The Zeus I'd seen. Hadn't seen the other one, but there are a few significant differences between these and what I was suggesting:
1. Both use low voltage rails. I'm talking about high voltage, like in a tube amp... say 350 to 450 volts.
2. The second amp is OTL. I'm talking about using regular tube output transformers.
3. Most classic tube designs I've seen do use some amount of global negative feedback.
This is true only by use of bad designed transformers.
The influence of sonic transmission of well designed transformers is much more lower than the influence between e. g. low and high idle current through the output
There are some top class amps (only few SS amplifiers like EAR M100 or Susan Parker's Kits) on the marked that have output transformers inside.
The main source for H2-H3 distortion isn't the transformer - so I think.
Last edited:
The same consideration I have also some time ago.
But a serious problem for me is the heatsink mounting of the MOSFETs - it is very necessary for the isolated voltage special washer assembly (low Rth). Furthermore, the MOSFET is running here in the linear instead switched mode with 50 mA quiescent current at 400 V supply voltage. There are only few high voltage types for linear applications, e. g. from
http://www.ixys.com/
Are there any experience from other guy's here regarded 350-450V solid state amps??
Because the follow brands also sell their output transformers to a wide range of commercial amplifier manufacturers, I guess, the employers there can tell you more about such kind of amplifiers if should be there some models:
1) Lundahl Transformers, audio transformer and tube amplifier transformer manufacturer
2) Welcome at Amplimo BV
3) PLITRON - audio transformers - toroidal transformers - toroids - output transformers - current transformers - power transformers - medical isolation transformers - power toroids
Last edited:
For more of an example here, I have a ca. 1960 Harman Kardon A300 integrated amp... push-pull with 7408 output tubes (similar to 6V6) and 355V applied to the center taps of the output transformers. It would probably be possible to build modules based around something like a Fairchild FQP2N70 to plug in in place of the 7408's. Since it's stereo, I could actually A/B compare the amp with tube outputs to the same amp with MOSFET outputs.
This is what I'm talking about here: not making an amp whose topology looks a little bit like a tube amp, but taking a tube amp design, mostly verbatim, and swapping N-channel FETs for the tubes.
Member
Joined 2009
Paid Member
Maybe you build a tube amplifier and replace the tubes with solid state so you can compare them ?
Solid State Replacement for the 1L6
Solid State Replacement for the 1L6
Interestingly, I think there are some perhaps poorly understood advantages to output transformers, beyond just the convenience of having taps to allow the same power output into different impedances. Just a suspicion on my part based on my experience building a tube amp. I note that McIntosh for a long time built solid state amps with output transformers, I think. Not and expert here.
Cheers,
Bob
They did and still do. The main benefit they quote is that, if your solid state output stage blows up, with an output transformer it won't be trying to feed 80V DC into your woofer's voice coil.
Of course, besides the impedance matching benefit, the output transformers give McIntosh a big advantage in the old "the best amp is the heaviest one" test. ;-)
Of course, besides the impedance matching benefit, the output transformers give McIntosh a big advantage in the old "the best amp is the heaviest one" test. ;-)
Well, it's not as if one can avoid coils and magnetic cores entirely in the playback chain.
Look up the Acoustat X10. It was a direct drive tube amp for electrostatic speakers. 5000V or so output with TV HO tubes for outputs. So, no output transformers, no step-up transformers, no voice coils, and no magnets!
Indeed. I forgot the existence of electrostatic speakers shortly after I heard a pair of Quads in 1982. This X10 is really scary - 5000V bias? Good for Hollywood action scenes where the villain is thrown head-first into the speaker and is instantly electrocuted. Presumably not recommended for bathrooms, kitchens and pool-sides.
Here's an interesting article on audio transformers: http://www.jensen-transformers.com/an/Audio Transformers Chapter.pdf
And here are some from Piltron: PLITRON - audio transformers - toroidal transformers - toroids - output transformers - current transformers - power transformers - medical isolation transformers - power toroids
And here's a DIYaudio thread all about it: http://www.diyaudio.com/forums/tubes-valves/157920-choosing-output-transformer.html#post2034780
I've no experience of audio transformers but it appears that they have plenty of bandwidth (25Hz-100kHz -3dB) and probably less THD than the speaker they'll be driving. Insanely expensive for "audiophile" quality by the look of it.
And here are some from Piltron: PLITRON - audio transformers - toroidal transformers - toroids - output transformers - current transformers - power transformers - medical isolation transformers - power toroids
And here's a DIYaudio thread all about it: http://www.diyaudio.com/forums/tubes-valves/157920-choosing-output-transformer.html#post2034780
I've no experience of audio transformers but it appears that they have plenty of bandwidth (25Hz-100kHz -3dB) and probably less THD than the speaker they'll be driving. Insanely expensive for "audiophile" quality by the look of it.
Last edited:
Member
Joined 2009
Paid Member
If you really like transformers you can use them alone as your amplifier - it's what they did before tubes were invented.
Here's a link to an example of a simple non-audio demonstration - oh and if you really like this link, go see the guys home made triode oscillator.
Homemade Magnetic Amplifiers.
Here's a link to an example of a simple non-audio demonstration - oh and if you really like this link, go see the guys home made triode oscillator.
Homemade Magnetic Amplifiers.
Gareth,
from a wider perspective, some problems may emerge, crawling up latently, intermittently much bigger than previously visualized under the influence of simulated laboratory model at random trip regime, having strategical focus on statistically predicting the past. In a somewhat deeper sense of analogy, it can be nothing else than a veiled virtual framework for generating mythic ruminations on a vaguely discernible formalistic phenomenon, refined to a high degree through superficial perseverance. Likewise, contemplating the general formula of parochial stationary opinion, which pragmatically examines various ostensible computational issues with unerring precision, including the fictional reminiscent characterization of abstract intersubjective field behavior, in a weakly ordered setting. Of course, it not just provides detailed quantitative insight into the classical sequential equidistance segmentation during arrangement exceptions, but also highlights standard difficulties involved, with the conceptual intent to neutralize the tactical influence of hypothetical pseudo-composite samples with a single cut. Now, to question the answers rather than to answer the questions, one being intimately associated with the other, more or less one in the other, we discover that due to irregular amorphous structural affinities in power BJTs, an empirical correction coefficient calibration is symbolistically invaluable. More theoretically, along with the studies that emphasize facilitating as well as integrating the concept of aspect as such, where a specific convulsive load is executed (twice), while contextualizing an autonomous code infiltration of foundational cross-genre representation on a significant scale, despite persistent determination, almost certainly loses its footing at cut off frequencies. In other words, at a certain level of ungoverned intensity, as a result of systematically growing coalescence rates, fluctuating conformers and other participating elements bring marked formation of sporadic turbulence into being, striking the laterally interior portion of the vertically peripheral zone, manifesting themselves as smoke particles, distinguishing core-criteria from academic affectation. Having said that, these legendary indicators reflect a chaotic non-equilibrium state of emergency, widely regarded as a disgrace, yet most indispensable to a truthful appreciation of semi-existent juxtapositional profile values. In the final analysis, given the fact that it is indeed a distinct variant with rather subtle dispersive impact, no effort should be spared to complement the results with additional evidence accumulated in the compliance agreement, strictly enforcing the possibilistic minimum sustainable increment method to closely fit the deconvolution calculations in accord with the error variance.
Let me put it this way:
from a wider perspective, some problems may emerge, crawling up latently, intermittently much bigger than previously visualized under the influence of simulated laboratory model at random trip regime, having strategical focus on statistically predicting the past. In a somewhat deeper sense of analogy, it can be nothing else than a veiled virtual framework for generating mythic ruminations on a vaguely discernible formalistic phenomenon, refined to a high degree through superficial perseverance. Likewise, contemplating the general formula of parochial stationary opinion, which pragmatically examines various ostensible computational issues with unerring precision, including the fictional reminiscent characterization of abstract intersubjective field behavior, in a weakly ordered setting. Of course, it not just provides detailed quantitative insight into the classical sequential equidistance segmentation during arrangement exceptions, but also highlights standard difficulties involved, with the conceptual intent to neutralize the tactical influence of hypothetical pseudo-composite samples with a single cut. Now, to question the answers rather than to answer the questions, one being intimately associated with the other, more or less one in the other, we discover that due to irregular amorphous structural affinities in power BJTs, an empirical correction coefficient calibration is symbolistically invaluable. More theoretically, along with the studies that emphasize facilitating as well as integrating the concept of aspect as such, where a specific convulsive load is executed (twice), while contextualizing an autonomous code infiltration of foundational cross-genre representation on a significant scale, despite persistent determination, almost certainly loses its footing at cut off frequencies. In other words, at a certain level of ungoverned intensity, as a result of systematically growing coalescence rates, fluctuating conformers and other participating elements bring marked formation of sporadic turbulence into being, striking the laterally interior portion of the vertically peripheral zone, manifesting themselves as smoke particles, distinguishing core-criteria from academic affectation. Having said that, these legendary indicators reflect a chaotic non-equilibrium state of emergency, widely regarded as a disgrace, yet most indispensable to a truthful appreciation of semi-existent juxtapositional profile values. In the final analysis, given the fact that it is indeed a distinct variant with rather subtle dispersive impact, no effort should be spared to complement the results with additional evidence accumulated in the compliance agreement, strictly enforcing the possibilistic minimum sustainable increment method to closely fit the deconvolution calculations in accord with the error variance.