Solid State and Tubes, Ohms and Watts
Feb 4, 2010 at 9:31 PM Thread Starter Post #1 of 14

eneloquent

Head-Fier
Joined
Dec 31, 2009
Posts
99
Likes
10
I think I've got it now - the idea of how impedance works with cans.

Now, I'm trying to figure it out with amps, and I'm hoping the answer is way more obvious than it seems to me.

I began trying to figure this out while looking for a new jack to plug my HD580s into. What I seem to have discovered is;
  1. Solid state amps, are capable of passing more voltage, the lower the impedance
  2. Tube amps, are capable of passing more voltage, the higher the impedance
I'm going to use Little Dot for examples here, but it seems consistent with desktop amps from other makers I've looked at.

LD mk V;
# Power Output:
# 187 mW @ 300 ohm
# 250 mW @ 120 ohm
# 500 mW @ 32 ohm

LD mk III;
# Power Output:
# 350 mW @ 300/600 ohm
# 300 mW @ 120 ohm
# 100 mW @ 32 ohm

The practical application seems that a SS amp such as this is better for lower impedance, harder to drive headphones such as the HE-5 (25 ohms, 87 dB 1mW sensitivity) while the Tube amp would be better suited to higher impedance phones such as Senns or even the DT-880 600ohm.


From the perspective of pure curiosity, can someone explain why this seems to be true? Is it something inherent to the way a SS amp generates power vs. the way a tube functions?

Are there any exceptions to this rule? (a SS amp that scales its output up with ohms, or can even attempt to drive a 300ohm set with at least 200mw)

Am I way off base here?
 
Feb 4, 2010 at 10:22 PM Post #2 of 14
Not necessarily true.

WA22 put out 2 watts into 32 ohms when using the higher power Sylvania 7236 tubes. I read somewhere that the WA6 is about 380 mw @ 32 ohm, and WA6SE is about 1.2 watts.
 
Feb 4, 2010 at 11:16 PM Post #3 of 14
Not exactly. Different designs of both categories will have different power output characteristics. The example you gave (MkIII) is an OTL-type tube amp, and they definitely shine when outputting into higher impedances. In fact, they'll not only put more power out, they'll heat up less themselves. But it's the specific design that's responsible, not the fact that it's a tube amp.
 
Feb 5, 2010 at 2:29 AM Post #4 of 14
Right - it's the design that dictates what is possible. But the two types of devices certainly differ. Here is a simplified view of how things work:

> Vacuum tubes are ideal as voltage sources. They do not produce a declining amount of voltage as impedance increases. In an OTL design, the tube itself drives the load, and will deliver different amounts of voltage into different loads, and will generally deliver less voltage into lower impedance loads, because it cannot increase its current delivery. When using an output transformer in a tube amp design, you can use it to step-up the current delivery using the transformer. Some tube amps will use different "taps' on the transformer so that the same voltage can be delivered into different loads.

> Solid state devices are ideal as current sources. They are rated to deliver a certain voltage into a certain load, but they will deliver more power (Watts) into decreasing loads up until the point the current capabilities of the device are reached. Looked at the other way, though, they are able to produce less and less power as the impedance rises.

You can design a SS amp with way more than enough power to drive 600 ohm headphones to the point of blowing them to kingdom come. But the actual output devices have to be up to the job - many of the op-amps used in headphone amps simply aren't.

It's also possible to build tube amps that will deliver several hundred watts into 8 ohm speaker loads. But here again, this usually requires a giant amp FULL of tubes.
 
Feb 5, 2010 at 2:33 AM Post #5 of 14
A-ha! So my assumption was wrong. Good!

Are there any SS amps in the same league (< $500) that can kick 200mW+ against 300ohm+?

Edit:

Quote:

Originally Posted by Skylab /img/forum/go_quote.gif
But here again, this usually requires a giant amp FULL of tubes.


ENGspace.jpg


My god, it's full of tubes!

... had to be done.

Thanks for the detailed explanation. In essence, it's not that these trends are a limitation of the architecture, but merely the most convenient/most straightforward implementation of said architecture?

The design/tech required to make a SS amp that could drive the hell out of a pair of 600ohm cans is feasible, but it requires scaling up the model - pushing the limits, and probably using some fairly expensive and uncommon parts?
 
Feb 5, 2010 at 2:47 AM Post #7 of 14
Not sure about amps under $500 - but the Audio-GD Phoenix can deliver 900mw into 600 ohms
biggrin.gif
 
Feb 5, 2010 at 4:29 PM Post #8 of 14
Quote:

Originally Posted by eneloquent /img/forum/go_quote.gif
I think I've got it now - the idea of how impedance works with cans.

Now, I'm trying to figure it out with amps, and I'm hoping the answer is way more obvious than it seems to me.

I began trying to figure this out while looking for a new jack to plug my HD580s into. What I seem to have discovered is;
  1. Solid state amps, are capable of passing more voltage, the lower the impedance
  2. Tube amps, are capable of passing more voltage, the higher the impedance
I'm going to use Little Dot for examples here, but it seems consistent with desktop amps from other makers I've looked at.

LD mk V;
# Power Output:
# 187 mW @ 300 ohm
# 250 mW @ 120 ohm
# 500 mW @ 32 ohm

LD mk III;
# Power Output:
# 350 mW @ 300/600 ohm
# 300 mW @ 120 ohm
# 100 mW @ 32 ohm

The practical application seems that a SS amp such as this is better for lower impedance, harder to drive headphones such as the HE-5 (25 ohms, 87 dB 1mW sensitivity) while the Tube amp would be better suited to higher impedance phones such as Senns or even the DT-880 600ohm.


From the perspective of pure curiosity, can someone explain why this seems to be true? Is it something inherent to the way a SS amp generates power vs. the way a tube functions?

Are there any exceptions to this rule? (a SS amp that scales its output up with ohms, or can even attempt to drive a 300ohm set with at least 200mw)

Am I way off base here?



It seems a bit like you are confusing voltage for power here, or I'm misinterpreting your post.

It basically works according these two formulas: P = V * I and V = R * I.
Where P is power, V is voltage, I current and R is resistance/impedance. To simplify this I'll just pretend I'm talking about DC here. In a simplified overview, an amplifier is just a voltage source and the voltage it puts out is proportional to its input voltage.
attachment.php

Here I made a simple model of an amplifier. You have the input voltage V, that's actually the voltage supplied by the source and that's probably the signal coming out of your DAC. That voltage signal is amplifier by a factor A[size=xx-small]g[/size], let's say it amplifies the signal by a factor of 5 (that's 5 dB). That's would then be the voltage difference as created at the terminals of the little diamond. The amplifier itself does have an output impedance though, that would be R[size=xx-small]o[/size] in this figure. The headphone itself would then be connected to the output terminals on the right hand side of the picture.

The voltage across the headphone would then be a factor of Ag, V and Ro and the actual voltage across the headphone is:

V[size=xx-small]headphone[/size] = [R[size=xx-small]headphone[/size]/(R[size=xx-small]headphone[/size]+R[size=xx-small]o[/size])]*(V*A[size=xx-small]g[/size])

The main difference between an OTL tube amplifier and a SS amplifier is the output impedance R[size=xx-small]o[/size]. The output impedance of a SS amplifier is usually very very low (less than 0,1 Ohm), while on an OTL tube amp it can be in the hundreds of Ohms and that makes a major difference in the amount of power the OTL tube amp can provide into lower impedance loads.

As an example: Suppose we have an OTL tube amp with an output impedance of 300 Ohm, a headphone of 30 Ohm, a headphone of 300 Ohm and the OTL tube amp is capable of generating a 10 Volt signal. In the case of the 300 Ohm headphone, the headphone would be getting a 5 Volt signal from the amplifier (calculate that yourself with the previously given formula). In the case of the 30 Ohm headphone, that headphone will be getting a ~0.9 Volt signal. That's a major difference in the actual voltage the headphone gets, while the amp itself is trying to generate the same signal for both headphones. The amount of power both headphones would get (P=V²/R) would then be; The 300 Ohm headphone gets 83 mW and the 30 Ohm headphone gets 27.5 mW.

A SS amp on the other hand could have an internal resistance of 0.1 Ohm for example. That would mean that our 300 Ohm headphone will actually get a voltage of 10 Volt (9.99666.... to be more precise). The 30 Ohm headphone will have a voltage of
9.96677 Volt. That means our 300 Ohm headphone gets 333 mW and our 30 Ohm headphone gets 3311 mW. That's quite a different story as what we had with the OTL tube amp.

This was all just a bit simplified and there is more to this. It is mainly meant to give you an idea of how it all works with regard to voltage, power, impedance and current. You also have to take into account that there is a huge variety of tube amps and a huge variety of SS amps and you shouldn't really generalize them. There are tube amps which can drive low impedance headphones just as well as some of the better SS amps and there are SS amps which can drive high impedance headphones just as well as some of the more expensive tube amps.

Quote:

Originally Posted by Skylab /img/forum/go_quote.gif
Right - it's the design that dictates what is possible. But the two types of devices certainly differ. Here is a simplified view of how things work:

> Vacuum tubes are ideal as voltage sources. They do not produce a declining amount of voltage as impedance increases. In an OTL design, the tube itself drives the load, and will deliver different amounts of voltage into different loads, and will generally deliver less voltage into lower impedance loads, because it cannot increase its current delivery. When using an output transformer in a tube amp design, you can use it to step-up the current delivery using the transformer. Some tube amps will use different "taps' on the transformer so that the same voltage can be delivered into different loads.

> Solid state devices are ideal as current sources. They are rated to deliver a certain voltage into a certain load, but they will deliver more voltage into decreasing loads up until the point the current capabilities of the device are reached. Looked at the other way, though, they are able to produce less and less voltage as the impedance rises.

You can design a SS amp with way more than enough power to drive 600 ohm headphones to the point of blowing them to kingdom come. But the actual output devices have to be up to the job - many of the op-amps used in headphone amps simply aren't.

It's also possible to build tube amps that will deliver several hundred watts into 8 ohm speaker loads. But here again, this usually requires a giant amp FULL of tubes.



That part about SS amps is not correct. SS amps have pretty similar behavior as tube amps, but their output impedance is much lower and it's easier to make a SS amp that provides more current than tube amps. A SS amp will, just like a tube amp, provide more voltage into a higher impedance load than a lower impedance load.

One more thing, a SS amp can actually be made to provide similar voltage swings as a tube amp. You could make a SS amp with a voltage swing of 1000 Volts if you wanted to. Not very useful for audio use though and it would be pretty expensive. They do use SS systems for powerlines though, but then we're talking about hundred thousands of volts though and you'll have to take special measures as you're exceeding the breakdown voltage of semiconductor devices.
Quote:

Originally Posted by Skylab /img/forum/go_quote.gif
Not sure about amps under $500 - but the Audio-GD Phoenix can deliver 900mw into 600 ohms
biggrin.gif



Or the SPL Auditor: Specifications: Sound Performance Lab
That's 1.7 Watt in a 600 Ohm headphone.

[size=xx-small]I accidentally hit reply
frown.gif
, still working on this post. I'll finish my post later though, as I'm about to have supper right now. <--scratch that, I'm 'done'[/size]
 
Feb 5, 2010 at 4:32 PM Post #9 of 14
Quote:

Originally Posted by Skylab /img/forum/go_quote.gif
It's also possible to build tube amps that will deliver several hundred watts into 8 ohm speaker loads. But here again, this usually requires a giant amp FULL of tubes.



TubeAmpA.jpg


eek.gif
eek.gif
eek.gif
 
Feb 7, 2010 at 12:55 AM Post #10 of 14
Quote:

Originally Posted by Helmore /img/forum/go_quote.gif

That part about SS amps is not correct. SS amps have pretty similar behavior as tube amps, but their output impedance is much lower and it's easier to make a SS amp that provides more current than tube amps. A SS amp will, just like a tube amp, provide more voltage into a higher impedance load than a lower impedance load.





You're right, I meant to say a SS will deliver more POWER (Watts) into the decreasing load, not more voltage.

Your whole post was excellent BTW, thanks for that!
 
Aug 9, 2011 at 7:35 AM Post #11 of 14
 
[size=10.0pt]I have no clue about how it works but last night i hooked up my Senns HD650 and my Grado SR125 to my MK III with a Grado 2 into 1 (lossless splitter).[/size]

[size=10.0pt]Here is what I found;[/size]

[size=10.0pt]SPL meter showed at the same volume[/size]

[size=10.0pt]Grado = 112  decibels [/size]

[size=10.0pt]Senns = 102 decibels[/size]

 

[size=10.0pt]As you know, Senns are 300 ohm and the Grado are 32 ohm[/size]

[size=10.0pt]So, same power why would the Grado be louder if it is harder to drive them?[/size]

[size=10.0pt][/size]

 
Aug 9, 2011 at 8:50 AM Post #12 of 14
Aug 9, 2011 at 9:27 AM Post #13 of 14
tell that to my vintage solid state amps which have an output impedance of 680ohms and can deliver up to 1w(1000mw) into that given 680ohms directly from the power amp sections.

Amps are voltage sources no matter if it's tubes,solid state or hybrids. what's most difficult with amps is low impedance driving cause of larger amount of current it requires to be given at the same output level. more current means more heat.
 
Aug 9, 2011 at 12:06 PM Post #14 of 14


Quote:
 
[size=10pt]I have no clue about how it works but last night i hooked up my Senns HD650 and my Grado SR125 to my MK III with a Grado 2 into 1 (lossless splitter).[/size]
[size=10pt]Here is what I found;[/size]
[size=10pt]SPL meter showed at the same volume[/size]
[size=10pt]Grado = 112  decibels [/size]
[size=10pt]Senns = 102 decibels[/size]
 
[size=10pt]As you know, Senns are 300 ohm and the Grado are 32 ohm[/size]
[size=10pt]So, same power why would the Grado be louder if it is harder to drive them?[/size]
[size=10pt][/size]


Another important techncial stat to consider is the sensitivity/efficiency of the headphone, often expressed as  XdB/1mW (and sometimes xdB/1mV, and do note that neither of these is a linear relationship). In a basic sense, you take your output voltage from the amp and calculate the power delivered across the headphone impedance, then calculate the volume from that power.
 
 

Users who are viewing this thread

Back
Top