Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › Driving the TakeT H2
New Posts  All Forums:Forum Nav:

Driving the TakeT H2 - Page 6

post #76 of 181
Steven, thanks for making those measurements. Can I assume that all those voltages are RMS?
post #77 of 181
Quote:
Originally Posted by stevenkelby View Post
Interesting, thanks. Don't really understand that 100% sorry, is the primary the wires before the output transformer? What does the ouutput transformer do in a tube amp? Does it increase the voltage as with the TR2? Or something else?

Any examples of suitable amps?
The primary is the input side of a transformer, secondary the output. An output trafo steps down the voltage and the impedance while stepping up the current. The specs of the TR2 are 16 ohm input and a 7000 ohms output. To find the ratio you divide the numbers and take a square of the result i.e. 20.92. So before the output transformer is high voltage, low current and high impedance, the perfect environment for Heil and electrostatic drivers.

Since the Heil is run single ended you could use a low powered SET amp and install a line from the primary to the headphones. There are hundreds of designs, kits and ready made amps out there and Single Power shouldn't have any problem building one, though it could take years...
post #78 of 181
Hey Serk, you've put in more effort than me with this! Those are RMS readings yes. With the sine wave readings (the list) I adjusted the volume knob to get the 1st value. Listening to music I just found the loudest bit of a song and looped 10 seconds of it. The values were the highest I saw. The meter refreshes pretty fast, maybe 10 or 20 times a second?

This is it:

http://www.jaycar.com.au/productView...Max=&SUBCATID=

Can't wait to hear these things at their full potential.
post #79 of 181
300 Vrms is definitely in the single side of a Blue Hawaii territory.
Or in balanced mode, with plenty of headroom available.

I was starting on a all solid state push pull design similar to the
dynahi, with + and - 125 volt rails, using mje15034/mje15035 as
the out transistors, biased at about 50ma. But you would have
to do this a balanced to get enough voltage, and still really not
enough. Push pull designs are pretty much limited to rail voltages
like this because higher voltage complementary devices don't
exist, and controlling the output bias stage current is massively
tricky, and the slightest mistake, and lots of magic smoke.
So a balanced or unbalanced current sourced single ended design
makes much more sense.


So a KGSS with the bias turned up and plenty of heatsinking or
a KGBH-SE seem like the right answer.

Due to the highly capacitive nature of the headphones, you are
unlikely to find a transformer that drives both the bass and high
frequencys correctly.
post #80 of 181
Quote:
Originally Posted by spritzer View Post
The primary is the input side of a transformer, secondary the output. An output trafo steps down the voltage and the impedance while stepping up the current. The specs of the TR2 are 16 ohm input and a 7000 ohms output. To find the ratio you divide the numbers and take a square of the result i.e. 20.92. So before the output transformer is high voltage, low current and high impedance, the perfect environment for Heil and electrostatic drivers.

Since the Heil is run single ended you could use a low powered SET amp and install a line from the primary to the headphones. There are hundreds of designs, kits and ready made amps out there and Single Power shouldn't have any problem building one, though it could take years...

Great thanks for that info, explains a lot and makes sense to me now. I did learn of these things many years ago, but have since forgotten, obviously.

Here is a part of a message I got from someone:

Tube amps are not designed to power a load off their high voltage output. While 1000 ohm is not much of a load, it still doesn't seem like the right thing to do and you are probably going to hear a lot of noise (hum and tube rush noise).. Impedance matching is critical, and I'm not so sure that running off of the high voltage output of a tube amp is going to be a good match with 1000 ohm... but I might be wrong about this. Takei would know best about his design (I hope).



What do you think of those thoughts?

Thanks.
post #81 of 181
Quote:
Originally Posted by kevin gilmore View Post
300 Vrms is definitely in the single side of a Blue Hawaii territory.
Or in balanced mode, with plenty of headroom available.

I was starting on a all solid state push pull design similar to the
dynahi, with + and - 125 volt rails, using mje15034/mje15035 as
the out transistors, biased at about 50ma. But you would have
to do this a balanced to get enough voltage, and still really not
enough.

So a KGSS with the bias turned up and plenty of heatsinking or
a KGBH-SE seem like the right answer.

Due to the highly capacitive nature of the headphones, you are
unlikely to find a transformer that drives both the bass and high
frequencys correctly.

Thank you for that Kevin, answered my questions.

Damn, you said the words I didn't want to hear.

"Blu$e Haw$ii"

Anyone want to buy some ED9s?

(JK)

(semi)

Surely there's a cheaper way? Anyway I'm off to bed now.
post #82 of 181
Quote:
Originally Posted by spritzer View Post
Since the Heil is run single ended you could use a low powered SET amp and install a line from the primary to the headphones. There are hundreds of designs, kits and ready made amps out there and Single Power shouldn't have any problem building one, though it could take years...

You still have to put a resistive load on the output of the transformer
or all sorts of bad things happen. But yes this definitely works and
probably sounds very good. Typically for most of those tiny SET amps
the B+ is only 250 to 350 and that is not going to be enough either.
You are going to need a B+ of about 800 volts. Also you are going
to need an output cap, because the plate of the tube is going to
sit at half of the B+ and the headphones might definitely not like that
amount of DC.
post #83 of 181
Quote:
Originally Posted by stevenkelby View Post
Great thanks for that info, explains a lot and makes sense to me now. I did learn of these things many years ago, but have since forgotten, obviously.

Here is a part of a message I got from someone:

Tube amps are not designed to power a load off their high voltage output. While 1000 ohm is not much of a load, it still doesn't seem like the right thing to do and you are probably going to hear a lot of noise (hum and tube rush noise).. Impedance matching is critical, and I'm not so sure that running off of the high voltage output of a tube amp is going to be a good match with 1000 ohm... but I might be wrong about this. Takei would know best about his design (I hope).



What do you think of those thoughts?

Thanks.
This is why we need an output cap and a resistive load. It's just a single cap and a resistor in the signal path and a resistor to ground.

Quote:
Originally Posted by kevin gilmore View Post
You still have to put a resistive load on the output of the transformer
or all sorts of bad things happen. But yes this definitely works and
probably sounds very good. Typically for most of those tiny SET amps
the B+ is only 250 to 350 and that is not going to be enough either.
You are going to need a B+ of about 800 volts. Also you are going
to need an output cap, because the plate of the tube is going to
sit at half of the B+ and the headphones might definitely not like that
amount of DC.
OTL does sound very good if it has enough power. The output cap is a necessary evil but it will be better then a TR2.

A single Blue Hawaii sounds like a good plan and those with deeper wallets and less sanity can go for a DHT based solution.
post #84 of 181
It occurs to me that balanced drive may not be a great idea - with the bimorph arrangement the H2 could have the outside electrodes grounded and be designed to meet minimum safety standards/be safe with only the inner layer driven to high V

Is the construction safe with the outer electrodes swinging +/- 2-300 V peaks?

Maybe with the H2 or other transformer the isolated secondaries give some extra safety but an active balanced amp would happily push those V right through your head to gnd


http://www.head-fi.org/forums/showpo...&postcount=131

shows the secondary of the h2 grounded
post #85 of 181
The dielectric strength of the film is 1kv so that should be fine but I'm not so sure about the wiring.
post #86 of 181
Quote:
Originally Posted by stevenkelby View Post

I measured the voltage either side of the TR2 transformer, with a 440Hz sine wave.

With 1V in, I got 23.1 out with the H2 connected and 33.7 with the H2 disconnected.

With 10V in, I got 200 out with the H2 connected and 290 with the H2 disconnected.

With 15V in, I got 303 out with the H2 connected and 420 with the H2 disconnected.

Then tonight, after I realized that last night the mute button was in, lowering the output by 20db last night, I took the same measurements with the same results, but did it in a quiet room and noticed that the transformer starts to hum at 5V input and it is very loud at 15V input. I only did that for a sec or 2 at a time and didn't push any higher as I didn't want to damage the TR2.

I then disconnected the TR2 and the Yamaha amp can put out 60V p-p with the 440Hz sine.

Listening to the H2 without the TR2, straight out of the Yamy at full pelt with it's possible 60V p-p, Brahms still didn't have near reference level SPL (105db?) but gave me serious chills nonetheless.

Losing the TR2 improves the sound no end for the same volume level. Deeper bass, higher treble, far more clarity. Neither make me happy enough though.

I did some more tests for Kevin and emailed the results.
post #87 of 181
So i have analyzed all the data and have come to the conclusion
that most of the numbers previously posted are correct, these things
like power and lots of it, the transformers are woefully inadequate to
supply anything like decent bass... And yes they really do suck power.
You would need transformers 4 times that physical size to actually
do the job correctly. Fact is that these things are about a 65 degree lagging
load.

So here are the answers. (for OTL amplifiers)

The KGBH-SE would work absolutely great out of the box with no
modifications except for a stax to take-t adapter.

The Singlepower ES-1 would require modifications to work correctly.
Those mods would include replacing the plate resistors in the output
stage with real current sources, and turning up the bias to
30ma per tube. Then possibly some extra tweeks to keep the dc output
level correct. And possibly a much bigger high voltage power supply.

All the stax amps and things like them would work, but the bass would
be very thin and possibly distorted. Stax transformer boxes would be
completely unacceptable.

If you like solid state, you can do this.
http://gilmore.chem.northwestern.edu/take-t.gif

This output stage replaces stage 4 on a KGSS. Please take note
that 40ma is according to my spice spinner the ideal output bias
current (minimum distortion).
THIS IS ABOUT 56 pure class A watts per channel.
Think large heatsink. LARGE HEATSINK. And the power supply would
have to be beefed up in addition. It would be very sweet.
post #88 of 181
It's great to have this all cleared up. I guess the H2's have more in common with the Jecklin Floats then the Stax/Sennheiser phones so the amp needs to be capable of more power delivery.

Does anybody know the TakeT XLR pinout so I can post the directions for a Taket->Stax adaptor?
post #89 of 181
Quote:
Originally Posted by kevin gilmore View Post
This output stage replaces stage 4 on a KGSS. Please take note
that 40ma is according to my spice spinner the ideal output bias
current (minimum distortion).
THIS IS ABOUT 56 pure class A watts per channel.
Think large heatsink. LARGE HEATSINK. And the power supply would
have to be beefed up in addition. It would be very sweet.
Kevin:

Is the ideal 40mA TakeT specific, or would this work (well) in the general electrostatic case? would would you expect to be the total current demands on the power supply?
post #90 of 181
Omega 2's are a 120 picofarad load.
at 10khz they are a 170k load.

The Take-T are a 60000 picofarad load.
at 10khz they are less than a 3k load.

You can certainly build an amp that drives the take-t well
and therefore would drive Omega2's. It would probably be
a severe waste of power to do so.

I figure that a KGSS appropriately equipped to drive take-t
would pull about 200 watts from the wall socket. All of which
would end up as heat one way or the other.

The 40ma number is where the lowest distortion of a modified
KGSS and a pair of take-t, taken as a system is.

For 6ca7's the number is 30ma.

To stay in class A all the time for those headphones requires 25ma.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Headphone Amps (full-size)
Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › Driving the TakeT H2