Re: Re: After the first watt
Yes, thanks, this needs to be corrected. I actually listen at 1 meter
quite a lot using nearfield monitors...
Though I wonder if the 9db correction factor includes the room sound
added in. The so-called reverberant field reduces the expected open-
field inverse-square-with-distance fall-off in power. The directivity
of the speaker would seem to be a factor also.
There must be some nice AES papers on the subject, also easy enough
to measure in my room with my speakers.
Cheers,
Michael
Brett said:
Yes, thanks, this needs to be corrected. I actually listen at 1 meter
quite a lot using nearfield monitors...
Though I wonder if the 9db correction factor includes the room sound
added in. The so-called reverberant field reduces the expected open-
field inverse-square-with-distance fall-off in power. The directivity
of the speaker would seem to be a factor also.
There must be some nice AES papers on the subject, also easy enough
to measure in my room with my speakers.
Cheers,
Michael
Re: #14
Real loudspeaker is not a well behaved 8 ohm resistor anyway...
A well matched triode's load line is much more forgiving natural
variances of that mechanical reality.
Excessive high damping factor is sometimes working against you.
Linear voltage is not necessarily the same as square law power.
When the loudspeaker load is constantly changing...
I am agreeing with Darius this time, if I havn't yet made that clear.
oldeurope said:If you want a high "damping factor", extremely low harmonics
and excellent linearity you should go and get a solid state amp.
Real loudspeaker is not a well behaved 8 ohm resistor anyway...
A well matched triode's load line is much more forgiving natural
variances of that mechanical reality.
Excessive high damping factor is sometimes working against you.
Linear voltage is not necessarily the same as square law power.
When the loudspeaker load is constantly changing...
I am agreeing with Darius this time, if I havn't yet made that clear.
#20
Hello Brett,
please note there is a difference between triode amps and tube amps in general.
The link shows a push pull [PP] design.
PP eliminates some triode characteristics this is why PP is not welcome.
The manufacturers of vintage tube amps tried to get a high "damping factor",
extremely low harmonics and excellent linearity.
Short- they tried to get the excellent results of a today's solid state amp (on the paper).
This makes no sense to me.
These days you must point out triode characteristics.
I am sorry but for me it is very difficult explaining this in English. I hope you know what I want to say.
I did the google search for you, here it is.
Please note, it uses a pentode, connected as a pentode, a coupling capacitor and cathode bypass caps ...
To me this isn't a pure triode design.
D.
Hello Brett,
please note there is a difference between triode amps and tube amps in general.
The link shows a push pull [PP] design.
PP eliminates some triode characteristics this is why PP is not welcome.
The manufacturers of vintage tube amps tried to get a high "damping factor",
extremely low harmonics and excellent linearity.
Short- they tried to get the excellent results of a today's solid state amp (on the paper).
This makes no sense to me.
These days you must point out triode characteristics.
I am sorry but for me it is very difficult explaining this in English. I hope you know what I want to say.
I did the google search for you, here it is.
Please note, it uses a pentode, connected as a pentode, a coupling capacitor and cathode bypass caps ...
To me this isn't a pure triode design.
D.
Thank you all for your input and comments.
I guess the Axiom is not what I expected it to be. Probably not worth $1500 to find out if good or not.
Brett:
You are right in saying I did not build a amp yet. And even when I will build one, the first ones will not be of my own design. I will start with the Zen clone and the Tubelab SE as a second amp.
The Amity is Push-Pull, but I will take a look at the circuit and the documentation.
What I need after these builds, is a way to understand how it all fits toghether. I have to lab it ! What is the best set-up or lab to learn all this ? What equipment is required ? Which are the best experiments ? Maybe I should start by creating a Spud amp.
I think that part swapping will only reveal the effect of the specific component and not really the possibilities of tubes, tubes designs and capabilities. By the same token, I am convinced the components have a great role to play in the final sound.
I have been reading litterature and forums opinions for the past year, but am missing the experience. The next step is ? ...
kenpeter:
The drivers are not linear loads but can it be partially fixed with a Zobel ?
oldeurope:
thanks for the Thorsten Legacy link. Beautiful simplicity of the driver and output stages. It's the first time I see a power supply section designed this way. The power supply is in my eyes as important as the input and driver stages. In some designs, the stages are somewhat isolated from the PS, but in most cases, it will participate in the final sound. From what I can understand, Thorsten Loesch was able the get the PSU final caps deliver all the current required to drive the amp without showing the non-linearities of the devices behind them.
I did not have time to follow all the references yet, but I will sometimes at the end of this week.
Are ther other great 300B designs that I should look at or try ?
I guess the Axiom is not what I expected it to be. Probably not worth $1500 to find out if good or not.
Brett:
You are right in saying I did not build a amp yet. And even when I will build one, the first ones will not be of my own design. I will start with the Zen clone and the Tubelab SE as a second amp.
The Amity is Push-Pull, but I will take a look at the circuit and the documentation.
What I need after these builds, is a way to understand how it all fits toghether. I have to lab it ! What is the best set-up or lab to learn all this ? What equipment is required ? Which are the best experiments ? Maybe I should start by creating a Spud amp.
I think that part swapping will only reveal the effect of the specific component and not really the possibilities of tubes, tubes designs and capabilities. By the same token, I am convinced the components have a great role to play in the final sound.
I have been reading litterature and forums opinions for the past year, but am missing the experience. The next step is ? ...
kenpeter:
The drivers are not linear loads but can it be partially fixed with a Zobel ?
oldeurope:
thanks for the Thorsten Legacy link. Beautiful simplicity of the driver and output stages. It's the first time I see a power supply section designed this way. The power supply is in my eyes as important as the input and driver stages. In some designs, the stages are somewhat isolated from the PS, but in most cases, it will participate in the final sound. From what I can understand, Thorsten Loesch was able the get the PSU final caps deliver all the current required to drive the amp without showing the non-linearities of the devices behind them.
I did not have time to follow all the references yet, but I will sometimes at the end of this week.
Are ther other great 300B designs that I should look at or try ?
"Hmmm, I dont like the diodes as a cathode resistor because of
their non linearity."
Using thermionic diodes in the cathode circuit are not necessarily causing a non-linearity. They have the same non-linearity curvature function as the g1-cathode does already, so they just lower the tube's Mu factor effectively. [new effective Gm1 = 1/(1/gm1 + RPdiode) ] (the tube's output Rp is also increased similarly, but I think final effective Mu still decreases) Why the Axiom does this is still a mystery though. But one might be able to dilute out some of the associated triode's g1 "island/inselbildung effect" this way. (but since it lowers effective Mu, and a lower Mu tube would have less island effect to begin with, one might just as well use a lower Mu tube to begin with)
Putting a thermionic diode in for the actual plate load, on the other hand, is just the classic voltage mirror circuit. The Aikido uses this approach for its first gain stage. The top (load) tube simply acts to halve the bottom triode's Rp (tracking non-linear Rps), and so halves the Mu effectively. This voltage mirror loading scheme really makes more sense as a load for a pentode to re-linearize it's output however.
The Axiom however has got a thermionic diode in series with the triode gain tube's output, so this would be increasing the triode's (via tracking non-linear R) plate Rp effectively. But since the real load is then a reduced fraction of the total load, it sees a reduced voltage output. I suppose one could argue that this would dilute out the load Z variation from the tube perspective, and so linearize gain, but at a huge cost in efficiency and damping factor.
I think the Axiom is interesting for pointing out these little known design options. But whether it is beneficial here, I will leave to the listener to decide that one.
Don
their non linearity."
Using thermionic diodes in the cathode circuit are not necessarily causing a non-linearity. They have the same non-linearity curvature function as the g1-cathode does already, so they just lower the tube's Mu factor effectively. [new effective Gm1 = 1/(1/gm1 + RPdiode) ] (the tube's output Rp is also increased similarly, but I think final effective Mu still decreases) Why the Axiom does this is still a mystery though. But one might be able to dilute out some of the associated triode's g1 "island/inselbildung effect" this way. (but since it lowers effective Mu, and a lower Mu tube would have less island effect to begin with, one might just as well use a lower Mu tube to begin with)
Putting a thermionic diode in for the actual plate load, on the other hand, is just the classic voltage mirror circuit. The Aikido uses this approach for its first gain stage. The top (load) tube simply acts to halve the bottom triode's Rp (tracking non-linear Rps), and so halves the Mu effectively. This voltage mirror loading scheme really makes more sense as a load for a pentode to re-linearize it's output however.
The Axiom however has got a thermionic diode in series with the triode gain tube's output, so this would be increasing the triode's (via tracking non-linear R) plate Rp effectively. But since the real load is then a reduced fraction of the total load, it sees a reduced voltage output. I suppose one could argue that this would dilute out the load Z variation from the tube perspective, and so linearize gain, but at a huge cost in efficiency and damping factor.
I think the Axiom is interesting for pointing out these little known design options. But whether it is beneficial here, I will leave to the listener to decide that one.
Don
smoking-amp said:
Mystery ?
No,
it is just a question of timing ... nothing more I want to get triode sound,
not diode, inductor, capacitor, transformer or "what ever" sound.
A good triode amp circuit points out "triode sound" only.
is not true!
Re: #20
As for the statement "PP eliminates some triode characteristics ": it is utter rubbish. You should study the design more thoroughly.
Using SE tube amps because of some pretensions about "sonic purity" is religious dogmatism and has no basis in reality.
The amp linked sets out to maximise the linearity of the amp by utilising the most linear compomemts in the best possible way. I do not think that Lynn took DF into account and the Amity design is based around Western Electric designs of the 30's. There is no global feedback so how id it minimised?oldeurope said:
As for the statement "PP eliminates some triode characteristics ": it is utter rubbish. You should study the design more thoroughly.
Using SE tube amps because of some pretensions about "sonic purity" is religious dogmatism and has no basis in reality.
Brett, I don't think it's so much a question of "purity" than it is a matter of a philosophy on what one wants the amp to do. If the amp is supposed to have a sound, then that's a different matter than designing an amp to have as little impact as possible on the sonics. Both are valid design goals, but mutually contradictory. I know that it would be a challenge for me to design a p-p amp that colors things the same way as an SET.
First of all, let go of any idea that your decisions about amp design at this point are based upon anything real, because at this point you know nothing, which is good as this is the best place to learn from, not having already decided what is best with no experience to back it up.lowtherdream said:
Then read everything you can about tube amps. Start with www.tubecad.com , especially the Grounded Cathode Amplifiers article. Next would be Morgan Jones' Valve Amplifiers book. A nit expensive, but very, very worthwhile. The companion Building Valve Amplifiers book is also a must.
I'd also suggest reading the Library at Lynn's site: lots of useful tidbits in there.
Before you start to build, learn about hihg voltage safety. The volyages in tube amps can harm or kill. I have been bitten before and it hurts. These days I work with high voltages all the time and am very, very careful.
Then get a decent soldering station, solder sucker and electronic hand tools. Practice some soldering. Some resistors, a decent soundcard and free measurement software will be enough to get you started. A CRO would be nice, but no necessary at this point.
THEN, breadboard a couple of simple amp circuits. I would suggest not working with DHT's at the moment as they can be difficult for the beginner to get quiet. The Legacy amp I mentioned earlier could be built with a trioded KT88 at first and a 300B substituted later.
A simple cheap PP is SY's Red Light District (do a search on this board for it).
At this point, do not obsess with 'audiophile' parts.
Thanks again for the comments and patience
I do not have building experience but, as I stated earlier, my ears tell me that SET is what sounds best (for the type of music I listen to). I was never attracted by any of the PP sound or complex designs at audio shows. Is it because there are no good PP amps or because they do not know how to demonstrate their value?
Here in Montreal, there are very few stores with tube amps. When they do, it's these very expensive and powerful arrangements that do not necessarily sound that good. The difficult part is to find a combination of components that give back the subtle vibrations of the instruments, the coloration of wood or brass, the perfection of the microphone or the sound take. If my goal was to listen to Tangerine Dream or Vangelis, I would probably select not to use a SE or tube amps at all. Electronic music is probably better voiced with other options. Isn't an amp a musical instrument after all? All the components participate in the final sound, including the chassis material. In the case of an wood instrument, even the lacquer will change the sound (it is also true for an electric guitar)http://www.mother-of-tone.com/lacquer.htm (more sugar in the Kool-Aid ?).
I will follow the reading recommendations and experiment with simple SE and PP designs (to give PP a fair chance). I have read all the warnings about high voltage and scared enough to be always careful. I have the soldering station, the oscilloscope and other basic tools. I am missing the soundcard and measurement software. That's next on the list I will also probably need a high-voltage variable PS.
I know there is a lot of bad info out there (Kool-Aid), but I need to build that good filter you are talking about. It is not easy to make sense of all of this
I do not have building experience but, as I stated earlier, my ears tell me that SET is what sounds best (for the type of music I listen to). I was never attracted by any of the PP sound or complex designs at audio shows. Is it because there are no good PP amps or because they do not know how to demonstrate their value?
Here in Montreal, there are very few stores with tube amps. When they do, it's these very expensive and powerful arrangements that do not necessarily sound that good. The difficult part is to find a combination of components that give back the subtle vibrations of the instruments, the coloration of wood or brass, the perfection of the microphone or the sound take. If my goal was to listen to Tangerine Dream or Vangelis, I would probably select not to use a SE or tube amps at all. Electronic music is probably better voiced with other options. Isn't an amp a musical instrument after all? All the components participate in the final sound, including the chassis material. In the case of an wood instrument, even the lacquer will change the sound (it is also true for an electric guitar)http://www.mother-of-tone.com/lacquer.htm (more sugar in the Kool-Aid ?).
I will follow the reading recommendations and experiment with simple SE and PP designs (to give PP a fair chance). I have read all the warnings about high voltage and scared enough to be always careful. I have the soldering station, the oscilloscope and other basic tools. I am missing the soundcard and measurement software. That's next on the list
I will also probably need a high-voltage variable PS.I know there is a lot of bad info out there (Kool-Aid), but I need to build that good filter you are talking about. It is not easy to make sense of all of this
The electric potion I drink sais to me: that linearity is good
for music (within bounds), bad for clipping (out of bounds).
Transitioning from linearity with artistic distortion is the key.
Clipping should be unfaithfully rendered and masked as
necessary. Hide clipping with 2nd order (Diatonic Octave)
or 2/3rd order (Diatonic Fifth) Harmonic. Musically related
to the orignial source.
Unless you intend to isolate the root note of the clipping
(assuming it wasn't completely transient), add or subtract
only the needed harmonics to coax it away from the rail,
there will be intermodulation. But 2nd harmonic produces
less offensive IMD than 3rd or higher....
All amps clip and/or distort. Especially those with dynamic
load or source impedances (Mikes and LoudSpeakers).
for music (within bounds), bad for clipping (out of bounds).
Transitioning from linearity with artistic distortion is the key.
Clipping should be unfaithfully rendered and masked as
necessary. Hide clipping with 2nd order (Diatonic Octave)
or 2/3rd order (Diatonic Fifth) Harmonic. Musically related
to the orignial source.
Unless you intend to isolate the root note of the clipping
(assuming it wasn't completely transient), add or subtract
only the needed harmonics to coax it away from the rail,
there will be intermodulation. But 2nd harmonic produces
less offensive IMD than 3rd or higher....
All amps clip and/or distort. Especially those with dynamic
load or source impedances (Mikes and LoudSpeakers).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.