It's explained in the text. Depends on how much gain you need. He does not show the gain calculations, but indicates that a single 6dj8 grounded cathode gain stage is probably not enough gain to drive the amp to full power output with a standard input.
It's not hard to make it flexible. Just use a single 6dj8 for the phase splitter - one triode for each channel, then use a separate dual triode for the gain stage. You can try different triodes as a grounded cathode amp in this position. Or you can use the common cathode gain stage, as he shows.
Sheldon
Riiiight...I figured the odds that OP had these particular headphones were pretty small, so I more or less ignored that option. However, the Audeze units are apparently of outstanding quality....and a regular b*tch to drive. The 2A3 route makes complete sense now, but as Sheldon indicates you'll really need to get/make a set of OPT's. I don't think the tubecad/Broskie 72Ohm 6AS7 design can be modified to take 2A3's, at least not without considerably raising the output impedance. Power output may still be just enough for those phones, but I'm really not sure. If an all-tube OTL design is preferred, then I'd stick to the 6AS7 solution as designed by Broskie. It's still all triodes, just not 2A3.
Forget about the headwize projects I linked to earlier; anything below 500mW output just isn't going to cut it.
Forget about the headwize projects I linked to earlier; anything below 500mW output just isn't going to cut it.
Balancing Act
this is what I want to build, the 2a3.
they offer the silver wire for OT, but the result is nothing better than copper wire OT.
what do you think about this amp ??
I don't think I have a willing to build mini tube amp after seeing this beast.
this is what I want to build, the 2a3.
they offer the silver wire for OT, but the result is nothing better than copper wire OT.
what do you think about this amp ??
I don't think I have a willing to build mini tube amp after seeing this beast.
You don't need balanced unless your sources have balanced outputs and you have long cable runs, but it won't hurt much other than your pocketbook.
The supply is a switching supply plus a linear supply?
The Sennheiser K1000 headphones require about 0.6V to drive them to 97dB. The Grado phones require 1mV for 98dB. Those are orders of magnitude more efficient than your headphones. Better make sure it will drive your phones to the levels you want.
Sheldon
Last edited:
it's not the balance I like, but the 2a3.
I don;t like balance either.
anyway... that's US$ 3900,- price !
sounds a lot for headphone amp.
anyone should design this... and let me know the result.
that will be a lot cheaper
btw I have 12b4 for my Preamp
I don;t like balance either.
anyway... that's US$ 3900,- price !
sounds a lot for headphone amp.
anyone should design this... and let me know the result.
that will be a lot cheaper
btw I have 12b4 for my Preamp
Here, models of headphone OPT with various secondaries winding :
SOWTER HEADPHONE TRANSFORMERS
The 8976 model with a simple SE 2A3 amp will do it.
Philippe
SOWTER HEADPHONE TRANSFORMERS
The 8976 model with a simple SE 2A3 amp will do it.
Philippe
Assuming that's .6Vrms that equates to 3mWrms into 120 Ohms. I think the Audeze units need a _lot_ more power than that.
I assume that is supposed to be 1mWrms. (1mVrms would be 31nW (yes, nano-Watt) into 32Ohms, or 1.6nW into 600Ohms). So that's not really an order of magnitude of difference, but 'just' 3 times more sensitive. The Audeze does require about 2 orders of magnitude more power than that (I guesstimate something between 250mW and 750mW for decent to quite loud levels).
That I can only agree with. This is not a 'build a headphone amp' excercise. This is more comparable to building an amp for sensitive full-size speakers, but with a somewhat higher output impedance.
Not according to their site. If the spec is true (may well be error), better have a pretty quiet amp.Grado Labs, Inc.
Indeed.
Last edited:
With the 18 ohm connection, the loadline would be about 5k. With the 75 ohm connection, it would be about 1.3k. Either could work, but distortion might be high for the latter.
The 9351 is also a possible choice, on the 40 ohm connection.
Sheldon
Good Lord, you're right, they consistently list sensitivity for 1mV, not 1mW. For the 32Ohm SR60i this would mean P = V^2/R = 0.03125uW I'm no physics expert, but I doubt that is enough energy to create enough air movement that can be detected by our ears.
Anyway, you're right, you better build a quiet amp for those phones. 0.01mV ripple on the output would still be like, what, -20dB or so? That can be pretty problematic.
Those Sowther xformers look pretty nice. I'd almost order a pair of the parafeed ones for my current budget. That way, I could get some decent (headphone) power from those ECL84's. But I think I'll stick to OTL and settle for bad impedance matching and a maximum power output of somewhere around 30mW.
Upon further review, it's gotta be an error, as they do list some at 1mW, and I assume that the construction is of a family. That would make it more like 180mV for around 98dB. Seems much more reasonable for typical amps. As you say, not enough energy to move the air. And these are not even in-ear phones.
Sheldon
guys,
now I have a crazy idea,
what about using 2A3 and 6AS7 for 1 channel ?
is this bad idea ?
or good idea ?
how to do that ?
now I have a crazy idea,
what about using 2A3 and 6AS7 for 1 channel ?
is this bad idea ?
or good idea ?
how to do that ?
It seems you're not serious about this project. If you are, then you need to calculate the voltage you want for the load you intend to drive. Then you need to know what your source output is. That's the minimum starting requirement.
Sheldon
Sheldon
I'm dead serious about this project, only I need 2A3 sound or 26.
since I have 4 quad match of 26 on my drawer.
just my question is it possible using 2A3 or 26 as a driver ? then 6AS7 as the final ?
like I said I'm no electronic guy who can calculate/design thing.
The load for 2A3/26 is 250Volt ? and 6AS7 is 200Volt ?
I have no idea what's the voltage is best for each tube.
I'm gonna use 50 Ohms Headphone.
ok, is there a software simulation for this ? design and amp ? I guess not ...
I hope so ...
since I have 4 quad match of 26 on my drawer.
just my question is it possible using 2A3 or 26 as a driver ? then 6AS7 as the final ?
like I said I'm no electronic guy who can calculate/design thing.
The load for 2A3/26 is 250Volt ? and 6AS7 is 200Volt ?
I have no idea what's the voltage is best for each tube.
I'm gonna use 50 Ohms Headphone.
ok, is there a software simulation for this ? design and amp ? I guess not ...
I hope so ...
OK, what's the sensitivity of your headphones? What is the RMS output of your source and its output impedance?
Sheldon
Sheldon
That's the problem. He has some really nice Audèze ESL units. 50 Ohms, but you'll need like 1 Wrms to drive them. That's about 7Vrms, or about 20Vpk-pk. I keep plugging John Broskie's design using 6AS7's which is specifically tailored to these power hungry headphones.
If a 2A3 sound is really needed, then the solution still stands to build any of the existing SE output stages with OPT's with an appropriate secondary.
If a 2A3 sound is really needed, then the solution still stands to build any of the existing SE output stages with OPT's with an appropriate secondary.
Anything is possible. But using a 2A3 to drive a 6AS7 doesn't make sense. A 26...well, perhaps, but the Broskie design isn't usable, as the mu of the 26 is too low and the maximum plate voltage rating would be exceeded. So some custom engineering would have to take place.
I'm with you Mastodon. The problem is that Calico is jumping all over the place. I'm trying to get the discussion back to first principles, and take it a step at a time. I was hoping to get Calico to understand his output requirements and source capabilities. From there, he can better evaluate his options and the limitations of various approaches.
You answered the first part of the question (I was hoping to get him to answer), so let's go with 7VRMS. Ok, what is the source output?
Sheldon
You answered the first part of the question (I was hoping to get him to answer), so let's go with 7VRMS. Ok, what is the source output?
Sheldon
Sheldon,
here is the spec of the headphone:
* Impedance: 50 Ohms, nominal
* Sensitivity: 91 dB/1mW
* Frequency Response: 5 Hz – 20 KHz, usable high frequency extension 50 KHz.
* Distortion: Less than 1% even at full output.
* Maximum power handling: 15W
* Maximum output: 133dB, 15W
* Active diaphragm area: 6.17 square inches.
* Maximum diaphragm excursion: 2.5mm p-p
here is the spec of the headphone:
* Impedance: 50 Ohms, nominal
* Sensitivity: 91 dB/1mW
* Frequency Response: 5 Hz – 20 KHz, usable high frequency extension 50 KHz.
* Distortion: Less than 1% even at full output.
* Maximum power handling: 15W
* Maximum output: 133dB, 15W
* Active diaphragm area: 6.17 square inches.
* Maximum diaphragm excursion: 2.5mm p-p
Last edited:
Ok, that's a start. I don't know how headphones are rated, but I assume the output is determined by the pressure level at your ear. Do you really need 133dB? That's hearing damage levels. Let's back it down to mere 121dB (120 is considered the pain threshold).
Now, for every 6dB increment the voltage changes by a factor of 2, and the power changes by a factor of 4 (2 squared). If 15W gets you to 133 dB, then at 12dB lower, the power requirement is 16 fold less (4X4), or about 1W. Let's say you need a minimum of 1 watt.
The OTL circuit described by JB requires a voltage swing of about 40V at the 6AS7 grid to get 4W output. That means a swing of 20V to get 1W output. With a u of 4, and a practical circuit gain of about 3, the 2A3 can't get you from 2.8V to 20V. So no, that won't work. You need an additional gain stage.
A single ended 2A3 will put out about 4W, but needs a swing on its grid of about 40V to do that. So you need a gain stage for that one too.
Either way, 2A3 into output transformer, or 2A3 into a 6AS7 OTL stage, another gain stage is needed.
Sheldon
Now, for every 6dB increment the voltage changes by a factor of 2, and the power changes by a factor of 4 (2 squared). If 15W gets you to 133 dB, then at 12dB lower, the power requirement is 16 fold less (4X4), or about 1W. Let's say you need a minimum of 1 watt.
The OTL circuit described by JB requires a voltage swing of about 40V at the 6AS7 grid to get 4W output. That means a swing of 20V to get 1W output. With a u of 4, and a practical circuit gain of about 3, the 2A3 can't get you from 2.8V to 20V. So no, that won't work. You need an additional gain stage.
A single ended 2A3 will put out about 4W, but needs a swing on its grid of about 40V to do that. So you need a gain stage for that one too.
Either way, 2A3 into output transformer, or 2A3 into a 6AS7 OTL stage, another gain stage is needed.
Sheldon
- Status
- Not open for further replies.