Don't worry that is just a symbolic price (for one speaker only!
) for something unique and it is certainly not for sale! And after trying to explain once how it works I will never try again just because of people like you and the transformer guy!! It's just to say that you cannot make your point just opening the mouth and questioning the abilities and means of the others that you don't know at all. You could hurt yourself a lot, as in this case. Envy is an ugly beast!I don't think I am missing anything. EI core thin laminations have low losses, nickel EI laminations have very low losses. You really don't know what you are talking about.
You can only dream my reference system, from the source to the room.
You can see the speaker in my avatar. If you have 200000 euros you can buy it!
As long as you can't give any datasheet of how low the losses are you talking crap i think.
Anyway a U808 has very high losses almost 1dB for a 5kOhm tap.
Look what happens if you change a good c-core for a better HiB core (same coils) :
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
I had started to think like this until the Audio Note HE-6 kit was alerted. And it seems to me that if AN reckon it worth preparing an expensive 9W SE amp with very costly 50-ohm output transformers, then it should be worth developing something DIY that works even better. Elsewhere on our tubes/valves forum we have shown that building DHT line amplifiers with #26s can now be done by anyone, and that any DIYer can get them quiet AND excellent-sounding - a real step forward.
In the same way, we can do an HE-6 amp without chickening out and using PSE EL84s - which will certainly be a compromise unworthy of expensive iron.
The fact that our DHT design will also drive speakers is an advantage, not a burden.
The solution for anyone content with 500mW output was shown hundreds of posts ago - just take a 4P1L, a catalogue transformer from Sowter (or whoever is closest to you), and build. If your interest stops there, fine - just unsubscribe.
Meanwhile, the fact of the smaller audience for a more extreme design is not enough to put me off. One reason: the market for this grade of headphones seems healthy enough. The other: if we can satisfy the highest calibre headphone demands at 5-8W power level using better driver circuits and lower noise power supplies, I think we will also have also developed a very useful speaker amplifier.
I think a summary of the discussion may be in order for those like yourself looking for practical info.
Probably you should read up on high gm IDHT spud headphone amps, they will power 95% of all headphones . Its relatively simple though and you can read up on a fairly good version at ecdesigns or dsvatisks website, they are perfect for HD800's. More linear than any two stage DHT amp. With the right transformer and B+ they sound better than anything you can buy commercially.
Don't build a DHT headphone amp if you only own Grados, Senns, Denons, etc these are efficient phones that need just a few milliwatts!
The only reason really to move to a DHT headphone amp are the hard to drive inefficient phones, not just the HE6 but the HE-500, HE5LE, LCD2, LCD3, AKG1k. This is why there are so few DIY DHT headphone amplifier builds, smart DIYers built the high gm idht spud's.
If you want a low cost DHT headphone amp just have your local custom transformer guy wind an M6 EI 5k primary with four 8 ohm secondaries, if you only need a watt or so use the 4P1L as an output tube. If you want more power and better linearity then pick the 300B. Use coleman filaments, ss B+ regulation, and separate chassis for the power txfmers/rectifiers.
Things get complicated and arguments start when we want a transformer that can handle the new planar phones and retain the finesse to handle low level detail for the sensitive phones. Its difficult and unchartered technical territory, not audiophile nonsense.
Sorry I haven't started my build yet but I want to get the transformer right first, and I am finishing up my DAC. I generally plan for several months before I build an amp this is a hobby not a race for me.
Edit: What Rod said too!
Last edited:
First, you can use that tone with your mother, not with me!
Second, I understand your frustration but copy & paste is not enough. A generic "normal C-core" can easily have more losses than best grade EI M6. Still waiting for some REAL thing.
Third, I have never said that M6 should be the choice for headphone amp. EI cores exist in better grades (i.e. lower loss), including nickel. The U-808 example was just to show you were wrong about the hypothetical consequences of lower leakage flux of C cores. YOUR example of "super" transformer demonstrates I am right.
And now I am really out of this thread. I have insisted on not ruling out EI for a reason..... I will keep it for myself and the others will continue their odissey of NOTHING! A proverb says: "Never give the gems to the pigs!" and it's really true!!
Last edited:
I am still waiting about some proof, all your statements are not well documented. A phd should know better.....
Some reading for the rest of us. Just for entertaining
http://www.tubesoundelectronics.de/tabellen/messtabelle_ot_etf2009.pdf
I have some old permalloy for free, modern materials and c-cores are so much better that i not use it anymore.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I was thinking hybrid push-pull with a solid-state phase splitter. Just for the heresy of it.
Nah. I doubt it will be a winner. Especially since your sockets for the 4p1l aren't teflon based.
I HATED all my teflon loctal sockets! The pins broke off, they wouldn't solder when they got old, and the inner sleeve would pull up when you pulled the tube out - teflon's very slippy!
I really like my "Chinese crap" ones - no problems with them.
The argument about core materials is very close to superfluous. M6 E/I core is not the equal of any of the nickle cores or amorphous cores for POWER TRANSFORMER performance. The loses are higher, no question. However, the M6 E/I core ceases to perform power functions above 500 Hz. Let me repeat that. M6 E/I core ceases to perform power functions above 500 Hz. This is important!!!!!
In E/I core, above 500 Hz, ALL power is transformed as direct antenna coupling. Direct flux coupling, for the portion of the AC signal that emits a B Field. At and above 500 Hz the core only controls the rise and dispersal time of the planar magnetic fields within the window and this is controlled by the permittivity of the material.
When core that does transform power above 500 Hz is utilized you are FORCED to decouple the antenna event. This FORCES you to loose ever more plate coupling for the E Field. In the least losey material available, amorphous C Core, you cannot have concentrically wound primary and secondary windings and you must have adequate space between layers of both windings to increase distributed capacitance. To fail to do these finds you with an AUDIO transformer with huge FR peaks.
To decouple the primary and secondary windings means you throw away ONE HALF of the information content possible from a transformer that is not designed for POWER transformation.
So, every designer of audio transformers is faced with a quandary. How metastable can you make the device, so that it will transform non linear signals properly? Each of the people involved in high performance audio transformers has made their choices based upon these trade offs, not on the core loss at 50 or 60 Hz.
And besides, HiB (M6) is not the best E/I grade available anyway. Check out Thomas & Skinner 0.009 thick M3 material.
In E/I core, above 500 Hz, ALL power is transformed as direct antenna coupling. Direct flux coupling, for the portion of the AC signal that emits a B Field. At and above 500 Hz the core only controls the rise and dispersal time of the planar magnetic fields within the window and this is controlled by the permittivity of the material.
When core that does transform power above 500 Hz is utilized you are FORCED to decouple the antenna event. This FORCES you to loose ever more plate coupling for the E Field. In the least losey material available, amorphous C Core, you cannot have concentrically wound primary and secondary windings and you must have adequate space between layers of both windings to increase distributed capacitance. To fail to do these finds you with an AUDIO transformer with huge FR peaks.
To decouple the primary and secondary windings means you throw away ONE HALF of the information content possible from a transformer that is not designed for POWER transformation.
So, every designer of audio transformers is faced with a quandary. How metastable can you make the device, so that it will transform non linear signals properly? Each of the people involved in high performance audio transformers has made their choices based upon these trade offs, not on the core loss at 50 or 60 Hz.
And besides, HiB (M6) is not the best E/I grade available anyway. Check out Thomas & Skinner 0.009 thick M3 material.
Last edited:
Hi Bas,
I know your post was tongue in cheek! But I seriously had so many problems with my teflon sockets I stopped using them. We could also go into a whole conversation of how to open up the locking mechanism. I finally found a quick way of doing it with a pliers followed by a thin-nose pliers. Any interest in going into this?
In terms of "prepping" I use a pot of flux quite a lot - dab the end of the solder into it before soldering. What else do you mean by "prepping"?
andy
I know your post was tongue in cheek! But I seriously had so many problems with my teflon sockets I stopped using them. We could also go into a whole conversation of how to open up the locking mechanism. I finally found a quick way of doing it with a pliers followed by a thin-nose pliers. Any interest in going into this?
In terms of "prepping" I use a pot of flux quite a lot - dab the end of the solder into it before soldering. What else do you mean by "prepping"?
andy
Maybe it's also nice to plot the impedance of a transformer.
Nice materials from T&S. I use 0.009" HiB(=M0) from Nicore or 0,002" SiFe from Arnold. For special transformers Amorphous or nanocrystalline.
I was thinking tonight about something i read before: "The argument about core materials is very close to superfluous"
I asked myself how big the difference would be in permeability between difference materials (just SiFe in this case) and I think between normal EI iron and HiB iron can be a factor 10 at least.
If this was not important why do people hear the difference?
I asked myself how big the difference would be in permeability between difference materials (just SiFe in this case) and I think between normal EI iron and HiB iron can be a factor 10 at least.
If this was not important why do people hear the difference?
I am pretty sure the moderators do not want us to stray into this.
However, the short form answer is that the core controls the rise and dispersal time of all leading and trailing edges of all AC signal regardless of frequency, even though the losses may preclude actual power transformation support. Everything else is controlled by the factors surrounding the E Field and how well the rise and dispersal of that field, during the momentary freeze while vector change is applied, matches that of the B Field.
More and deeper discussions can be developed here but I suggest you PM me if you want to know what little I know.
Bud
However, the short form answer is that the core controls the rise and dispersal time of all leading and trailing edges of all AC signal regardless of frequency, even though the losses may preclude actual power transformation support. Everything else is controlled by the factors surrounding the E Field and how well the rise and dispersal of that field, during the momentary freeze while vector change is applied, matches that of the B Field.
More and deeper discussions can be developed here but I suggest you PM me if you want to know what little I know.
Bud