the saturation voltage for a signal into a transformer increases directly with frequency. You can put more voltage (and power) through a transformer the higher the frequency you feed it. That's why switching power supplies can be so small, by the way. The frequency they're fed is so high that they don't need to be big.

The bottom line here is that turning it up to 10, 12, or 50 may have a much smaller effect on the output transformer saturating than whether you're playing the open E string a lot. Sure, turning it down helps; but it only matters if the guy who designed the transformer (if he had a clue about transformer design; there aren't many of those) cheaped out and didn't put in enough iron so the full power supply voltage at low E - or drop-D or whatever would saturate the iron with volt-seconds. If they properly estimated the lowest incoming frequency and the volts available from the power supply ==> you can't saturate the output transformer by turning it up <==. If they did a poor job, or didn't know what they were doing, all bets are off, of course. And if the boss's newly hired MBA nephew who's being groomed for taking the place over, gets to dink with the design, you're pretty sure to get economies in the wrong places.

The thing about all the mythos and folk knowledge about tube amps is that it's all focused on what someone can see. It's this line of reasoning: "We know these pink transformers sound bad, therefore, if we paint any transformer pink, the sound gets worse, even if it used to be good." There are real issues inside transformers that make them complicated to deal with, but almost none of them are visible from the outside.

But knowing that has never stopped humans from creating a belief system.