[b0dhi]

I use an unusual pair of headphones, the TakeT H2 (see here). Since they run on high voltage, normally you need to use the accompanying TR2 transformer with a speaker amp to be able to power them properly. Recently though I modified mine, with the help of Mr Takei, to be able to run off a speaker amp without the transformer at a good volume. (I replaced the internal 4.3K resistors with 100R/5W ones.)

The new sound was very impressive. It was lacking the "hard edge" of the previous configuration, although without sacrificing any of the explosive dynamics the H2 inherently have. Detail in all areas of the sound also increased.

The mod had a a side-effect though. As a result of reducing the impedance of the H2 (I'm told), the bass has has become very thin. This is a deal-breaker for me, so I'm thinking of striking a compromise by changing the resistors to 1 or 2k/10W, which should result in balanced bass, while being more sensitive than the original 4.3kOhms. At the moment I'm driving them with a speaker amp that puts out about 40V into them, resulting in a good volume. Prior to the mod, the same amp could only get them to low volumes.

The impedance of the headphones is infinite at DC, 3000R at 1khz and about 100R at 20Khz.

My long winded question is, do any amp wizards out there know of an all-tube amp design that can put out 150-200V, and enough current, into the H2? I'd want it to be OTL ofcourse, and probably wont use coupling capacitors either due to the high-Z at DC inherent in the headphones.

Alternatively, does anyone know of a ~60W "all-tube" trafo-coupled push-pull amp/DIY kit whose primary I could tap into? Or maybe something cheap to look out for on ebay?

Edit: found a few tube kits, although I have no idea which would be more suitable:
Ella KT88 EL34 PP Amplifier | Diy HiFi Supply
Kits (Edison 60 Push-pull version)

 

[nikongod]

I think Kevin Gilmore was working on a direct-drive amp for the T2's a while ago. Maybe it was solid-state.

You *will* probably need a coupling cap. The drivers may be OK for lots of DC offset, but its good to keep that safely away from your gourd. The good news is that since the load is pretty much purely reactive, you dont need as large of a coupling cap as a 100-ohm dynamic would require. you can probably use 1uf (maybe less) with great bass. If you are going to put a resistor across the cans, that becomes the target impedance for bass roll-off, it will still probably be less than 20uf.

Most of the P-P tube designs are focused on pentodes, and the plate impedance is quite high. The bass will probably be good, but the highs will suck. Even a P-P triode stage will have difficulty running this kind of load OTL without resorting to paralleling sections.

If you are willing to parallel sections, you can probably get those cans singing loud and proud with either a 6c33 or 5998 (both triodes in the bottle paralleled) push-pull driving off of the plate. [sarcasm]Ooh, your going to have such fun calculating load lines for a reactive load[/sarcasm]

 

[b0dhi]

Thanks for your response. I noticed you used the word "calculation". I'm not sure what happened after that

I have the schematic to Dr Gilmore's H2 amp around here somewhere, but I wouldn't trust myself building a high voltage amp without instructions just off the schematic without the board, particularly when the writing on the schematic isn't legible. There's also that stuff about memory distortion that makes me want to move toward tubes (see here). But also I really like the idea of running the headphones right off the the tubes.

I was actually moving from headphones into speakers recently but by chance I heard the H2 driven directly (albeit, still not optimally) without the trafo in the way and was really impressed at the improvement. As someone that's run around amp matching the GS1000, I know what a picky dynamic can be, but the H2 are something else. They can morph almost into another headphone by the way they're driven and sourced, which leads me to believe there is an exceptional amount of potential in them (no pun intended). I'm really curious to see what they can reveal when they're hooked straight up to the tubes.

Designing an amp is out of the question for me, so would either of those amps work well if I just bypassed the output transformers (coupling caps not withstanding)? Are there any other options?

 

[nikongod]

Quote:

Originally Posted by b0dhi /img/forum/go_quote.gif Designing an amp is out of the question for me, so would either of those amps work well if I just bypassed the output transformers (coupling caps not withstanding)? Are there any other options?

I dont think it will work. The 100-ohm load at high frequencies will be very rough on the tubes, and they will at the very least drop out of class-A trying to get 200V into it.

Trying to put 50v into the headphones should be doable, but the output impedance of an off-the-shelf transformer coupled amp (with transformers "bypassed") will still probably disagree with the load the headphone represents.

Im curious what the Gilmore design looks like. I am afraid of how much power it pulls out of the wall though.

 

[b0dhi]

The schematic is here: http://gilmore.chem.northwestern.edu/take-t.gif.

What would be a reasonable impedance for the plates to see? I can raise the impedance of the headphones (which I probably will do anyway, due to the bass issue). Originally the internal resistors were 4.3k, but it needed about 250-300V to get good volume. The impedances I mentioned earlier were for 100R resistors only. Would, say, a 500Ohm minimum load on the plates be ok?