The all DHT SET Headphone Amp

[iko]

I agree. I'm just not sure that amorphous is the right way to go. Sometime in the future I'll try to wind my own OPT on M4-11 c-core from these guys.

[regal]

I don't have the skills to wind a quality OPT, doubt many do.

I remember Onetics was offering custom transformers with a litiz secondary that was supposed to have low insertion loss but I don't think they were c-core or amorphous.

[Pano]

Ask Bud, I'm sure he'll have plenty to say.

[BudP]

I always have plenty to say, but I don't know what I am talking about here.
I haven't read the thread and probably will not. Not arrogance, just not interested at the moment.

This is, as I understand it, a DHT SE amplifier, driving inefficient headphones? So, perhaps a 1/2 watt of power needed absolute max? Sounds like 45 tube territory to me.

People I would recommend for high resolution OPT's are:

Pieter Truffaunt of Tribute Audio
Dave Slagle of Intact Audio
Myself, of course
Possibly Per Lundahls small signal transformers, if one with enough inductance is available.
Possibly Peter Sowters nickle core items.

If there are a group of you who can decide upon one provider, you can probably get a better price. Certainly works that way with me.

Level 3 from me is M3 E/I core, ultra high performance dielectrics, passively self demagnetizing core structure and Litz wire secondary. Cost would be a bit over $500 each for a pair for a 45 tube SE amp. Could be less for quantities over 6 pieces and pricing for 300 B or 2A3 alternatives would be similar.

I will provide quite a bit more internal gradient structure for notes and transients than other makers. I will also accept slightly more distortion than they will. So, slightly softer overall sound, with quite a bit more information than you will get from an equivalent amorphous core transformer. My less expensive levels sound as good, but without as much gradient information being transformed. You won't hate them and level 1 would cost $154 each.

Sorry if this upsets normal policies, but you should have some pricing guidelines in mind if you intend to go looking for the very best performance available. None of the folks I have mentioned will disappoint you. Each will provide a different solution, with performance good enough that the choice becomes one of personal taste. I know Pieter frequents diy Audio so perhaps you can find him and invite him in for comment. I don't know about Dave, but I would be surprised if he wasn't also available.

Bud

[regal]

Thanks bud our issue is losses with big secondaries. For the new inefficient headphones the max is roughly 1W at 38 ohms. My current "custom" tranformer with a 32 ohm secondary (not from you) has a secondary DCR of 12 ohms, so much power is "wasted" as to force the tube to swing into non linear parts of the loadline. How would you deal with that issue that a headphone secondary presents? One of our own headphone amp innovators strongly advices nickle parafeed to keep the trafo small, efficient and low in copper dcr. But most of us are a bit parafeed shy, with the difficulty building a silent higher voltage B+ as well as the cost oof the parafed cap, and the whole tank rsoance issue.

[BudP]

Regal,

What you need is either amophous core C core or rather large E/Core. A parafeed amplifier can be quite good. You must have an audio choke, as opposed to a power choke. I make this kind, perhaps some others do also. As for the capacitor, Reliacap multi cap PPFX will equal a very good OPT in performance, with the caveat that amplitude causes bloom in location in a perceived field of sound. Not as big an issue with headphones as it is with speakers. Maybe Doc's designer has something like this laying around. Certainly my first choice for a parafeed amp designer.

May I ask why you aren't thinking out of the box here? How about lateral Fets driving a transformer? The primary inductance needed for the transformer would be very low, hence low secondary DCR too. Likely a single gain stage, running the Fet at very low tube voltages. Could be PP or SE and the sound will rival the best a 45 tube can do. I know this for a fact. I have heard amplifiers built along these lines. From Gary Pimm, a superbly musical build, all solid state, all transformer coupled. Another built by Rene Jaeger (think Grass Valley group, Micro Sonics, Berkely Alpha Dac, HDCD, etc.) using a SE 300 B tube as a driver stage and PP lateral Fets as the power stage. Transformer coupled and gorgeous sound quality. For that matter, a driver stage from one of Papa's amps, into a 4:1 line level buffer transformer might do.

You really need to get primary turns down in number. This requires high permeability, large size, or low impedance (250 ohms) and moderate voltages.


Bud

[regal]

Bud thanks for the info. Not really interested in Mosfets as the commercial sector has a $350 USA made mosfet amp called the Lyr that makes a diy Mosfet headamp make about as much sense as homemaking paper. But DIYing a DHT output headamp makes economic sense since thee are rare and expensive in the commercial sector.

With a headphone like this we are still looking for the best 2 stage combination. The 45 doesn't fair too well. The sims show its better to use higher owered valves since they are more linear swinging a smaller % of there max Plate Vp-p. The 307A triode connected looks the best bang for the buck if you can find them. But they run at a voltage a maybe a bit too high for parafeed.

For single feed, would you also recommend an amorphous C-Core, or d you think a EI M series core with your livitz secondary could keep up with parafeed for low level detail?

Attached is a cmparison of THD at a few millwats to 1W for various 2 stage combination as predicted by LTspice under extremely ideal conditions.

[regal]

Added 4p1l-307A and Gain. If anyone sees what looks like an error or counter-intuitive let me know. These are more relative numbers than anythig as it is simple spice. If anyone has a model for the 841 (old thoriated fil valve with gain of 30) please let me know.

[Bas Horneman]

Originally Posted by BudP

Pieter Truffaunt of Tribute Audio

Pieter Treurniet.

[regal]

Fixed typo in chart post #508 had a typo. Added 6sn7-307a