transformer form factor: toroidal vs bobbin sonic impact: depth dimension 3D

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

klem

Well-known member
Joined
Oct 12, 2005
Messages
172
Location
SW
couldn't find much in searching

i notice with ribbons i particularly enjoy (as i would with any other mic type), those give greater depth or dimension to the source. I don't know how else to describe it, maybe 'realism' or 3D? only in the past year or so as i regained interest in mics did i realize many of my favorites had the toroidal as the sole common factor (vs. various ribbon thickness and length, head basket geometry, etc.) some companies go so far as to explicitly associate a model's depth as being due using a toroidal tx.

so, any thoughts as to what it is about toroidals that helps impart this? i suppose part of the trouble is that depth and dimension are not as easy to quantify as a FR graph, so let's see where the discussion leads...
 
Capacitance can be a problem with toroids, you have to wind "pies" in order to reduce this effect.

Also, using insulation to keep the hi voltage from arcing into the secondary can be a hassle, much easier with a bobbin wound coil.


Then there is the weirdest problem. It is easy to magnetize a toroidal core with unbalanced DC . This will lower the perm and thus inductance of the coil. So your tone could be inconsistent depending on the state of the core. And if the amp gets shut down or turned on at the wrong moment, a voltage spike could magnetize the core. I have never heard a toroid wound transformer that sounded any good. Which is why nobody makes them except Plitron if I recall correctly.

I tried winding a toroid output for a bass amp and it sucked. Don,t know why. Toroid wound transformers are supposed to have low leakage inductance, but apparently this benefit was outweighed by the other problems.
 
I know that several console manufacturers used toroids for line output transformers. The first that comes to mind are the "pumpkins" used on Neve 51 (and maybe VR?) series desks.

https://groupdiy.com/threads/schematic-needed-for-neve-ev10852-output-amp-card.12451/
The pix on this one look more orange in color (hence pumpkin):

https://reverb.com/item/16522784-marinair-t1801-and-vt24499-from-neve-33135-neve-vr
ADM (Audio Design Manufacturing, which made desks for TV stations) used what appear to be toroids and potted in a metal can. I have several that I removed from some ADM switching modules. Examples:

https://reverb.com/item/30591594-adm-audio-output-transformers-lot-of-5-transformers
I think the ADMs are 1:1 but unsure about the pumpkins.

Finally. Amek 9098 desks ("by Rupert Neve the designer") had some toroids beneath a metal cover on the input modules. These had a tertiary feedback winding. Example in attached pic.

So maybe a toroid has some usefulness?

Bri
 

Attachments

  • IMG_0118.JPG
    IMG_0118.JPG
    1.2 MB · Views: 2
I can not say anything about "depth dimension 3D" in this context.

I have never heard a toroid wound transformer that sounded any good. Which is why nobody makes them except Plitron if I recall correctly.
Nobody? The RK series from Haufe is completely designed in toroidal core technology. (Ring kern) This means that the whole radio and television, as well as professional studios in Germany over a long period of time had these things successfully in use. Really countless RK transformer with those excellent results were achieved. By the way, the other company that also manufactures RK transformers in the same village is called Pikatron. Was also often used.
 
I've come across quite a few Toroids with shorted turns or breakdown between primary and secondary ,
Most modern EI power transformers have a split bobbin ,keeping pri-sec well apart electrically ,
Also the layering means the voltage difference between adjacent layers is kept small and where high voltages do exist extra insulation
In toroids you sometimes get points where quite large voltage differences exist with only the wire insulation seperating ,
Ive dismantled a few broken toroids and more often than not you'll find a weak spot where the arching occurred .

Interesting to note the heavy tropicalisation on the Neve toroids ,
some extra data here about their use ,
https://gearspace.com/board/so-much...874085-neve-a4792-air-montserrat-console.html
 
Power transformers are a whole different kettle of kippers compared to something as mundane as a low turns ratio line level output transformer.

I guess I need to do some tests with those ADM transformers I have collected here...ASSuming they are toroids like I suspect they are.

Bri
 
Since Tubetec veered this thread into torodial power transformers, I haven't found them to be less reliable on average.

That being said, I recall a large one which was part of a very large Crest V12 "live" desk that failed....apparently with shorted turn(s). That desk was shipped with two identical PSU units (main and backup) that had the various rails "diode OR'ed" inside the desk.

When troubleshooting, I observed the toroid in the defective PSU made an audible (acoustic) hum with only the primary connected. Fortunately, Crest/Peavey had a replacement in stock.
 

Attachments

  • IMG_9843.JPG
    IMG_9843.JPG
    1,000.6 KB · Views: 0
  • IMG_9845.JPG
    IMG_9845.JPG
    995.8 KB · Views: 0
  • IMG_9846.JPG
    IMG_9846.JPG
    1 MB · Views: 0
Small toroid xfmrs are more expensive than EI types. Claimed advantages for toroid signal xfmrs are essentially extended bandwidth, but EI types are not deficient there. Jensen have output xfmrs with a -3dB BW of 15MegaHz!
OTOH, the saturation characteristics of toroids are not as pleasing as standard cores. The utter absence of gap in toroids results in a very sharp distortion increase when saturation is reached, contrary to other types that have some gap, which results in a smoother transition.
In order to avoid this, toroid xfmrs need to be somewhat overspec'd, which also increases cost.
There is no doubt that for operation within strict limits, toroids provide pristine transparency, which also can be perceived as lack of character.
The use of global NFB around conventional xfmrs makes the discussion moot.
Power transformers is a very different subject. The main advantage is the reduced stray magnetic field, and somewhat reduced bulk. However, too often, off-the-shelf toroids are designed for maximum nominal power, because it's the most immediate point of comparison. Th edrawback is that the core operates close to saturation, which, in case of elevated mains voltage, results in significant magnetic field, distorted, which creates nasty buzz.
Some manufacturers spec their 58mm diameter xfmrs at 30VA. For clean radiated field, they should be rated 15VA: halving the VA results in much lower nasty harmonics radiated.
 
Off the top of my head, expensive (Samar line, Coles 4038) to inexpensive (Bumblebee), and some others in between (OPR, Stager) use toroidals in their ribbon mics, and many of their product descriptions use similar verbiage to what I mentioned in the OP. While I am generally allergic to marketing hyperbole, there does seem to be a consensus among many manufacturers that toroidals provide a superior representation of the source compared to bobbin types. It would be great to drill down on toroidals in a microphone context in this thread. What are the technical characteristics and differences that may lend to the sonic differences?

This also gets at the limitations of a frequency response plot or other specs. It's nice to know whether something is dark, flat, hyped in any particular frequency area, but it says nothing about whether a mic might sound lifeless/dull or have detailed dimension / realism of the source.

I may lack the best way to describe what I'm hearing, and I know there are some naysayers and borderline contrarians here, but I hope this can be discussed in good faith. And a bit of levity ( @NOON ) is never bad!
 
I don’t think empiricism is the same as contrarianism.

Since it’s so easy for expectations to color our perception (even if we have lots of knowledge or experience!), it’s understandable that an empirically-minded community like this one would place greater value on data and evidence over subjective perception alone.

Of course where it gets really interesting is when the empirical data support the subjective impressions. We are fortunate to live in a time where highly-sophisticated measurement gear is available.

If you have a hypothesis about the toroidal form factor having a “sound,” the very best course would be to set about testing that hypothesis (if you lack the equipment or methodology to do so, perhaps you could partner with someone who shares the question and has the chops).

But until the hypothesis is tested (with that one variable isolated), there’s unfortunately not much to discuss (good faith or otherwise) that would find an audience here.
 
I have a masters in epidemiology and biostatistics; I'm pro-empiricism, understand IV/DV and the limitations in determining causality. So although I think I understand where you are coming from, it's a bit of a leap to interpret my earlier post as anti-data driven. Rather, I specifically asked about "technical characteristics" that may yield my, and indeed others, subjective/qualitative impressions. My hope was that, given the vast experience and knowledge-base here, folks more well-versed than me have already explored the subject and could share their impressions. Just like any other healthy message board. If I was looking to start a mic business, sure, I'd do plenty of testing. But that's not my intent. This is more just conversation over a beer type of interest/curiosity. I don't see why anything other than the "very best course" precludes worthiness of discussion here. But maybe I am indeed misjudging the audience here. I hope not!

As for contrarianism and good faith, that had nothing to do with FR or other empirical data. More of a preemptive strike against the replies I notice a few seasoned members here take where, rather than directly answer a question asked in good faith, they are borderline chastising the OP for having the question when they could technically figure it out themselves with our current access to information. As if we all had infinite time and experience. It assumes the person is asking out of laziness, and by and large I don't see that. If they were being lazy, sure, UTFSF/LMGTFY. But they seem to both assume the worst and go out of their way not to help answer the question. It gets old. I could link to threads, but I don't wish to kick up dust, and this thread is having enough trouble as-is staying on topic with the immediate jump to non-microphone toroidals.
 
Okay,

I wonder whether we can find a more-objective way of describing what “depth/dimension/3D” would describe for a mono source?

I ask not to invalidate that subjective impression, but rather because I fear we may all imagine slightly-different things when hearing those words.

It’s sort of like describing a beer as having “bite.” Someone might hear that and imagine lots of carbonation; another might imagine lots of hops flavor. They could wax poetic in conversation, feel like they’re understanding one another, and yet completely talk past one another as they imagine different things.

If we can establish something a bit more objectively-descriptive of the phenomenon you perceive, maybe we could form a hypothesis as to why a toroidal transformer might yield that particular quality.

But in so doing, to attribute the quality to the toroidal nature we’d first have to isolate and rule out the effect of other variables—ones that may be less-conspicuous than the donut shape, but potentially equally-impactful (core material, ratio, etc).

One final question—are we sure the Coles 4038 has a toroidal transformer? I own some and have never heard that before; I’ve also never opened mine up to see. Do you remember where you learned that?
 
I can't remember where I originally read the 4038 having a toroidal tx, but the spec sheet I just found sure makes it sound like there's a toroid in there.
https://coleselectroacoustics.com/wp-content/uploads/2021/12/4038Spec.pdf
As for a definition, I'm unsure how to better put into words what I'm hearing. If you feel that necessarily disqualifies any further discussion on the matter, I think that's throwing the baby out with the bathwater, as well as discounting other useful ways of understanding. [Despite the en-vogue focus on quantitative, empirically derived data (now I'm sounding like the contrarian), qualitative methods do have merit as well, and indeed are the bread and butter of many people's paychecks that are much smarter than the average joe. Hate to digress and give much room for this, but I'm bemused by the somewhat rigid necessary vs sufficient taxonomy in approaching such a topic.] Still, I get your concern about not having much commonality in meaning in such a discussion, one person's 'detailed' can be another's 'bright'.

Putting all that aside, IMHO mono sources, not just stereo, can still have depth and varying degrees of it depending on mic. Obviously there's no width to a source, but I think of when I've heard a ribbon on a single source, and there was... a greater... richness, robustness, detail, vibrancy, life-like qualities than the 421 it was paired with on an amp, adjacent Hardy M2 channels (but I really don't want to chase down IV to the point of cable capacitance details - I'd propose it's more helpful to stay big picture, this post is long enough as is). Unsure why you are zeroing in on just mono though. While I did use the singular 'mic' in an earlier post in reference to specs, I think it's easier to hear some of these qualities if you put a pair of modern 414s vs. 4038s in Blumlein. So yes, apples to oranges in terms of transducer types, but the end result is very, here's that word again, qualitatively different.

Maybe maximizing power transfer has a major role here, same for lack of degradation by magnetic fields (makes me think of the word robust above). And indeed, it would be helpful to change just the tx type in otherwise constant mono and stereo configs, introduce strong magnetic fields with the different txs to hear the impact on the sound. Thanks abbey for the contribution!
 
I’ve got no problem speaking qualitatively—and indeed I spend virtually my whole life in that space. It makes all the sense in the world to deal in qualitative language when the artist wants the tambourine to sound “thick and chunky” or the Rhodes to sound “syrupy.” I can make sense of that in the creation of art.

But the idea that one shape of transformer core can sound more “3D” is a bit harder for me to parse, because that’s the domain of engineering (I’d similarly feel confused if someone said that a suspension bridge felt more “dimensional” to walk across than a cantilever bridge!)

If it makes a difference to the mic’s ability to turn sound pressure into alternating current with fidelity, we have very sophisticated tools that may be able to measure such a distinction and supplement the qualitative with the quantitative.

Some of the reluctance—and I admit to fitting this description myself (after many years of engaging in the very same audio sommelier-style language)—is just a bit of fatigue with discourse that contains little in the way of information, yet manages to inspire the mythologizing of pedantic (and sometimes imaginary!) details (to the point, too frequently, of magical thinking).

In other words—we love audio so much that we get a little carried away with ourselves, and before you know it people are, instead of recording music, fretting that a mic they used to like now isn’t good enough.
 
About the difficulty of finding a common value to the descriptive language: the hen does not care about the taste of their eggs.
As a designer, I never based a design on elusive euphonic qualities, I always started with a set of definite measurable performance targets.
As a performer, SE and producer, I had to navigate between different ways of expressing perception, and finally managed to acquire an understanding of the vernacular, but honestly, qualifying the sound of a single mic as more or less "deep", or "3D" is beyond me.
Maybe the ratio between direct and reverberated sound results in a different sense of space.
I certainly understand the dynamic relationship between a singer and a mic, based on the interaction between the mic's directionality/proximity effect/frequency response and the singer's expression, though.
Back to the subject, I'm utterly unable to point out what in a toroid xfmr makes it depth-prone...:unsure:
 
Back
Top