Tubes 'n transformers

 

Gold Member
Username: Magfan

USA

Post Number: 2645
Registered: Oct-07
Just curious.
Why don't you see more tube amps w/Toroidal Transformers? do they overload differently or simply worse? I can see it for output, perhaps, but power? None with required higher voltage output?

I think I've seen one so equipped, but don't remember if/ when.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 16978
Registered: May-04
.

I don't know a particular reason, but I would guess it has at least a small amount to do with the placement of the transformer relative to the circuitry itself. The main advantage of torroids in consumer audio is their lower levels of EMI radiation. Or, more appropriately, their more compact radiation field as the torroid contains its radiation inward where the shape minimizes noise in neighboring circuitry. Since most tubed power amplifiers have a (step up type) power transformer which sits above the (steel) chassis, the transformer isn't as likely to inject its noise into the circuitry through a radiation field. Whether the average operating Voltages in a tubed amp have any degree of influence on how much a transformer can affect the circuit is a subject I've not seen discussed. But I would assume the effects of transformer induced noise fields will have a more dramatic effect on 50VDC circuits than on 550VDC circuits.

Torroids are sometimes promoted as being more compact than a E core but I doubt that is a much of a selling point in tubed amplifiers due to their inherent bulk. In the high end market more tubed comonents use some form of point to point wiring which is all but impossible in solid state due to the higher degree of complexity required but more conventional transistor circuits. So space is not at a premium in tubed gear as it is in solid state. At that point, all other things being equal, most designers would select the lower cost component with equal performance.


I can't think of any US or British tubed gear which used or uses a torroid. A few Italian manufacturers have used them; Graff and Pathos come to mind; http://www.hifiaudiosound.com/amplifier_graaf_gm20.htm However, when you consider these are Italian amps, simple looks might have influenced the decision making when it came to choosing transformers.




.
 

Gold Member
Username: Hawkbilly

Nova Scotia Canada

Post Number: 1446
Registered: Jul-07
Both Bryston and Blue Circle have used toroidal transformers in their designs. Blue Circle in some of their tube gear....although I'm not sure if they used it in their main tubed components or in their hybrid designs.
 

Gold Member
Username: Magfan

USA

Post Number: 2648
Registered: Oct-07
The reason I asked was it seems a little odd, that toroids seem have less stray emissions than EI or any of the other alphabet soup of conventional transformers. Yet, in both power supply and output, they are favored overwhelmingly in tube designs.
I'm curious, too, if there are any Toroidal autoformers?

Tubes are suseptible to mechanical vibes....leading to microphonics. I'm curious if the same wouldn't be true from a 60hz transformer in the ps. A toroid may limit that radiation? I'm just thinking out loud here, that IF tubes really are more prone to EMR/RFI that a quieter transformer can be nothing but to the good.

It appears that toroids are nearly in 100% of SS while conventional transformers are nearly 100% of tube gear. I'm just curious as to possible reasons.

Even Jan is partially stumped, though informed speculation is good enough for me to get on the scent.

Would anybody care to know if I actually found something substansive?
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 16985
Registered: May-04
.

Sure. Where are you going to find your answer?
 

Gold Member
Username: Magfan

USA

Post Number: 2649
Registered: Oct-07
I saw something about this a (long) while ago. I'll look into it and see if I can find something substansive.

Stay tuned::
 

Gold Member
Username: Magfan

USA

Post Number: 2650
Registered: Oct-07
http://www.plitron.com/standard-toroidal-transformers/tube-audio-transformers/tu be-output-transformers/

Plitron makes output toroids.

Now, maybe the VERY high bandwidth is a problem? Would or could there be a feedback or oscillation problem with a transformer which is good to say.....100khz?
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 16987
Registered: May-04
.

"Plitron makes output toroids."


Not being an engineer I can't see any advantage to an output transformer being a torroid other than being able to build a less tall packing box..



"Now, maybe the VERY high bandwidth is a problem? Would or could there be a feedback or oscillation problem with a transformer which is good to say.....100khz?"


High bandwidth? On the output transformer? Or, on the power transformer? Everyone who cares to can disagree but wide bandwidth has for decades been an advantage in an output transformer. McIntosh and Citation designed and wound their output transformers to have a bandwidth well above 50kHz way back in the 1960's. The arguments they typically ran into would be where the 50kHz signal was originating vs the cost of bulding for an non-existent signal. But having bandwidth above and beneath the 20-20kHz range was considered good engineering by anyone's standards.


Again, not being an engineer, I can't envision how or why a wide bandwidth transformer would or could cause problems with oscillation. On the contrary, such an overbuilt transformer should make the amplifier more stable. Feedback? Of what sort? I don't see where you're headed with this, leo.



.
 

Gold Member
Username: Magfan

USA

Post Number: 2651
Registered: Oct-07
Just asking:
Also, there are a number of DIY folks, which I won't link, who use other than traditional transformers. Some of the hard-core Japanese DIY guys use Toroids.
I believe high bandwidth helps with moving phase shift out of hearing range. Others feel some advantage to bandwidth to as high as 100khz with the claim of evidence of human perception to those frequencies.
Not headed anywhere in particular. So far, at least one of my questions has been answered in the affirmative by Plitron. They exist!

Now, from a consistancy standpoint, I can't see how anyone who makes a (albeit minor) issue of wire insulation can not at least be curious about the obvious physical differences between transformer types. I'm further curious about electrical differences between transformer types. I know there are some measurable diffs.

Jan, if you'd like to help, just ask one of your engineer buddies. Get the abridged version, perhaps.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 16990
Registered: May-04
.

"I believe high bandwidth helps with moving phase shift out of hearing range. Others feel some advantage to bandwidth to as high as 100khz with the claim of evidence of human perception to those frequencies"


I've never seen, or heard, anyone claim humans have even perception out to 100kHz let alone "hearing". The case can be made, however, that extending the bandwidth of an audio amplifier to five times higher and lower than the conventional 20-20kHz hearing range will benefit the amplifier with improved rise time and slew rates within the range of human hearing. So, while we may not hear or even perceive what is happening at 100kHz, our preceptual skills can distinguish betweem two otherwise identical amplifiers when one has truncated the bandwidth to a more typical 20-20kHz.


Improved rise time and slewing rates will equate to a more accurate reproduction of a square wave across a broader frequency range with less overshoot, less rounding off and less ringing. Improved square wave performance can then be said to improve phase performance but there is more to phase than simply increasing the bandwidth as square wave performance can easily be destroyed by excessive NFB application. Since a square wave represents all frequencies applied to the amplifier simultaneously, many feel it is a more stringent test for fidelity than any other "conventional" test bench measurement.

I can't say there is a definitive correlation but when I read about any amplifier that has very good square wave response I also tend to see an amplifier with extremely low IM distortion and well damped upper harmonics. Sam Tellig in Stereophile a few years back interviewed a particular designer who took this concept to a not all that outrageous extreme. Not just insisting the overall amplifier from input to output must have excellent square wave response, he insisted each stage of his amplifier must have its own capacity to pass ideal square waves. The end result was a very transparent, "fast" and clean sounding amplifier according to ST. Measurements confirmed the very low THD and IM but more importantly, IMO, the very low levels of upper harmonic distortions above the third harmonic. As a push/pull design there was an inherent tendency to damp second order harmonic distortion which gave the amp a sonic fingerprint which was described as more tube like that solid state and more solid state than tube like.

Matti Otala was a designer for HK and later Electrocompaniet who worked with the idea of extended bandwidth, high current delivery and, subsequently, accurate square wave performance to establish the concept of TIM - transient induced modulation distortion - in the late 1970's. http://www.blackdahlia.com/html/tip_75.html


I think if you begin to look at the square wave tests of various amplifiers, you'll also see that correlation between bandwidth and distortions. But extended bandwidth has been around for many decades. I don't have the specs for the tubed Citation gear but my solid state Citation 12 from the early '70's had a bandwidth of 4-120kHz +/-0.5dB. My Mac tubed amps from '61are spec'd to a bandwidth of 10-100kCycles, +0dB/ -1dB - quite a feat for a transformer coupled design.



"I can't see how anyone who makes a (albeit minor) issue of wire insulation can not at least be curious about the obvious physical differences between transformer types. I'm further curious about electrical differences between transformer types. I know there are some measurable diffs."



I might be more curious if I were ever to have a go at designing a transformer coupled amp from the ground up. But that's unlikely to happen at this point. What I have contented myself to know are the most successful tubed power amplifier companies in the 1950-60's were those who understood transformers. The best companies wound their own transformers in house to their own specs. Dynaco probably wouldn't have existed as we know them had they not begun as a transformer winding company. Transformer winding is an art which takes the mathematics and physics of transformer design and makes a whole which is more than the sum of its parts.



I no longer have contact with any "engineer buddies", leo. Besides, engineers can only ever tell you what they already know. If you want to know something they don't know, you're probably going to have to find another engineer and then another engineer and then another ...




.
 

New member
Username: Phemenwa

Michigan

Post Number: 1
Registered: Feb-12
Cost
« Previous Thread Next Thread »



Main Forums

Today's Posts

Forum Help

Follow Us