OTL = Low Impedance, Transformer Coupled = High?
May 11, 2009 at 6:25 AM Thread Starter Post #1 of 32

Covenant

Headphoneus Supremus
Joined
Jun 2, 2005
Posts
2,773
Likes
61
Or is it the other way around?
confused_face_2.gif


How do tube topologies relate to the kinds of headphones you use them with? I'm under the impression that low impedance cans require plenty of current, wheras high impedance benefit from voltage swing. So which topology matches which requirement better?
 
May 11, 2009 at 7:48 AM Post #3 of 32
Things without an output transformer will be better at driving high impedance.
So transformer coupled = low.
And yes, low impedance does require more current to achieve desired power vs the same headphone with a higher resistance.
 
May 11, 2009 at 8:02 AM Post #5 of 32
Quote:

Originally Posted by glitch39 /img/forum/go_quote.gif
x3.

however, an transformer coupled amp, if designed properly, can also sound as good if not better than an OTL in driving high-z cans.



This is the part that confuses me. Looking at Woo's lineup in particular, people seem to be pairing the OTL amps with low-impedance headphones (in particular, WA2 with Grados and the like) while using the WA6 and WA6SE with Sennheisers.

I know in the case of the WA6/SE there are impedance switches and whatnot, but wouldnt it make more sense to use the topology designed to match your headphones?

I suppose what I'm trying to ascertain is that, should it always be a rule of thumb to pair the topology to the headphone, or do other factors come into play as well?
 
May 11, 2009 at 8:07 AM Post #6 of 32
Not particularly familiar with the WooAudio, but I think once you get into the higher end of things you can get amps of either type that drive phones of crazy impedance and they recommend crazy loads like 8 ohms to 6000. In which case I think that's more a case of people wanting an amp that sounds good and will drive a large number of phones rather than several okay sounding amps that drive one headphone expertly.
I think the LDMIII i have recomends something like 32 to 600, however it is OTL. It drives 770/80 okay, but IMO doesnt sound any better than a 4p straight out of the STX (perhaps even a little worse). I dont have a higher ohm phone atm to compare, but yeah.
 
May 11, 2009 at 8:16 AM Post #7 of 32
Quote:

Originally Posted by Covenant /img/forum/go_quote.gif
...should it always be a rule of thumb to pair the topology to the headphone?..


I just found one. Pair your HD600 with WA3 as I did with my HD650.
 
May 11, 2009 at 10:31 AM Post #8 of 32
OTL is easy to drive high impedance Cans but some high current OTL headphone can drive Low impedance Cans very well. Transformers out can drive low or high impedance but depend on design only.
 
May 11, 2009 at 2:46 PM Post #9 of 32
all dynamic headphones are low impedance loads compared to tube circuit impedances

some monster tubes can directly drive high Z headphones at very low efficiencies

to get the hundreds of mA a K701 or Grado might require on dynamic music peaks directly from tubes exceeds even most audiophiles' concept of proportion and requires paralleling output tubes - violating the simplicity that tube aficionados also seem to highly value

with an output transformer single output tube SE circuits or push-pull outputs can easily drive any headphone load that the transformer is wired for - including both "high" and "low" Z headphone taps is simple
 
May 11, 2009 at 2:58 PM Post #10 of 32
How an amp drives headphones depends on the output impedance of the amp. Amps have an output impedance.

The easiest way to think of this is connecting two water pipes of different diameters. If you connect two 1" pipes, then water will flow just like it did before the joint. If the pipes are different diameters, you're going to have a change in water pressure.

It's similar with impedance. If the output impedance matches the headphone impedance, then the power will flow without loss. However, an impedance mismatch affects power transfer - there is a formula for working this out.

If you want a simple rule, then always make sure the amp's output impedance is lower than the impedance of the headphones. The lower the output impedance is compared to the headphone impedance, the higher the damping factor is. You want the best damping factor possible - that is the control the amp has over the driver.

OTL tube amps tend to have a high output impedance. That's because tubes generally have a high output impedance. If you try to push a low impedance can with one, you lose a lot of power and the amp has poor control over the headphone. On exception is the Zana Deux. The 6C33C tube has a low output impedance, which lets you drive Grados from an OTL. That's why it's special - there isn't another amp on the market that does that.

Output transformers will give you a lower output impedance if you wind them for that. There are a few drawbacks. One, output transformers are expensive. Well, decent ones are. You've got to pay at least $200 for a pair, often more, just for the output transformers. Second, they steal a little power from the circuit. That's not the end of the world, but you have to plan for it. Third, transformers have their own sound, which is not the same as OTL.

Solid state always has low output impedance, so that's why solid state amps are almost always OTL. McIntosh made some transformer-coupled solid state amps, but you rarely see that configuration used.
 
May 11, 2009 at 4:08 PM Post #11 of 32
Hello Eric,

thanks a lot for your explanation, now I understood it at least half
smily_headphones1.gif


But I do have some more Questions:
You are telling that the Zana Deux is the only tube amp without transformers which has a low output impedance to drive Grados well.
I think the s2A3 has also driven the Grados quite well, but I assume it had quite a higher output impedance?
And what about the Zana Deux Transformer, will it drive the Grados even better than the ZD ?

I think I read somewhere that transformer driven amps do sound different as amps without transformers, but I have not idea if this is true and what "different" means in this case.

I am really interested in this theme as I am willing to buy an amp in this category for Grado use but do not know which way to go.
The ZD, ZDT, BA (Aficionado) or would perhaps even a s2A3 do the job quite well?

Living in Germany makes it quite impossible to go to a meet to hear the differences, so I need some advice...

Thanks for your help Jochen
 
May 11, 2009 at 10:29 PM Post #12 of 32
Quote:

Originally Posted by moonboy403 /img/forum/go_quote.gif
It's the other way around.

OTL = High
Transformer Coupled = Low

Technically speaking.



OTL means output transformerless. OTL can be low or high output impedance depending on the circuit topology. With tubes one typically uses a cathode follower circuit to attain low output impedance. OTL tube circuits ususally include a very large series capacitor in the output to block DC voltages from reaching the headphones.

Transformer coupled outputs, usually used with tubes, can also be low or high output impedance. This depends on the internal impedance of the tube and the output transformer's step down ratio.

Regarding amplifier output impedance and damping factor, there seems to be a lot of misunderstanding on this topic towards which I would like to provide some insight: Some headphones, particularly the higher impedance European models, are designed and tuned around the IEC 61938 specification for headphone amplification which is 120 ohm output impedance. The amplifier output impedance acts as a voltage divider with the complex impedance of the headphone and modifies the frequency response accordingly.

Other headphones, particularly some of the lower impedance type, like to be powered with an amplifier which has low output impedance. This way the amplifier acts more as a voltage source with much less interaction with the headphone's impedance.

High damping factor isn't necessarily desired. In my opinion, what is desired is what the headphone designer/manufacturer intended. A headphone designed and tuned for low impedance drive should be powered accordingly; likewise, a headphone designed and tuned around the IEC standard of 120 ohm should be powered by an amplifier with this output. An IEC-tuned headphone powered by a low impedance amplifier might sound lean, overly damped and thin in the bass, and bright. This is not a fault of the headphone or amplifier but a mismatch between the two.
 
May 11, 2009 at 10:43 PM Post #13 of 32
Great posts, Donald North and Uncle Erik. I am saving and compiling some of this information so that when it gets asked again in a few months
smily_headphones1.gif
 
May 11, 2009 at 10:43 PM Post #14 of 32
Quote:

Originally Posted by Uncle Erik /img/forum/go_quote.gif
You want the best damping factor possible - that is the control the amp has over the driver.


Quote:

Originally Posted by Donald North
High damping factor isn't necessarily desired. In my opinion, what is desired is what the headphone designer/manufacturer intended. A headphone designed and tuned for low impedance drive should be powered accordingly; likewise, a headphone designed and tuned around the IEC standard of 120 ohm should be powered by an amplifier with this output. An IEC-tuned headphone powered by a low impedance amplifier might sound lean, overly damped and thin in the bass, and bright. This is not a fault of the headphone or amplifier but a mismatch between the two.


I'm getting some conflicting information here. If damping factor correlates to the control the amp has over the driver, how could being "overly damped" be a bad thing?

Uncle Erik seems to be saying that an amp with a lower output impedance than the headphone being used is the right choice, wheras Doland seems to be taking the stance that the two must be matched and equal. Or am I taking too simplistic a view?
 
May 11, 2009 at 11:10 PM Post #15 of 32
Quote:

Originally Posted by Covenant /img/forum/go_quote.gif
I'm getting some conflicting information here. If damping factor correlates to the control the amp has over the driver, how could being "overly damped" be a bad thing?

Uncle Erik seems to be saying that an amp with a lower output impedance than the headphone being used is the right choice, wheras Doland seems to be taking the stance that the two must be matched and equal. Or am I taking too simplistic a view?



Yes, there are some differences between what Erik and I say
bigsmile_face.gif


First, I am not saying the amplifier output impedance and headphone impedance should be identical. I am suggesting the headphone should be powered by an amplifier with output impedance which the headphone's designer intended and used to tune them.

My background is transducer and loudspeaker design. In loudspeakers there is a dimensionless figure of merit called Q which pertains to the speaker's damping of its fundamental resonance frequency. The Q is a function of the speaker's mass, suspension compliance and resistive losses, strength of the motor (product of the voice coil and magnet structure), and electrical impedance. This Q parameter is often used also in defining the tuning of transducer in a sealed enclosure. A Q of 0.5 is defined as critically damped as the diaphragm has minimal overshoot and returns to its zero rest position as quickly as possible. A Q less than 0.5 is considered over damped, while a Q greater than 1 is usually considered under damped.

If you take any transducer and add a series resistor to it, you increase its Q and decrease its damping.

With loudspeakers, they're normally designed to see an amplifier with very low output impedance, less than 0.5 ohm. This is how 99% of the hi-fi amplifiers in the world market are. An amplifier with higher output impedance reduces the speaker's low frequency damping and modifies the speaker's frequency response. Hi-fi tube amplifiers typically have an output impedance of 1-3 ohms depending on topology and if they use feedback or not. Do note: You can design a loudspeaker to sound "right" when feed from a higher output impedance. This is what I do with my high efficiency loudspeakers designed for use with low wattage SET tube amplifiers.

Now let's look at headphones: A headphone transducer might be designed to have the right (according to its designer) damping Q when fed from a low impedance amplifier, meaning for proper sound the designer does not want the headphone to see much amplifier series output impedance to decrease the damping. Other headphone transducers are intentionally designed over damped with low Q, knowing that they'll be used with amplifiers with higher output impedance (perhaps the IEC standard of 120 ohms) which subsequently decreases the headphone's damping and brings them "in tune".

Ultimately there is no absolute right or wrong. What matters is what sounds best to you.
 

Users who are viewing this thread

Back
Top