 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
75.168.143.116
In Reply to: RE: perhaps you can enlighten me? You are correct I've not been keeping up... posted by rogerh113 on September 09, 2021 at 17:34:05
The issue here is the OP is using speakers that are bit low in efficiency for use with SETs. I find that if you push an SET hard enough that it sounds 'dynamic' you're pushing it too hard, since that dynamic quality is caused by distortion. This is why you see so many comments about how 'dynamic' an SET is for its power output in reviews. They're simply asking too much power of them.
I suggested that he get an amp of at least 60 watts but to be successful with that kind of power he's going to need an amp that he can push to at least 90% of full power. You can't do that with SETs, not if you want to hear what they are all about.
This particular transformer is a bit curious as its equipped with a feedback winding. Have you heard it with the feedback applied?
Follow Ups:
No Ralph that is called clipping and it sounds unpleasant not dynamic.
Peter Van Willenswaard demonstrated that a SET will generate several times higher voltage on dynamic peaks than its static rating. Waveforms looked not real good but he declared it still sounded audibly clean.
I tend to agree, providing the power
supply can quickly supply the peak's
requirements without causing the amp
to go into clipping or time-delay
distortions.The "extra dynamics" that people note
when pushing a SET with an ordinary
power supply is current sag and LAG.
This causes the speaker woofer cone to misbehave,
causing the effect by pushing and pulling the woofer cone
as an AFTER EFFECT (time-lagged) effect of
the power supply sagging and restoring..This effect is initially started by a musical
low-frequency pulse, but it has departed
from following the music-- it departs from the music--
it's not following the music ON TIME.
The time sequence is LATE and distorted. The dynamic effect
is fake......The effect is eliminated by giving the SE amp low-DCR,
fast attack and quick restore power, with minimal timing
issues and minimal voltage & current sag when pulsed by
a low frequency "hit". The more repetitive these pulses
are, the more the conventional power supply lags behind the
music (in SE operation).. in many aspects, pushing and pulling
the woofer cone, when it shouldn't.Power supplies for SE amps must be low DCR, super fast, and
must be much larger than the amp is calculated to need....,
for instance, a power supply for a one-watt amp should be
the same electrical size as one for a 75 watt push/pull amp.At one audio show, a visitor asked me to take the front grille
off of my speakers so he could watch the woofer cones.He said, "if I see that cone move more than 1/32nd
inch, I'm outta here! And I expect great bass."He later said, "now that amp has a great power
supply".Next year, he asked if that pair of amps was still
around-- he wanted them.The man would go all over the Hotel, listening to music and
looking at woofer cones!-Dennis-
Edits: 10/02/21 10/02/21 10/02/21
-unless we're using different definitions. Since I have an technical/engineering background, I'm used to seeing amps clip on an oscilloscope and so define clipping as that point when a sine wave can no longer be increased in amplitude and so the amp clips off the top bit- that is where the term comes from.
But Stereophile defines clipping as 'anything above 1% THD' which is a different thing altogether!
So I'm not talking about clipping since we're talking about an amp that might only be making 50% of full power (although at that power level its in danger of being overloaded due to the logarithmic nature of the ear).
Yes, only with the feedback applied, per the schematic. I gather Hirata Tango made custom order transformer sets specifically for certain Shishido designs. Apparently the Shishido designs were initially produced commercially in Japan by Fine Arts, and later by Wavac (using ISO Tango or later xformers).
Not questioning your intent. OP asked specifically for SET options that he could try. I agree with you that limited power SET would likely not work acceptably with his speakers. Given the 6db down point at 30 Hz, bottom end of the speakers might be limited anyway. Don't know.
There is potentially a large difference between pure performance specs and musicality. Specs certainly do not assure musicality, and some pretty poor spec equipment can sound great. Listener impressions, like many on this thread, can be very useful (taken with caution).
There is potentially a large difference between pure performance specs and musicality. Specs certainly do not assure musicality, and some pretty poor spec equipment can sound great. Listener impressions, like many on this thread, can be very useful (taken with caution).
IME this is everything to do about the distortion signature. If there isn't a nice 2nd or 3rd harmonic to mask the higher orders, the latter will be audible as harshness and brightness since the ear converts distortion into tonality. The 2nd and 3rd are relatively innocuous so the ear can tolerate a lot of it.
I'm of the opinion that the distortion signature is more important than the amount of distortion. The signature will say how smooth the amp is. If you can get the same signature but at a lower overall level, you can get more detail.
So SETs do fairly well in this regard because of the 2nd harmonic. But you can get a better distortion signature by going fully differential and balanced from input to output. This will cause the predominate harmonic to be the 3rd (which is treated by the ear same as the 2nd) but at about 1/10th the level. Further, because distortion is compounded less from stage to stage throughout the circuit, the overall distortion will be lower and as the order of the harmonic is increased, its amplitude will fall off faster.
The SET has what is called a Quadratic Non-Linearity and the fully differential amp has a Cubic Non-linearity.
IMO/IME the cubic non-linearity is better because its smoother and more detailed. Its also quite a bit lower distortion. For example, our amps make about 0.5%THD with no feedback while an SET does about 10% before clipping. But the SET might make 7 watts, while one of our amps might make 60, 140 or even 200 watts. If you run them at the same power levels the SET can run, the distortion is a lot lower since both amps have the property of distortion dropping to unmeasurable as you decrease the output power.
So that's one thing. Another thing to look for is that the distortion should be the same at 100Hz, 1kHz and 10kHz. A lot of solid state amps (which otherwise have great specs) actually have distortion rising as frequency is increased; their feedback is falling off with frequency due to insufficient gain bandwidth product. This is part of the reason they are bright and harsh (the other being that feedback has suppressed the lower ordered harmonics).
So yeah, the specs don't tell the whole story but we do know how the amp is going to sound based on what we can measure. Its just that the measurements you need to see never show up on the spec sheets, and for that matter what's often considered important quite often isn't; its marketing rather than engineering.
If there isn't a nice 2nd or 3rd harmonic to mask the higher orders
Hey Ralph,
do you have a good reference on this? I have seen it mentioned in passing but have never seen a good paper on the concept.
dave
the article below isn't it but its certainly worth a read, from Nelson Pass.
I'd like pointers to the articles.
It is interesting that Nelson wrote the linked info back in 2013 and since then I hear he has been playing with the phase of the harmonics in successive stages so any 2nd generated remained in phase rather than nulled. Take a simple driver and output with a typical H2 dominant distortion spectra and the nature of the phase inversion puts the even order harmonics from each stage out of phase from the other creating a new overall spectra where the evens partially null leaving the odds to possibly dominate. The story goes Nelson would inset a 1:1 inverting transformer between the stages to keep the evens summing in phase so 2nd dominates and each following order declines.
The above is a rough paraphrase from my fuzzy memory... if anyone knows nelson's current thoughts please chime in.
dave
If you google "preamp variable 2nd harmonic added" you see that people are getting 'tube sound' with plug-ins and the like that add a variable 2nd harmonic.
Nelson made a harmonic generator that did that- he sold kits at the Burning Amp festival. Link below.
I am specifically looking for a citing in support of the following statement:The 2nd and 3rd are relatively innocuous so the ear can tolerate a lot of it.
this seems to suggest they are on equal footing and 1% 2nd is equivalent to 1% third. My experience has been that even order distortions are "friendlier" than odd and as the order increases the level of distortion needs to go down substantially to avoid objectionable sonics.
In respect to Nelson's 2d generator, I am having a hard time understanding what he actually means by the phase of the H2. As near as I can tell he is operating under the belief that there will already be some H2 in the signal and by adding additional H2 from an external source the phase relationships of the two are important.
I dove into spice and built a simple 300B stage that has 1.3% distortion with this spectra.
I then attenuated the signal to make it a unity gain stage and fed it to the grid of a second 300B operating in the same manner. I could then reverse the phase of the two "H2 generators" to illustrate what I believe to be the two conditions (H2+ and H2-) that nelson is referring to and came up with these two spectra.
if we then compare these harmonic profiles to nelsons genral tonal description things become much more (or less??) clear
from listening tests we learn that there is a tendency to
interpret negative phase 2nd as giving a deeper soundstage and improved localization than
otherwise. Positive phase seems to put the instruments and vocals closer and a little more
in-your-face with enhanced detail.It does seem to me that this is not a 100% direct correlation to exactly what nelson is discussing but I cannot envision how the phase of any given distortion spectra is relevant until it is combined with another spectra so you can get sum and difference effects. The biggest 2nd order generator in a system is typically the air speaker driver interface so none of this exists in a vacuum.
dave
Edits: 09/21/21
Take a look- see what you think.
the two things that strike me in the article (and the linked article by Lynn Olson) that jump out at me are this thought referenced to Norman Crowhurst:
Research by Norman Crowhurst points out that feedback mostly reduces the level of the 2nd and 3rd harmonics, leaving the upper harmonics more or less alone, or sometimes at higher level than before feedback. It doesn't take very much level of a 5th, 7th, or 9th harmonic to make the sound "bitter", "dry", "clinical", "cold" "small" "dynamically limited" or "harsh". This is all part of the psychoacoustics of distortion and masking.
This is all in reference to feedback reducing distortion and has nothing to do with masking even though both 2nd and 3rd are mentioned.
Later on in the article he does get into masking but at that point the "masking" generator is a tubed circuit that generates only 2nd order distortion and the "tube noise" raises the noise floor to bury all of the existing harmonics higher than 2nd in the increased noise floor.
both of the points that I extracted from a precursory glance do not seem to support the idea that the third harmonic is benign and helps mask higher orders. He does call 2nd order "sonic gold" but in all of the examples the third is way down and not seemingly a factor in the "sonic flavor" I do like that they give both the measured objective and the sonic subjective results for a number of different music selections which allows the reader to get a better feel for the relationships at hand.
dave
The pattern definitely matters, which is implied here by Crowhurst as the higher orders are largely left alone and the 2nd and 3rd are reduced having a negative impact on the sound.
Cheever also finds that SPL affects the sensitivity to high orders and the ear is more tolerant at higher SPL. This is why having no high order harmonics at lower SPL is critical because our hearing sensitivity to them is the highest.
-
I agree that anything odd other than the 3rd is not perceived well. They become more audible the further they are from the fundamental tone. That is why the 3rd is considered innocuous as the 2nd (and to a lessor degree, the 4th). I read this somewhere in the last 2 weeks but pounding google hasn't resulted in the page I viewed.
If you want to look at an example of where the 3rd shows up, look at any sort of analog tape recording. A properly working tape machine will have a primary distortion amount of the 3rd generated at 0VU which might be 3% at +3VU. Its also a product of output transformers although I've not spent a lot of time on the latter.
A lot of people say that a good reel to reel can't be touched by any other format; it would appear that the 3rd harmonic is relatively innocuous on that basis alone.
It seems that the 2nd and 3rd are somehow helpful to assist the ear/brain system in making out soundstage depth; I was interested to see that Nelson commented on that too. I've noticed using direct mic feeds in the studio if you use an amp that makes significant lower ordered harmonic distortion as opposed to one that does not, you seem to get a more accurate sense of depth out of the monitor system. I don't know why this is but Nelson's comment about this is not the only one I've seen. I'd love to see more research on that but I'm not holding my breath...
There are manufacturers that design to specs and those that design to produce sound. Pretty sure that both can not be the #1 consideration (since there is only one first).
The Tango winders probably came from generations of winders, since skills/jobs were generally passed down. Unlikely that these were trained engineers, but their connection with the product allowed them to create exceptional transformers. Which came first - the transformer or the spec? Meet the spec rather than fully optimize the xformer? There is a reason for exceptional products, and it goes beyond technology.
I consider/buy equipment based on sound rather than spec. Then I can be happy listening to music, rather than being proud that I have an expensive, high spec system that just sounds the way it sounds.
Everyone I knew that was playing around with building OPTs had a garage full of prototypes. And certainly with any transformer-coupled amp the OPT is the heart of its nature and quality.
The OTLs allowed me to get crazy bandwidth. Before I started bandwidth limiting in the voltage amplifier some of my prototypes (this was in the 1970s) could be used as a Radio Frequency booster amplifier- I used one to make 50 watts as a booster for a walkie talkie at 27MHz. But that sort of bandwidth can cause tweeters to get damaged if any RF gets into your system so we never made any production amps with that sort of bandwidth.
But we do make them that have no phase shift at 20Hz or 20KHz. Since I've also produced LPs from some recordings that I've done I have a pretty good idea of what that actually sounds like. I recommend any audiophile do something similar if they can; having a real reference recording on hand really allows you to know what you're hearing.
I used to think that there were things that could not be measured that we are hearing. I don't think that anymore; instead I've come to realize that usually what's happening is that most of what's important that we can measure never gets reported on the spec sheets. To that end, I feel that the distortion signature (the spectrum of harmonics generated) is more important that what the THD is; if you can get the signature right then its OK to get low THD but not the other way 'round!
Phase shift is important in my book. You don't need bandwidth to get it right if you run enough feedback, but if you're not running enough or if you run none then you need bandwidth. You can't hear phase shift at any one frequency but the ear will perceive it as a coloration if it occurs over a spectrum. So an amp that is rolled off at 50KHz can sound dark as a result.
I like what the head engineer at Scott said about this decades ago:
"If it measures good and sounds bad, -- it is bad. If it sounds good and measures bad, -- you've measured the wrong thing."
I would simply add that it helps to also understand how the ear perceives sound and how that relates to engineering. Most spec sheets ignore human hearing perceptual rules.
Make a single ended OTL!
http://aries-cerat.eu/products/amplifiers/collatio-series
I spotted a few false comments on that page.
There are class A OTLs. There are OTLs that are zero feedback. This statement from that page is entirely false:
Feedback is mandatory in this designs to cancel out the artifacts generated by the problems of not true complimentary driving of the signal, and of course cross-over distortion.
Cross-over distortion isn't a thing if the circuit is designed correctly... I could go on but its clear that there is misinformation on that page.
Good perspective. Hopefully one day I will get to hear one of your OTL amps.
OTL has interested me for several years, so I had the objective of renovating a Futterman H3 just to hear it. The right occasion / opportunity never occurred. An Atma-Sphere would likely be a bit more practical. It certainly would provide an interesting and relevant listening experience of a unique approach to music amplification.
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
