It's only an opinion
Sheesh Joel ! You should take things less seriously . I have given you my explanation and believe you me I have tried a hell of a lot of 9 pinners in my time , like I said I prefer the sound of octals and loctals , it's my preference , folks may have different tastes . Up to you what you like , I'm not going to try to force a different opinion from you
316a
Sheesh Joel ! You should take things less seriously . I have given you my explanation and believe you me I have tried a hell of a lot of 9 pinners in my time , like I said I prefer the sound of octals and loctals , it's my preference , folks may have different tastes . Up to you what you like , I'm not going to try to force a different opinion from you
316a
Here's one guy's set of measurements for a number of tubes. Unfortunately they're only for one bias point, but I think they're interesting anyway, especially as they include spectra.
http://users.rcn.com/joepage/
What would be an interesting experiment, would be for someone to build an automated rig to test a variety tubes over a range of operating conditions and frequencies, including distortion spectra. Comparison of their performance against the most used operating points for different types could be very revealing. Or not.
http://users.rcn.com/joepage/
What would be an interesting experiment, would be for someone to build an automated rig to test a variety tubes over a range of operating conditions and frequencies, including distortion spectra. Comparison of their performance against the most used operating points for different types could be very revealing. Or not.
Re: Interesting, but....
Why? I've always measured amps to try to minimise distortion in them. Measuring tubes would be even better as it would allow a more detailed perspective during the design process. A distortion spectra vs load vs op point vs output swing would remove a lot of hype and mumbo jumbo that goes along with audio. If f'rinstance, when testing 6SN7s and looking at the result of my proposed test, it might become clear that some of the more popular op points and loads used for this tube favour a particular spectrum, then it would be easier to design an amp for that particular sound. Or you could look at another op point and load to get more or less distortion to taste. My preference us usually for less, but I would rather have more 2H only, than a lower % figure with a wide spectrum. It would also help see what's going on in cascade stages, with how the distortions on each stage modulate each other and how the stages work individually and together over a variety of conditions.
The last point touches on one of my pet hates. Many people will look at the distortion figures of a tube amp, and using your SE845 as an example, say 5% at 40W, yuk, it must be terribly coloured. When looked at in the context of it's usage in the system, perhaps a 10W peak is about your max typically, with much of the important musical information being 40 - 60 dB below this. What's your amp's distortion spectra at 1/100W? And how does that compare to a class AB1 60dB NFB SS amp (your typical receiver) at that level? Typical SS amps can be really ugly down there, and if you really want to experience that, try listening to one on some horns.
Good and experienced designers can do all of this with their ears, but the test gear would be a useful tool in helping to minimise (or at least make conscious) biasses when listening to gear in a development or evaluation sequence.
The only absolute sound that exists in audio is the one you absolutely love to listen to.
A test setup like I described wouldn't be too hard to build, but I can't do the software side of it, so I've been reluctant to try it.
dhaen said:
Why? I've always measured amps to try to minimise distortion in them. Measuring tubes would be even better as it would allow a more detailed perspective during the design process. A distortion spectra vs load vs op point vs output swing would remove a lot of hype and mumbo jumbo that goes along with audio. If f'rinstance, when testing 6SN7s and looking at the result of my proposed test, it might become clear that some of the more popular op points and loads used for this tube favour a particular spectrum, then it would be easier to design an amp for that particular sound. Or you could look at another op point and load to get more or less distortion to taste. My preference us usually for less, but I would rather have more 2H only, than a lower % figure with a wide spectrum. It would also help see what's going on in cascade stages, with how the distortions on each stage modulate each other and how the stages work individually and together over a variety of conditions.
The last point touches on one of my pet hates. Many people will look at the distortion figures of a tube amp, and using your SE845 as an example, say 5% at 40W, yuk, it must be terribly coloured. When looked at in the context of it's usage in the system, perhaps a 10W peak is about your max typically, with much of the important musical information being 40 - 60 dB below this. What's your amp's distortion spectra at 1/100W? And how does that compare to a class AB1 60dB NFB SS amp (your typical receiver) at that level? Typical SS amps can be really ugly down there, and if you really want to experience that, try listening to one on some horns.
Good and experienced designers can do all of this with their ears, but the test gear would be a useful tool in helping to minimise (or at least make conscious) biasses when listening to gear in a development or evaluation sequence.
The only absolute sound that exists in audio is the one you absolutely love to listen to.
A test setup like I described wouldn't be too hard to build, but I can't do the software side of it, so I've been reluctant to try it.
Newb's comment
Brett,
As a complete novice here & still doing a lot of reading I thought that it was interesting how the tubes measured did not tend to lose gain over the audio spectrum & if anything seemed to have lower distortion figures @ 20KHz than 50Hz.
From what I understand SS can have higher distortion @ higher freq. due to less feedback?. (bearing in mind that I only looked at 10 or so of the small signal tube results)
Cheers
Paul
Brett,
As a complete novice here & still doing a lot of reading I thought that it was interesting how the tubes measured did not tend to lose gain over the audio spectrum & if anything seemed to have lower distortion figures @ 20KHz than 50Hz.
From what I understand SS can have higher distortion @ higher freq. due to less feedback?. (bearing in mind that I only looked at 10 or so of the small signal tube results)
Cheers
Paul
Tube distortion figures
I thought I would put in my .02 because I just checked the distortion of several tubes amps. I figured I would see mega distortion figures but actually they were below 1% and it doesn't seem noticeable as compared to SS gear. When SS gear is at 1% I can really hear the distortion and it isn't pleasant to the ear.
Like I said...Just my .02
Joe
I thought I would put in my .02 because I just checked the distortion of several tubes amps. I figured I would see mega distortion figures but actually they were below 1% and it doesn't seem noticeable as compared to SS gear. When SS gear is at 1% I can really hear the distortion and it isn't pleasant to the ear.
Like I said...Just my .02
Joe
Re: Re: Interesting, but....
Yes, and this was brought into bold relief when I did a listening comparison on my Altec's between my two solid state amps and my tube gear. It is a quite obvious difference, and I'm not one to fall for "voodoo".
And you are absolutely on the money - if people are truly concerned about distortion figures, then the question they should all be asking is "what is my THD at 80mW?"!
Brett said:
Yes, and this was brought into bold relief when I did a listening comparison on my Altec's between my two solid state amps and my tube gear. It is a quite obvious difference, and I'm not one to fall for "voodoo".
And you are absolutely on the money - if people are truly concerned about distortion figures, then the question they should all be asking is "what is my THD at 80mW?"!
Brett,
I agree with everything said in your last post.
What I was thinking about was the enormous variation from valve to valve, both when new, and as they age.
The sheer amount of data that would result, would be daunting!
There is also the issue of distortion "cancellation". Improving one stage in isolation can cause an overall deterioration. I'm not saying this is a good situation, but it does happen. The last time I was in a position to test an amp on a spectrum analyser, I found that rediculously small changes - even in ambient temperature changed the distortion spectrum. That must have been a cancellation effect.
I don't want to put anyone off. Valves are imperfect, but their attractions for me outweigh the defects.
Cheers,
I agree with everything said in your last post.
What I was thinking about was the enormous variation from valve to valve, both when new, and as they age.
The sheer amount of data that would result, would be daunting!
There is also the issue of distortion "cancellation". Improving one stage in isolation can cause an overall deterioration. I'm not saying this is a good situation, but it does happen. The last time I was in a position to test an amp on a spectrum analyser, I found that rediculously small changes - even in ambient temperature changed the distortion spectrum. That must have been a cancellation effect.
I don't want to put anyone off. Valves are imperfect, but their attractions for me outweigh the defects.
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.